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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 565 - (show annotations) (download)
Sun Oct 31 18:18:48 2010 UTC (4 years, 1 month ago) by ph10
File size: 193188 byte(s)
Added parentheses argument to -o and --only-matching options of pcregrep.

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 12. Added an optional parentheses number to the -o and --only-matching options
68 of pcregrep.
69
70
71 Version 8.10 25-Jun-2010
72 ------------------------
73
74 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
75 THEN.
76
77 2. (*ACCEPT) was not working when inside an atomic group.
78
79 3. Inside a character class, \B is treated as a literal by default, but
80 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
81 causes the error). The code is unchanged, but I tidied the documentation.
82
83 4. Inside a character class, PCRE always treated \R and \X as literals,
84 whereas Perl faults them if its -w option is set. I have changed PCRE so
85 that it faults them when PCRE_EXTRA is set.
86
87 5. Added support for \N, which always matches any character other than
88 newline. (It is the same as "." when PCRE_DOTALL is not set.)
89
90 6. When compiling pcregrep with newer versions of gcc which may have
91 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
92 declared with attribute warn_unused_result" were given. Just casting the
93 result to (void) does not stop the warnings; a more elaborate fudge is
94 needed. I've used a macro to implement this.
95
96 7. Minor change to pcretest.c to avoid a compiler warning.
97
98 8. Added four artifical Unicode properties to help with an option to make
99 \s etc use properties (see next item). The new properties are: Xan
100 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
101
102 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
103 use Unicode properties. (*UCP) at the start of a pattern can be used to set
104 this option. Modified pcretest to add /W to test this facility. Added
105 REG_UCP to make it available via the POSIX interface.
106
107 10. Added --line-buffered to pcregrep.
108
109 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
110 studied, and the match started with a letter with a code point greater than
111 127 whose first byte was different to the first byte of the other case of
112 the letter, the other case of this starting letter was not recognized
113 (#976).
114
115 12. If a pattern that was studied started with a repeated Unicode property
116 test, for example, \p{Nd}+, there was the theoretical possibility of
117 setting up an incorrect bitmap of starting bytes, but fortunately it could
118 not have actually happened in practice until change 8 above was made (it
119 added property types that matched character-matching opcodes).
120
121 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
122 possible starting bytes for non-anchored patterns.
123
124 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
125 \R, and also a number of cases that involve Unicode properties, both
126 explicit and implicit when PCRE_UCP is set.
127
128 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
129 input, it could crash or give wrong results if characters with values
130 greater than 0xc0 were present in the subject string. (Detail: it assumed
131 UTF-8 input when processing these items.)
132
133 16. Added a lot of (int) casts to avoid compiler warnings in systems where
134 size_t is 64-bit (#991).
135
136 17. Added a check for running out of memory when PCRE is compiled with
137 --disable-stack-for-recursion (#990).
138
139 18. If the last data line in a file for pcretest does not have a newline on
140 the end, a newline was missing in the output.
141
142 19. The default pcre_chartables.c file recognizes only ASCII characters (values
143 less than 128) in its various bitmaps. However, there is a facility for
144 generating tables according to the current locale when PCRE is compiled. It
145 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
146 characters, are recognized by isspace() and therefore were getting set in
147 these tables, and indeed these tables seem to approximate to ISO 8859. This
148 caused a problem in UTF-8 mode when pcre_study() was used to create a list
149 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
150 which of course cannot start UTF-8 characters. I have changed the code so
151 that only real ASCII characters (less than 128) and the correct starting
152 bytes for UTF-8 encodings are set for characters greater than 127 when in
153 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
154 altogether.)
155
156 20. Added the /T option to pcretest so as to be able to run tests with non-
157 standard character tables, thus making it possible to include the tests
158 used for 19 above in the standard set of tests.
159
160 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
161 reference to a subpattern the other side of a comment that contains an
162 opening parenthesis caused either an internal compiling error, or a
163 reference to the wrong subpattern.
164
165
166 Version 8.02 19-Mar-2010
167 ------------------------
168
169 1. The Unicode data tables have been updated to Unicode 5.2.0.
170
171 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
172 configured.
173
174 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
175 original author of that file, following a query about its status.
176
177 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
178 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
179
180 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
181 quantifier applied to a forward-referencing subroutine call, could compile
182 incorrect code or give the error "internal error: previously-checked
183 referenced subpattern not found".
184
185 6. Both MS Visual Studio and Symbian OS have problems with initializing
186 variables to point to external functions. For these systems, therefore,
187 pcre_malloc etc. are now initialized to local functions that call the
188 relevant global functions.
189
190 7. There were two entries missing in the vectors called coptable and poptable
191 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
192 I've fixed the data, and added a kludgy way of testing at compile time that
193 the lengths are correct (equal to the number of opcodes).
194
195 8. Following on from 7, I added a similar kludge to check the length of the
196 eint vector in pcreposix.c.
197
198 9. Error texts for pcre_compile() are held as one long string to avoid too
199 much relocation at load time. To find a text, the string is searched,
200 counting zeros. There was no check for running off the end of the string,
201 which could happen if a new error number was added without updating the
202 string.
203
204 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
205
206 11. \K was not working if it appeared in an atomic group or in a group that
207 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
208 \K is "not well defined" if used in an assertion. PCRE now accepts it if
209 the assertion is positive, but not if it is negative.
210
211 12. Change 11 fortuitously reduced the size of the stack frame used in the
212 "match()" function of pcre_exec.c by one pointer. Forthcoming
213 implementation of support for (*MARK) will need an extra pointer on the
214 stack; I have reserved it now, so that the stack frame size does not
215 decrease.
216
217 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
218 item in branch that calls a recursion is a subroutine call - as in the
219 second branch in the above example - was incorrectly given the compile-
220 time error "recursive call could loop indefinitely" because pcre_compile()
221 was not correctly checking the subroutine for matching a non-empty string.
222
223 14. The checks for overrunning compiling workspace could trigger after an
224 overrun had occurred. This is a "should never occur" error, but it can be
225 triggered by pathological patterns such as hundreds of nested parentheses.
226 The checks now trigger 100 bytes before the end of the workspace.
227
228 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
229
230
231 Version 8.01 19-Jan-2010
232 ------------------------
233
234 1. If a pattern contained a conditional subpattern with only one branch (in
235 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
236 computed the wrong minimum data length (which is of course zero for such
237 subpatterns). This could cause incorrect "no match" results.
238
239 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
240 the pattern is reset in the first branch, pcre_compile() failed with
241 "internal error: code overflow at offset...". This happened only when
242 the reset was to the original external option setting. (An optimization
243 abstracts leading options settings into an external setting, which was the
244 cause of this.)
245
246 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
247 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
248 assertion pattern did not match (meaning that the assertion was true), it
249 was incorrectly treated as false if the SKIP had been reached during the
250 matching. This also applied to assertions used as conditions.
251
252 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
253 assertion subpattern, including such a pattern used as a condition,
254 unpredictable results occurred, instead of the error return
255 PCRE_ERROR_DFA_UITEM.
256
257 5. The C++ GlobalReplace function was not working like Perl for the special
258 situation when an empty string is matched. It now does the fancy magic
259 stuff that is necessary.
260
261 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
262 removed. (These were left over from very, very early versions of PCRE.)
263
264 7. Some cosmetic changes to the code to make life easier when compiling it
265 as part of something else:
266
267 (a) Change DEBUG to PCRE_DEBUG.
268
269 (b) In pcre_compile(), rename the member of the "branch_chain" structure
270 called "current" as "current_branch", to prevent a collision with the
271 Linux macro when compiled as a kernel module.
272
273 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
274 prevent a collision with the Linux macro when compiled as a kernel
275 module.
276
277 8. In pcre_compile() there are some checks for integer overflows that used to
278 cast potentially large values to (double). This has been changed to that
279 when building, a check for int64_t is made, and if it is found, it is used
280 instead, thus avoiding the use of floating point arithmetic. (There is no
281 other use of FP in PCRE.) If int64_t is not found, the fallback is to
282 double.
283
284 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
285 2005 (difference between two addresses compared to an unsigned value).
286
287 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
288 custom one, because of the following reported problem in Windows:
289
290 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
291 under Win32.
292 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
293 therefore missing the function definition.
294 - The compiler thus generates a "C" signature for the test function.
295 - The linker fails to find the "C" function.
296 - PCRE fails to configure if asked to do so against libbz2.
297
298 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
299 messages were output:
300
301 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
302 rerunning libtoolize, to keep the correct libtool macros in-tree.
303 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
304
305 I have done both of these things.
306
307 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
308 most of the time, it *can* run out if it is given a pattern that contains a
309 runaway infinite recursion. I updated the discussion in the pcrestack man
310 page.
311
312 13. Now that we have gone to the x.xx style of version numbers, the minor
313 version may start with zero. Using 08 or 09 is a bad idea because users
314 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
315 interpreted as invalid octal numbers. I've updated the previous comment in
316 configure.ac, and also added a check that gives an error if 08 or 09 are
317 used.
318
319 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
320 causing partial matching to fail when the end of the subject matched \W
321 in a UTF-8 pattern where \W was quantified with a minimum of 3.
322
323 15. There were some discrepancies between the declarations in pcre_internal.h
324 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
325 their definitions. The declarations used "const uschar *" and the
326 definitions used USPTR. Even though USPTR is normally defined as "const
327 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
328 reported that: "This difference in casting confuses some C++ compilers, for
329 example, SunCC recognizes above declarations as different functions and
330 generates broken code for hbpcre." I have changed the declarations to use
331 USPTR.
332
333 16. GNU libtool is named differently on some systems. The autogen.sh script now
334 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
335 (FreeBSD).
336
337 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
338 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
339 comment: "Figure out how to create a longlong from a string: strtoll and
340 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
341 instance, but it only takes 2 args instead of 3!"
342
343 18. A subtle bug concerned with back references has been fixed by a change of
344 specification, with a corresponding code fix. A pattern such as
345 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
346 refers, was giving matches when it shouldn't. For example, xa=xaaa would
347 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
348 same bug. Such groups have to be quantified to be useful, or contained
349 inside another quantified group. (If there's no repetition, the reference
350 can never match.) The problem arises because, having left the group and
351 moved on to the rest of the pattern, a later failure that backtracks into
352 the group uses the captured value from the final iteration of the group
353 rather than the correct earlier one. I have fixed this in PCRE by forcing
354 any group that contains a reference to itself to be an atomic group; that
355 is, there cannot be any backtracking into it once it has completed. This is
356 similar to recursive and subroutine calls.
357
358
359 Version 8.00 19-Oct-09
360 ----------------------
361
362 1. The table for translating pcre_compile() error codes into POSIX error codes
363 was out-of-date, and there was no check on the pcre_compile() error code
364 being within the table. This could lead to an OK return being given in
365 error.
366
367 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
368 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
369 in a Windows environment.
370
371 3. The pcregrep --count option prints the count for each file even when it is
372 zero, as does GNU grep. However, pcregrep was also printing all files when
373 --files-with-matches was added. Now, when both options are given, it prints
374 counts only for those files that have at least one match. (GNU grep just
375 prints the file name in this circumstance, but including the count seems
376 more useful - otherwise, why use --count?) Also ensured that the
377 combination -clh just lists non-zero counts, with no names.
378
379 4. The long form of the pcregrep -F option was incorrectly implemented as
380 --fixed_strings instead of --fixed-strings. This is an incompatible change,
381 but it seems right to fix it, and I didn't think it was worth preserving
382 the old behaviour.
383
384 5. The command line items --regex=pattern and --regexp=pattern were not
385 recognized by pcregrep, which required --regex pattern or --regexp pattern
386 (with a space rather than an '='). The man page documented the '=' forms,
387 which are compatible with GNU grep; these now work.
388
389 6. No libpcreposix.pc file was created for pkg-config; there was just
390 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
391
392 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
393 when UCP support is not needed, by modifying the Python script that
394 generates it from Unicode data files. This should not matter if the module
395 is correctly used as a library, but I received one complaint about 50K of
396 unwanted data. My guess is that the person linked everything into his
397 program rather than using a library. Anyway, it does no harm.
398
399 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
400 was a minimum greater than 1 for a wide character in a possessive
401 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
402 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
403 character. Chaos in the form of incorrect output or a compiling loop could
404 result.
405
406 9. The restrictions on what a pattern can contain when partial matching is
407 requested for pcre_exec() have been removed. All patterns can now be
408 partially matched by this function. In addition, if there are at least two
409 slots in the offset vector, the offset of the earliest inspected character
410 for the match and the offset of the end of the subject are set in them when
411 PCRE_ERROR_PARTIAL is returned.
412
413 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
414 synonymous with PCRE_PARTIAL, for backwards compatibility, and
415 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
416 and may be more useful for multi-segment matching.
417
418 11. Partial matching with pcre_exec() is now more intuitive. A partial match
419 used to be given if ever the end of the subject was reached; now it is
420 given only if matching could not proceed because another character was
421 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
422 string "Z", which now yields "no match" instead of "partial match". In the
423 case of pcre_dfa_exec(), "no match" is given if every matching path for the
424 final character ended with (*FAIL).
425
426 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
427 if the pattern had a "must contain" character that was already found in the
428 earlier partial match, unless partial matching was again requested. For
429 example, with the pattern /dog.(body)?/, the "must contain" character is
430 "g". If the first part-match was for the string "dog", restarting with
431 "sbody" failed. This bug has been fixed.
432
433 13. The string returned by pcre_dfa_exec() after a partial match has been
434 changed so that it starts at the first inspected character rather than the
435 first character of the match. This makes a difference only if the pattern
436 starts with a lookbehind assertion or \b or \B (\K is not supported by
437 pcre_dfa_exec()). It's an incompatible change, but it makes the two
438 matching functions compatible, and I think it's the right thing to do.
439
440 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
441 so that the demonstration program is easily available in environments where
442 PCRE has not been installed from source.
443
444 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
445 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
446 library.
447
448 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
449 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
450 is not the first non-POSIX option to be added. Clearly some people find
451 these options useful.
452
453 17. If a caller to the POSIX matching function regexec() passes a non-zero
454 value for nmatch with a NULL value for pmatch, the value of
455 nmatch is forced to zero.
456
457 18. RunGrepTest did not have a test for the availability of the -u option of
458 the diff command, as RunTest does. It now checks in the same way as
459 RunTest, and also checks for the -b option.
460
461 19. If an odd number of negated classes containing just a single character
462 interposed, within parentheses, between a forward reference to a named
463 subpattern and the definition of the subpattern, compilation crashed with
464 an internal error, complaining that it could not find the referenced
465 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
466 [The bug was that it was starting one character too far in when skipping
467 over the character class, thus treating the ] as data rather than
468 terminating the class. This meant it could skip too much.]
469
470 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
471 /g option in pcretest when the pattern contains \K, which makes it possible
472 to have an empty string match not at the start, even when the pattern is
473 anchored. Updated pcretest and pcredemo to use this option.
474
475 21. If the maximum number of capturing subpatterns in a recursion was greater
476 than the maximum at the outer level, the higher number was returned, but
477 with unset values at the outer level. The correct (outer level) value is
478 now given.
479
480 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
481 PCRE did not set those parentheses (unlike Perl). I have now found a way to
482 make it do so. The string so far is captured, making this feature
483 compatible with Perl.
484
485 23. The tests have been re-organized, adding tests 11 and 12, to make it
486 possible to check the Perl 5.10 features against Perl 5.10.
487
488 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
489 pattern matches a fixed length string. PCRE did not allow this; now it
490 does. Neither allows recursion.
491
492 25. I finally figured out how to implement a request to provide the minimum
493 length of subject string that was needed in order to match a given pattern.
494 (It was back references and recursion that I had previously got hung up
495 on.) This code has now been added to pcre_study(); it finds a lower bound
496 to the length of subject needed. It is not necessarily the greatest lower
497 bound, but using it to avoid searching strings that are too short does give
498 some useful speed-ups. The value is available to calling programs via
499 pcre_fullinfo().
500
501 26. While implementing 25, I discovered to my embarrassment that pcretest had
502 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
503 study optimizations had never been tested with that matching function.
504 Oops. What is worse, even when it was passed study data, there was a bug in
505 pcre_dfa_exec() that meant it never actually used it. Double oops. There
506 were also very few tests of studied patterns with pcre_dfa_exec().
507
508 27. If (?| is used to create subpatterns with duplicate numbers, they are now
509 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
510 on the other side of the coin, they are no longer allowed to have different
511 names, because these cannot be distinguished in PCRE, and this has caused
512 confusion. (This is a difference from Perl.)
513
514 28. When duplicate subpattern names are present (necessarily with different
515 numbers, as required by 27 above), and a test is made by name in a
516 conditional pattern, either for a subpattern having been matched, or for
517 recursion in such a pattern, all the associated numbered subpatterns are
518 tested, and the overall condition is true if the condition is true for any
519 one of them. This is the way Perl works, and is also more like the way
520 testing by number works.
521
522
523 Version 7.9 11-Apr-09
524 ---------------------
525
526 1. When building with support for bzlib/zlib (pcregrep) and/or readline
527 (pcretest), all targets were linked against these libraries. This included
528 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
529 libraries. This caused unwanted dependencies to be created. This problem
530 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
531 pcretest is linked with readline.
532
533 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
534 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
535 moved outside it again, because FALSE and TRUE are already defined in AIX,
536 but BOOL is not.
537
538 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
539 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
540
541 4. The pcregrep documentation said spaces were inserted as well as colons (or
542 hyphens) following file names and line numbers when outputting matching
543 lines. This is not true; no spaces are inserted. I have also clarified the
544 wording for the --colour (or --color) option.
545
546 5. In pcregrep, when --colour was used with -o, the list of matching strings
547 was not coloured; this is different to GNU grep, so I have changed it to be
548 the same.
549
550 6. When --colo(u)r was used in pcregrep, only the first matching substring in
551 each matching line was coloured. Now it goes on to look for further matches
552 of any of the test patterns, which is the same behaviour as GNU grep.
553
554 7. A pattern that could match an empty string could cause pcregrep to loop; it
555 doesn't make sense to accept an empty string match in pcregrep, so I have
556 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
557 seems to be how GNU grep behaves.
558
559 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
560 start or after a newline", because the conditional assertion was not being
561 correctly handled. The rule now is that both the assertion and what follows
562 in the first alternative must satisfy the test.
563
564 9. If auto-callout was enabled in a pattern with a conditional group whose
565 condition was an assertion, PCRE could crash during matching, both with
566 pcre_exec() and pcre_dfa_exec().
567
568 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
569 used for matching.
570
571 11. Unicode property support in character classes was not working for
572 characters (bytes) greater than 127 when not in UTF-8 mode.
573
574 12. Added the -M command line option to pcretest.
575
576 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
577
578 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
579
580 16. Added comments and documentation about mis-use of no_arg in the C++
581 wrapper.
582
583 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
584 from Martin Jerabek that uses macro names for all relevant character and
585 string constants.
586
587 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
588 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
589 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
590 these, but not everybody uses configure.
591
592 19. A conditional group that had only one branch was not being correctly
593 recognized as an item that could match an empty string. This meant that an
594 enclosing group might also not be so recognized, causing infinite looping
595 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
596 with the subject "ab", where knowledge that the repeated group can match
597 nothing is needed in order to break the loop.
598
599 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
600 exec(), but without supplying a callout function, matching went wrong.
601
602 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
603 leak if the size of the offset vector was greater than 30. When the vector
604 is smaller, the saved offsets during recursion go onto a local stack
605 vector, but for larger vectors malloc() is used. It was failing to free
606 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
607 error, in fact).
608
609 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
610 heapframe that is used only when UTF-8 support is enabled. This caused no
611 problem, but was untidy.
612
613 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
614 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
615 included within another project.
616
617 24. Steven Van Ingelgem's patches to add more options to the CMake support,
618 slightly modified by me:
619
620 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
621 not building pcregrep.
622
623 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
624 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
625
626 25. Forward references, both numeric and by name, in patterns that made use of
627 duplicate group numbers, could behave incorrectly or give incorrect errors,
628 because when scanning forward to find the reference group, PCRE was not
629 taking into account the duplicate group numbers. A pattern such as
630 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
631
632 26. Changed a few more instances of "const unsigned char *" to USPTR, making
633 the feature of a custom pointer more persuasive (as requested by a user).
634
635 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
636 pcretest.c, inside #ifndefs, because it seems they are sometimes already
637 pre-defined.
638
639 28. Added support for (*UTF8) at the start of a pattern.
640
641 29. Arrange for flags added by the "release type" setting in CMake to be shown
642 in the configuration summary.
643
644
645 Version 7.8 05-Sep-08
646 ---------------------
647
648 1. Replaced UCP searching code with optimized version as implemented for Ad
649 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
650 stage table and inline lookup instead of a function, giving speed ups of 2
651 to 5 times on some simple patterns that I tested. Permission was given to
652 distribute the MultiStage2.py script that generates the tables (it's not in
653 the tarball, but is in the Subversion repository).
654
655 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
656 scripts.
657
658 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
659 a group with a zero qualifier. The result of the study could be incorrect,
660 or the function might crash, depending on the pattern.
661
662 4. Caseless matching was not working for non-ASCII characters in back
663 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
664 It now works when Unicode Property Support is available.
665
666 5. In pcretest, an escape such as \x{de} in the data was always generating
667 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
668 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
669 truncation.
670
671 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
672
673 7. Added two (int) casts to pcregrep when printing the difference of two
674 pointers, in case they are 64-bit values.
675
676 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
677 test 2 if it fails.
678
679 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
680 and a #define of that name to empty if it is not externally set. This is to
681 allow users of MSVC to set it if necessary.
682
683 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
684 the convenience functions in the pcre_get.c source file.
685
686 11. An option change at the start of a pattern that had top-level alternatives
687 could cause overwriting and/or a crash. This command provoked a crash in
688 some environments:
689
690 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
691
692 This potential security problem was recorded as CVE-2008-2371.
693
694 12. For a pattern where the match had to start at the beginning or immediately
695 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
696 pcre_dfa_exec() could read past the end of the passed subject if there was
697 no match. To help with detecting such bugs (e.g. with valgrind), I modified
698 pcretest so that it places the subject at the end of its malloc-ed buffer.
699
700 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
701 exec() might read past the end of the data buffer in UTF-8 mode.
702
703 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
704 the data contained the byte 0x85 as part of a UTF-8 character within its
705 first line. This applied both to normal and DFA matching.
706
707 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
708 /^[^d]*?$/8 failed to match "abc".
709
710 16. Added a missing copyright notice to pcrecpp_internal.h.
711
712 17. Make it more clear in the documentation that values returned from
713 pcre_exec() in ovector are byte offsets, not character counts.
714
715 18. Tidied a few places to stop certain compilers from issuing warnings.
716
717 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
718 supplied by Stefan Weber. I made a further small update for 7.8 because
719 there is a change of source arrangements: the pcre_searchfuncs.c module is
720 replaced by pcre_ucd.c.
721
722
723 Version 7.7 07-May-08
724 ---------------------
725
726 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
727 a string to a long long, pretend we don't even have a long long." This is
728 done by checking for the strtoq, strtoll, and _strtoi64 functions.
729
730 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
731 pre-7.6 versions, which defined a global no_arg variable instead of putting
732 it in the RE class. (See also #8 below.)
733
734 3. Remove a line of dead code, identified by coverity and reported by Nuno
735 Lopes.
736
737 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
738
739 (1) The include/exclude patterns were being applied to the whole pathnames
740 of files, instead of just to the final components.
741
742 (2) If there was more than one level of directory, the subdirectories were
743 skipped unless they satisfied the include/exclude conditions. This is
744 inconsistent with GNU grep (and could even be seen as contrary to the
745 pcregrep specification - which I improved to make it absolutely clear).
746 The action now is always to scan all levels of directory, and just
747 apply the include/exclude patterns to regular files.
748
749 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
750 --exclude_dir in the tests to avoid scanning .svn directories.
751
752 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
753 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
754 doesn't support NULs in patterns.
755
756 7. Added some missing "const"s to declarations of static tables in
757 pcre_compile.c and pcre_dfa_exec.c.
758
759 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
760 caused by fix #2 above. (Subsequently also a second patch to fix the
761 first patch. And a third patch - this was a messy problem.)
762
763 9. Applied Craig's patch to remove the use of push_back().
764
765 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
766 matching function regexec().
767
768 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
769 which, however, unlike Perl's \g{...}, are subroutine calls, not back
770 references. PCRE supports relative numbers with this syntax (I don't think
771 Oniguruma does).
772
773 12. Previously, a group with a zero repeat such as (...){0} was completely
774 omitted from the compiled regex. However, this means that if the group
775 was called as a subroutine from elsewhere in the pattern, things went wrong
776 (an internal error was given). Such groups are now left in the compiled
777 pattern, with a new opcode that causes them to be skipped at execution
778 time.
779
780 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
781 to the way PCRE behaves:
782
783 (a) A lone ] character is dis-allowed (Perl treats it as data).
784
785 (b) A back reference to an unmatched subpattern matches an empty string
786 (Perl fails the current match path).
787
788 (c) A data ] in a character class must be notated as \] because if the
789 first data character in a class is ], it defines an empty class. (In
790 Perl it is not possible to have an empty class.) The empty class []
791 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
792 The negative empty class [^] matches any one character, independently
793 of the DOTALL setting.
794
795 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
796 non-existent subpattern following a character class starting with ']' and
797 containing () gave an internal compiling error instead of "reference to
798 non-existent subpattern". Fortunately, when the pattern did exist, the
799 compiled code was correct. (When scanning forwards to check for the
800 existencd of the subpattern, it was treating the data ']' as terminating
801 the class, so got the count wrong. When actually compiling, the reference
802 was subsequently set up correctly.)
803
804 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
805 it was being rejected as not supported by pcre_dfa_exec(), even though
806 other assertions are supported. I have made pcre_dfa_exec() support
807 (*FAIL).
808
809 16. The implementation of 13c above involved the invention of a new opcode,
810 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
811 cannot be changed at match time, I realized I could make a small
812 improvement to matching performance by compiling OP_ALLANY instead of
813 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
814 on the OP_ANY path.
815
816 17. Compiling pcretest on Windows with readline support failed without the
817 following two fixes: (1) Make the unistd.h include conditional on
818 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
819
820 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
821 ncurses library to be included for pcretest when ReadLine support is
822 requested, but also to allow for it to be overridden. This patch came from
823 Daniel Bergström.
824
825 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
826 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
827 any errors with the current Unicode tables. Thanks to Peter Kankowski for
828 spotting this.
829
830
831 Version 7.6 28-Jan-08
832 ---------------------
833
834 1. A character class containing a very large number of characters with
835 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
836 overflow.
837
838 2. Patch to cut out the "long long" test in pcrecpp_unittest when
839 HAVE_LONG_LONG is not defined.
840
841 3. Applied Christian Ehrlicher's patch to update the CMake build files to
842 bring them up to date and include new features. This patch includes:
843
844 - Fixed PH's badly added libz and libbz2 support.
845 - Fixed a problem with static linking.
846 - Added pcredemo. [But later removed - see 7 below.]
847 - Fixed dftables problem and added an option.
848 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
849 HAVE_LONG_LONG.
850 - Added readline support for pcretest.
851 - Added an listing of the option settings after cmake has run.
852
853 4. A user submitted a patch to Makefile that makes it easy to create
854 "pcre.dll" under mingw when using Configure/Make. I added stuff to
855 Makefile.am that cause it to include this special target, without
856 affecting anything else. Note that the same mingw target plus all
857 the other distribution libraries and programs are now supported
858 when configuring with CMake (see 6 below) instead of with
859 Configure/Make.
860
861 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
862 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
863 exported in the Windows port". It has not yet been confirmed that the patch
864 solves the problem, but it does no harm.
865
866 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
867 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
868 with CMake, and also correct the comment about stack recursion.
869
870 7. Remove the automatic building of pcredemo from the ./configure system and
871 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
872 of a program that users should build themselves after PCRE is installed, so
873 building it automatically is not really right. What is more, it gave
874 trouble in some build environments.
875
876 8. Further tidies to CMakeLists.txt from Sheri and Christian.
877
878
879 Version 7.5 10-Jan-08
880 ---------------------
881
882 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
883 values in parens when parsing an RE using the C++ wrapper."
884
885 2. Negative specials like \S did not work in character classes in UTF-8 mode.
886 Characters greater than 255 were excluded from the class instead of being
887 included.
888
889 3. The same bug as (2) above applied to negated POSIX classes such as
890 [:^space:].
891
892 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
893 defined or documented. It seems to have been a typo for PCRE_STATIC, so
894 I have changed it.
895
896 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
897 first named subpattern) and a construct such as (?&a) would reference the
898 first named subpattern whose name started with "a" (in other words, the
899 length check was missing). Both these problems are fixed. "Subpattern name
900 expected" is now given for (?&) (a zero-length name), and this patch also
901 makes it give the same error for \k'' (previously it complained that that
902 was a reference to a non-existent subpattern).
903
904 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
905 this is right because (?- can be followed by option settings as well as by
906 digits. I have, however, made the messages clearer.
907
908 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
909 than the number used in the conditional) now cause a compile-time error.
910 This is actually not compatible with Perl, which accepts such patterns, but
911 treats the conditional as always being FALSE (as PCRE used to), but it
912 seems to me that giving a diagnostic is better.
913
914 8. Change "alphameric" to the more common word "alphanumeric" in comments
915 and messages.
916
917 9. Fix two occurrences of "backslash" in comments that should have been
918 "backspace".
919
920 10. Remove two redundant lines of code that can never be obeyed (their function
921 was moved elsewhere).
922
923 11. The program that makes PCRE's Unicode character property table had a bug
924 which caused it to generate incorrect table entries for sequences of
925 characters that have the same character type, but are in different scripts.
926 It amalgamated them into a single range, with the script of the first of
927 them. In other words, some characters were in the wrong script. There were
928 thirteen such cases, affecting characters in the following ranges:
929
930 U+002b0 - U+002c1
931 U+0060c - U+0060d
932 U+0061e - U+00612
933 U+0064b - U+0065e
934 U+0074d - U+0076d
935 U+01800 - U+01805
936 U+01d00 - U+01d77
937 U+01d9b - U+01dbf
938 U+0200b - U+0200f
939 U+030fc - U+030fe
940 U+03260 - U+0327f
941 U+0fb46 - U+0fbb1
942 U+10450 - U+1049d
943
944 12. The -o option (show only the matching part of a line) for pcregrep was not
945 compatible with GNU grep in that, if there was more than one match in a
946 line, it showed only the first of them. It now behaves in the same way as
947 GNU grep.
948
949 13. If the -o and -v options were combined for pcregrep, it printed a blank
950 line for every non-matching line. GNU grep prints nothing, and pcregrep now
951 does the same. The return code can be used to tell if there were any
952 non-matching lines.
953
954 14. Added --file-offsets and --line-offsets to pcregrep.
955
956 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
957 infinitely looping recursion. The bug was that positive lookaheads were not
958 being skipped when checking for a possible empty match (negative lookaheads
959 and both kinds of lookbehind were skipped).
960
961 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
962 inclusion of <windows.h> to before rather than after the definition of
963 INVALID_FILE_ATTRIBUTES (patch from David Byron).
964
965 17. Specifying a possessive quantifier with a specific limit for a Unicode
966 character property caused pcre_compile() to compile bad code, which led at
967 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
968 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
969 caused the error; without that there was no problem.
970
971 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
972
973 19. Added --enable-pcretest-libreadline.
974
975 20. In pcrecpp.cc, the variable 'count' was incremented twice in
976 RE::GlobalReplace(). As a result, the number of replacements returned was
977 double what it should be. I removed one of the increments, but Craig sent a
978 later patch that removed the other one (the right fix) and added unit tests
979 that check the return values (which was not done before).
980
981 21. Several CMake things:
982
983 (1) Arranged that, when cmake is used on Unix, the libraries end up with
984 the names libpcre and libpcreposix, not just pcre and pcreposix.
985
986 (2) The above change means that pcretest and pcregrep are now correctly
987 linked with the newly-built libraries, not previously installed ones.
988
989 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
990
991 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
992 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
993 UTF-8 newline character). The key issue is that the pattern starts .*;
994 this means that the match must be either at the beginning, or after a
995 newline. The bug was in the code for advancing after a failed match and
996 checking that the new position followed a newline. It was not taking
997 account of UTF-8 characters correctly.
998
999 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1000 character classes. PCRE was not treating the sequence [:...:] as a
1001 character class unless the ... were all letters. Perl, however, seems to
1002 allow any characters between [: and :], though of course it rejects as
1003 unknown any "names" that contain non-letters, because all the known class
1004 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1005 for example, whereas PCRE did not - it did not recognize a POSIX character
1006 class. This seemed a bit dangerous, so the code has been changed to be
1007 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1008 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1009 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1010 Perl does, and where it didn't before.
1011
1012 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1013 Windows environments %n is disabled by default.
1014
1015
1016 Version 7.4 21-Sep-07
1017 ---------------------
1018
1019 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1020 means that a class such as [\s] counted as "explicit reference to CR or
1021 LF". That isn't really right - the whole point of the change was to try to
1022 help when there was an actual mention of one of the two characters. So now
1023 the change happens only if \r or \n (or a literal CR or LF) character is
1024 encountered.
1025
1026 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1027 of both had grown to the point where there were only 3 bits left.
1028 Fortunately, there was spare space in the data structure, and so I have
1029 moved the internal flags into a new 16-bit field to free up more option
1030 bits.
1031
1032 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1033 but did not set the internal JCHANGED flag - either of these is enough to
1034 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1035 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1036 start sets both bits.
1037
1038 4. Added options (at build time, compile time, exec time) to change \R from
1039 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1040
1041 5. doc/pcresyntax.html was missing from the distribution.
1042
1043 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1044 compatibility, even though it is no longer used.
1045
1046 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1047 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1048 windows.h file is present (where different names are used). [This was
1049 reversed later after testing - see 16 below.]
1050
1051 8. Changed all #include <config.h> to #include "config.h". There were also
1052 some further <pcre.h> cases that I changed to "pcre.h".
1053
1054 9. When pcregrep was used with the --colour option, it missed the line ending
1055 sequence off the lines that it output.
1056
1057 10. It was pointed out to me that arrays of string pointers cause lots of
1058 relocations when a shared library is dynamically loaded. A technique of
1059 using a single long string with a table of offsets can drastically reduce
1060 these. I have refactored PCRE in four places to do this. The result is
1061 dramatic:
1062
1063 Originally: 290
1064 After changing UCP table: 187
1065 After changing error message table: 43
1066 After changing table of "verbs" 36
1067 After changing table of Posix names 22
1068
1069 Thanks to the folks working on Gregex for glib for this insight.
1070
1071 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1072 unicode-properties was also set.
1073
1074 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1075
1076 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1077 checked only for CRLF.
1078
1079 14. Added casts to pcretest.c to avoid compiler warnings.
1080
1081 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1082
1083 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1084 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1085 entirely. This removes changes made in 7 above.
1086
1087 17. The CMake files have been updated, and there is now more information about
1088 building with CMake in the NON-UNIX-USE document.
1089
1090
1091 Version 7.3 28-Aug-07
1092 ---------------------
1093
1094 1. In the rejigging of the build system that eventually resulted in 7.1, the
1095 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1096 brackets there is not right, since it causes compilers to look for an
1097 installed pcre.h, not the version that is in the source that is being
1098 compiled (which of course may be different). I have changed it back to:
1099
1100 #include "pcre.h"
1101
1102 I have a vague recollection that the change was concerned with compiling in
1103 different directories, but in the new build system, that is taken care of
1104 by the VPATH setting the Makefile.
1105
1106 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1107 when the subject happened to end in the byte 0x85 (e.g. if the last
1108 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1109 characters but of course it shouldn't be taken as a newline when it is part
1110 of another character. The bug was that, for an unlimited repeat of . in
1111 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1112 characters when looking for a newline.
1113
1114 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1115
1116 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1117 in debug output.
1118
1119 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1120 long printing in the pcrecpp unittest when running under MinGW.
1121
1122 6. ESC_K was left out of the EBCDIC table.
1123
1124 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1125 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1126 limit also applies to "virtual nesting" when a pattern is recursive, and in
1127 this case 1000 isn't so big. I have been able to remove this limit at the
1128 expense of backing off one optimization in certain circumstances. Normally,
1129 when pcre_exec() would call its internal match() function recursively and
1130 immediately return the result unconditionally, it uses a "tail recursion"
1131 feature to save stack. However, when a subpattern that can match an empty
1132 string has an unlimited repetition quantifier, it no longer makes this
1133 optimization. That gives it a stack frame in which to save the data for
1134 checking that an empty string has been matched. Previously this was taken
1135 from the 1000-entry workspace that had been reserved. So now there is no
1136 explicit limit, but more stack is used.
1137
1138 8. Applied Daniel's patches to solve problems with the import/export magic
1139 syntax that is required for Windows, and which was going wrong for the
1140 pcreposix and pcrecpp parts of the library. These were overlooked when this
1141 problem was solved for the main library.
1142
1143 9. There were some crude static tests to avoid integer overflow when computing
1144 the size of patterns that contain repeated groups with explicit upper
1145 limits. As the maximum quantifier is 65535, the maximum group length was
1146 set at 30,000 so that the product of these two numbers did not overflow a
1147 32-bit integer. However, it turns out that people want to use groups that
1148 are longer than 30,000 bytes (though not repeat them that many times).
1149 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1150 made it possible to implement the integer overflow checks in a much more
1151 dynamic way, which I have now done. The artificial limitation on group
1152 length has been removed - we now have only the limit on the total length of
1153 the compiled pattern, which depends on the LINK_SIZE setting.
1154
1155 10. Fixed a bug in the documentation for get/copy named substring when
1156 duplicate names are permitted. If none of the named substrings are set, the
1157 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1158 empty string.
1159
1160 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1161 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1162 because the ] is interpreted as the first data character and the
1163 terminating ] is not found. PCRE has been made compatible with Perl in this
1164 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1165 cause memory overwriting.
1166
1167 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1168 string has been matched (to stop an infinite loop). It was not recognizing
1169 a conditional subpattern that could match an empty string if that
1170 subpattern was within another subpattern. For example, it looped when
1171 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1172 condition was not nested. This bug has been fixed.
1173
1174 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1175 past the start of the subject in the presence of bytes with the top bit
1176 set, for example "\x8aBCD".
1177
1178 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1179 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1180
1181 14. Optimized (?!) to (*FAIL).
1182
1183 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1184 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1185 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1186 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1187 does: it's just the validity check that is more restrictive.
1188
1189 16. Inserted checks for integer overflows during escape sequence (backslash)
1190 processing, and also fixed erroneous offset values for syntax errors during
1191 backslash processing.
1192
1193 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1194 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1195
1196 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1197 caused an overrun.
1198
1199 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1200 something other than just ASCII characters) inside a group that had an
1201 unlimited repeat caused a loop at compile time (while checking to see
1202 whether the group could match an empty string).
1203
1204 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1205 [\P{Any}] did so. (Error in the code for printing property names.)
1206
1207 21. An orphan \E inside a character class could cause a crash.
1208
1209 22. A repeated capturing bracket such as (A)? could cause a wild memory
1210 reference during compilation.
1211
1212 23. There are several functions in pcre_compile() that scan along a compiled
1213 expression for various reasons (e.g. to see if it's fixed length for look
1214 behind). There were bugs in these functions when a repeated \p or \P was
1215 present in the pattern. These operators have additional parameters compared
1216 with \d, etc, and these were not being taken into account when moving along
1217 the compiled data. Specifically:
1218
1219 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1220 length.
1221
1222 (b) An item such as \pL+ within a repeated group could cause crashes or
1223 loops.
1224
1225 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1226 "reference to non-existent subpattern" error.
1227
1228 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1229
1230 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1231 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1232
1233 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1234
1235 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1236 character were causing crashes (broken optimization).
1237
1238 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1239 \p or \P) caused a compile-time loop.
1240
1241 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1242 break. For example, the unstudied pattern [\r\n]A does not match the string
1243 "\r\nA" because change 7.0/46 below moves the current point on by two
1244 characters after failing to match at the start. However, the pattern \nA
1245 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1246 the same is true. There doesn't seem any very clean way out of this, but
1247 what I have chosen to do makes the common cases work: PCRE now takes note
1248 of whether there can be an explicit match for \r or \n anywhere in the
1249 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1250 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1251 pattern has explicit CR or LF references.
1252
1253 29. Added (*CR) etc for changing newline setting at start of pattern.
1254
1255
1256 Version 7.2 19-Jun-07
1257 ---------------------
1258
1259 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1260 which is apparently normally available under Windows.
1261
1262 2. Re-jig the pcregrep tests with different newline settings in an attempt
1263 to make them independent of the local environment's newline setting.
1264
1265 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1266
1267 4. Some of the "internals" tests were previously cut out when the link size
1268 was not 2, because the output contained actual offsets. The recent new
1269 "Z" feature of pcretest means that these can be cut out, making the tests
1270 usable with all link sizes.
1271
1272 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1273 stack recursion. This gives a massive performance boost under BSD, but just
1274 a small improvement under Linux. However, it saves one field in the frame
1275 in all cases.
1276
1277 6. Added more features from the forthcoming Perl 5.10:
1278
1279 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1280 recursion call. It refers to the nth most recently opened parentheses.
1281
1282 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1283 to be opened parentheses.
1284
1285 (c) Conditions that refer to capturing parentheses can be specified
1286 relatively, for example, (?(-2)... or (?(+3)...
1287
1288 (d) \K resets the start of the current match so that everything before
1289 is not part of it.
1290
1291 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1292
1293 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1294 reference syntax.
1295
1296 (g) (?| introduces a group in which the numbering of parentheses in each
1297 alternative starts with the same number.
1298
1299 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1300
1301 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1302 PCRE_INFO_JCHANGED.
1303
1304 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1305 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1306 for detecting groups that can match an empty string.
1307
1308 9. A pattern with a very large number of alternatives (more than several
1309 hundred) was running out of internal workspace during the pre-compile
1310 phase, where pcre_compile() figures out how much memory will be needed. A
1311 bit of new cunning has reduced the workspace needed for groups with
1312 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1313 workspace instead of running out of the 4096 that are available.
1314
1315 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1316
1317 11. Applied patch from Google to remove an optimization that didn't quite work.
1318 The report of the bug said:
1319
1320 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1321 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1322 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1323
1324 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1325 it matched the wrong number of bytes.
1326
1327
1328 Version 7.1 24-Apr-07
1329 ---------------------
1330
1331 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1332 that is more "standard", making use of automake and other Autotools. There
1333 is some re-arrangement of the files and adjustment of comments consequent
1334 on this.
1335
1336 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1337 for recursive directory scanning broke on some systems because the files
1338 are not scanned in any specific order and on different systems the order
1339 was different. A call to "sort" has been inserted into RunGrepTest for the
1340 approprate test as a short-term fix. In the longer term there may be an
1341 alternative.
1342
1343 3. I had an email from Eric Raymond about problems translating some of PCRE's
1344 man pages to HTML (despite the fact that I distribute HTML pages, some
1345 people do their own conversions for various reasons). The problems
1346 concerned the use of low-level troff macros .br and .in. I have therefore
1347 removed all such uses from the man pages (some were redundant, some could
1348 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1349 HTML has been updated to handle .nf/.fi and to complain if it encounters
1350 .br or .in.
1351
1352 4. Updated comments in configure.ac that get placed in config.h.in and also
1353 arranged for config.h to be included in the distribution, with the name
1354 config.h.generic, for the benefit of those who have to compile without
1355 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1356
1357 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1358 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1359 makevp.bat for the current PCRE, using the additional files
1360 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1361
1362 6. A Windows user reported a minor discrepancy with test 2, which turned out
1363 to be caused by a trailing space on an input line that had got lost in his
1364 copy. The trailing space was an accident, so I've just removed it.
1365
1366 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1367 that is needed.
1368
1369 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1370 as "const" (a) because they are and (b) because it helps the PHP
1371 maintainers who have recently made a script to detect big data structures
1372 in the php code that should be moved to the .rodata section. I remembered
1373 to update Builducptable as well, so it won't revert if ucptable.h is ever
1374 re-created.
1375
1376 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1377 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1378 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1379 support is not required. This saves 1.5-2K of code, which is important in
1380 some applications.
1381
1382 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1383 so as not to refer to the tables, even though these functions will never be
1384 called when UTF-8 support is disabled. Otherwise there are problems with a
1385 shared library.
1386
1387 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1388
1389 (a) It was defining its arguments as char * instead of void *.
1390
1391 (b) It was assuming that all moves were upwards in memory; this was true
1392 a long time ago when I wrote it, but is no longer the case.
1393
1394 The emulated memove() is provided for those environments that have neither
1395 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1396 is clearly not the case, as these two bugs were recently reported.
1397
1398 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1399 and Detrail to create the HTML documentation, the .txt form of the man
1400 pages, and it removes trailing spaces from listed files. It also creates
1401 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1402 case, it wraps all the #defines with #ifndefs. This script should be run
1403 before "make dist".
1404
1405 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1406 with Unicode property support.
1407
1408 (a) For a maximizing quantifier, if the two different cases of the
1409 character were of different lengths in their UTF-8 codings (there are
1410 some cases like this - I found 11), and the matching function had to
1411 back up over a mixture of the two cases, it incorrectly assumed they
1412 were both the same length.
1413
1414 (b) When PCRE was configured to use the heap rather than the stack for
1415 recursion during matching, it was not correctly preserving the data for
1416 the other case of a UTF-8 character when checking ahead for a match
1417 while processing a minimizing repeat. If the check also involved
1418 matching a wide character, but failed, corruption could cause an
1419 erroneous result when trying to check for a repeat of the original
1420 character.
1421
1422 13. Some tidying changes to the testing mechanism:
1423
1424 (a) The RunTest script now detects the internal link size and whether there
1425 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1426 values substituted by "configure". (The RunGrepTest script already did
1427 this for UTF-8.) The configure.ac script no longer substitutes the
1428 relevant variables.
1429
1430 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1431 with length and offset values. This means that the output is different
1432 for different internal link sizes. Test 2 is skipped for link sizes
1433 other than 2 because of this, bypassing the problem. Unfortunately,
1434 there was also a test in test 3 (the locale tests) that used /B and
1435 failed for link sizes other than 2. Rather than cut the whole test out,
1436 I have added a new /Z option to pcretest that replaces the length and
1437 offset values with spaces. This is now used to make test 3 independent
1438 of link size. (Test 2 will be tidied up later.)
1439
1440 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1441 segmentation fault instead of returning the appropriate error message.
1442
1443 15. In multiline mode when the newline sequence was set to "any", the pattern
1444 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1445 This doesn't seem right; it now treats the CRLF combination as the line
1446 ending, and so does not match in that case. It's only a pattern such as ^$
1447 that would hit this one: something like ^ABC$ would have failed after \r
1448 and then tried again after \r\n.
1449
1450 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1451 in an attempt to make files that differ only in their line terminators
1452 compare equal. This works on Linux.
1453
1454 17. Under certain error circumstances pcregrep might try to free random memory
1455 as it exited. This is now fixed, thanks to valgrind.
1456
1457 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1458 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1459 was because its rules for how to advance for /g after matching an empty
1460 string at the end of a line did not allow for this case. They now check for
1461 it specially.
1462
1463 20. pcretest is supposed to handle patterns and data of any length, by
1464 extending its buffers when necessary. It was getting this wrong when the
1465 buffer for a data line had to be extended.
1466
1467 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1468 CRLF as a newline sequence.
1469
1470 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1471 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1472 I have nevertheless tidied it up.
1473
1474 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1475
1476 24. Added a man page for pcre-config.
1477
1478
1479 Version 7.0 19-Dec-06
1480 ---------------------
1481
1482 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1483 moving to gcc 4.1.1.
1484
1485 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1486 sys/time.h, which is documented as needed for this function. It doesn't
1487 seem to matter on Linux, but it showed up on some releases of OS X.
1488
1489 3. It seems that there are systems where bytes whose values are greater than
1490 127 match isprint() in the "C" locale. The "C" locale should be the
1491 default when a C program starts up. In most systems, only ASCII printing
1492 characters match isprint(). This difference caused the output from pcretest
1493 to vary, making some of the tests fail. I have changed pcretest so that:
1494
1495 (a) When it is outputting text in the compiled version of a pattern, bytes
1496 other than 32-126 are always shown as hex escapes.
1497
1498 (b) When it is outputting text that is a matched part of a subject string,
1499 it does the same, unless a different locale has been set for the match
1500 (using the /L modifier). In this case, it uses isprint() to decide.
1501
1502 4. Fixed a major bug that caused incorrect computation of the amount of memory
1503 required for a compiled pattern when options that changed within the
1504 pattern affected the logic of the preliminary scan that determines the
1505 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1506 that the computed length was too small. The symptoms of this bug were
1507 either the PCRE error "internal error: code overflow" from pcre_compile(),
1508 or a glibc crash with a message such as "pcretest: free(): invalid next
1509 size (fast)". Examples of patterns that provoked this bug (shown in
1510 pcretest format) are:
1511
1512 /(?-x: )/x
1513 /(?x)(?-x: \s*#\s*)/
1514 /((?i)[\x{c0}])/8
1515 /(?i:[\x{c0}])/8
1516
1517 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1518 is now done differently.
1519
1520 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1521 wrapper classes; (b) implement a new function in the C++ scanner that is
1522 more efficient than the old way of doing things because it avoids levels of
1523 recursion in the regex matching; (c) add a paragraph to the documentation
1524 for the FullMatch() function.
1525
1526 6. The escape sequence \n was being treated as whatever was defined as
1527 "newline". Not only was this contrary to the documentation, which states
1528 that \n is character 10 (hex 0A), but it also went horribly wrong when
1529 "newline" was defined as CRLF. This has been fixed.
1530
1531 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1532 was being set to -1 for the "end of line" case (supposedly a value that no
1533 character can have). Though this value is never used (the check for end of
1534 line is "zero bytes in current character"), it caused compiler complaints.
1535 I've changed it to 0xffffffff.
1536
1537 8. In pcre_version.c, the version string was being built by a sequence of
1538 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1539 string (as it is for production releases) called a macro with an empty
1540 argument. The C standard says the result of this is undefined. The gcc
1541 compiler treats it as an empty string (which was what was wanted) but it is
1542 reported that Visual C gives an error. The source has been hacked around to
1543 avoid this problem.
1544
1545 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1546 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1547 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1548 of them did).
1549
1550 10. Originally, pcretest opened its input and output without "b"; then I was
1551 told that "b" was needed in some environments, so it was added for release
1552 5.0 to both the input and output. (It makes no difference on Unix-like
1553 systems.) Later I was told that it is wrong for the input on Windows. I've
1554 now abstracted the modes into two macros, to make it easier to fiddle with
1555 them, and removed "b" from the input mode under Windows.
1556
1557 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1558
1559 12. Added -help and --help to pcretest as an official way of being reminded
1560 of the options.
1561
1562 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1563 and pcrecpp.cc because they annoy compilers at high warning levels.
1564
1565 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1566
1567 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1568 scripts are not C programs :-) and which was not noticed because it works
1569 on Linux.
1570
1571 16. pcretest is supposed to handle any length of pattern and data line (as one
1572 line or as a continued sequence of lines) by extending its input buffer if
1573 necessary. This feature was broken for very long pattern lines, leading to
1574 a string of junk being passed to pcre_compile() if the pattern was longer
1575 than about 50K.
1576
1577 17. I have done a major re-factoring of the way pcre_compile() computes the
1578 amount of memory needed for a compiled pattern. Previously, there was code
1579 that made a preliminary scan of the pattern in order to do this. That was
1580 OK when PCRE was new, but as the facilities have expanded, it has become
1581 harder and harder to keep it in step with the real compile phase, and there
1582 have been a number of bugs (see for example, 4 above). I have now found a
1583 cunning way of running the real compile function in a "fake" mode that
1584 enables it to compute how much memory it would need, while actually only
1585 ever using a few hundred bytes of working memory and without too many
1586 tests of the mode. This should make future maintenance and development
1587 easier. A side effect of this work is that the limit of 200 on the nesting
1588 depth of parentheses has been removed (though this was never a serious
1589 limitation, I suspect). However, there is a downside: pcre_compile() now
1590 runs more slowly than before (30% or more, depending on the pattern). I
1591 hope this isn't a big issue. There is no effect on runtime performance.
1592
1593 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1594 newline (only possible for the last line of a file) and it was a
1595 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1596
1597 19. Added additional timing features to pcretest. (1) The -tm option now times
1598 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1599 separate command line item, by a number that specifies the number of
1600 repeats to use when timing. The default is 50000; this gives better
1601 precision, but takes uncomfortably long for very large patterns.
1602
1603 20. Extended pcre_study() to be more clever in cases where a branch of a
1604 subpattern has no definite first character. For example, (a*|b*)[cd] would
1605 previously give no result from pcre_study(). Now it recognizes that the
1606 first character must be a, b, c, or d.
1607
1608 21. There was an incorrect error "recursive call could loop indefinitely" if
1609 a subpattern (or the entire pattern) that was being tested for matching an
1610 empty string contained only one non-empty item after a nested subpattern.
1611 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1612 incorrectly, because the \d was being skipped in the check.
1613
1614 22. The pcretest program now has a new pattern option /B and a command line
1615 option -b, which is equivalent to adding /B to every pattern. This causes
1616 it to show the compiled bytecode, without the additional information that
1617 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1618 is the same as /B/I).
1619
1620 23. A new optimization is now able automatically to treat some sequences such
1621 as a*b as a*+b. More specifically, if something simple (such as a character
1622 or a simple class like \d) has an unlimited quantifier, and is followed by
1623 something that cannot possibly match the quantified thing, the quantifier
1624 is automatically "possessified".
1625
1626 24. A recursive reference to a subpattern whose number was greater than 39
1627 went wrong under certain circumstances in UTF-8 mode. This bug could also
1628 have affected the operation of pcre_study().
1629
1630 25. Realized that a little bit of performance could be had by replacing
1631 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1632
1633 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1634
1635 27. Possessive quantifiers such as a++ were previously implemented by turning
1636 them into atomic groups such as ($>a+). Now they have their own opcodes,
1637 which improves performance. This includes the automatically created ones
1638 from 23 above.
1639
1640 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1641 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1642 the first matched character to be a colon. This applied both to named and
1643 numbered groups.
1644
1645 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1646
1647 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1648 building PCRE on Windows easier, so I have included it in the distribution.
1649
1650 31. There is now a check in pcretest against a ridiculously large number being
1651 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1652 loop, the loop is abandoned.
1653
1654 32. Forward references to subpatterns in conditions such as (?(2)...) where
1655 subpattern 2 is defined later cause pcre_compile() to search forwards in
1656 the pattern for the relevant set of parentheses. This search went wrong
1657 when there were unescaped parentheses in a character class, parentheses
1658 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1659
1660 33. "Subroutine" calls and backreferences were previously restricted to
1661 referencing subpatterns earlier in the regex. This restriction has now
1662 been removed.
1663
1664 34. Added a number of extra features that are going to be in Perl 5.10. On the
1665 whole, these are just syntactic alternatives for features that PCRE had
1666 previously implemented using the Python syntax or my own invention. The
1667 other formats are all retained for compatibility.
1668
1669 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1670 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1671 also .NET compatible.
1672
1673 (b) A recursion or subroutine call to a named group can now be defined as
1674 (?&name) as well as (?P>name).
1675
1676 (c) A backreference to a named group can now be defined as \k<name> or
1677 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1678 5.10, are also .NET compatible.
1679
1680 (d) A conditional reference to a named group can now use the syntax
1681 (?(<name>) or (?('name') as well as (?(name).
1682
1683 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1684 groups (named and numbered) that are never evaluated inline, but can be
1685 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1686 is always false. There may be only one alternative in such a group.
1687
1688 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1689 as the simple (?(R). The condition is true only if the most recent
1690 recursion is that of the given number or name. It does not search out
1691 through the entire recursion stack.
1692
1693 (g) The escape \gN or \g{N} has been added, where N is a positive or
1694 negative number, specifying an absolute or relative reference.
1695
1696 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1697 some "unreachable code" warnings.
1698
1699 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1700 things, this adds five new scripts.
1701
1702 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1703 There were also incompatibilities regarding the handling of \Q..\E inside
1704 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1705 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1706
1707 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1708 matches an empty string, and forcibly breaks the loop. There were bugs in
1709 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1710 against aaaa the result was just "a" rather than "aaaa", for example. Two
1711 separate and independent bugs (that affected different cases) have been
1712 fixed.
1713
1714 39. Refactored the code to abolish the use of different opcodes for small
1715 capturing bracket numbers. This is a tidy that I avoided doing when I
1716 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1717 The new approach is not only tidier, it makes it possible to reduce the
1718 memory needed to fix the previous bug (38).
1719
1720 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1721 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1722 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1723 mode.
1724
1725 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1726 report.
1727
1728 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1729 copy construction and assignment in the C++ wrapper.
1730
1731 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1732 couple of bugs that could have given wrong results in the "--newline=crlf"
1733 case.
1734
1735 44. Added a number of casts and did some reorganization of signed/unsigned int
1736 variables following suggestions from Dair Grant. Also renamed the variable
1737 "this" as "item" because it is a C++ keyword.
1738
1739 45. Arranged for dftables to add
1740
1741 #include "pcre_internal.h"
1742
1743 to pcre_chartables.c because without it, gcc 4.x may remove the array
1744 definition from the final binary if PCRE is built into a static library and
1745 dead code stripping is activated.
1746
1747 46. For an unanchored pattern, if a match attempt fails at the start of a
1748 newline sequence, and the newline setting is CRLF or ANY, and the next two
1749 characters are CRLF, advance by two characters instead of one.
1750
1751
1752 Version 6.7 04-Jul-06
1753 ---------------------
1754
1755 1. In order to handle tests when input lines are enormously long, pcretest has
1756 been re-factored so that it automatically extends its buffers when
1757 necessary. The code is crude, but this _is_ just a test program. The
1758 default size has been increased from 32K to 50K.
1759
1760 2. The code in pcre_study() was using the value of the re argument before
1761 testing it for NULL. (Of course, in any sensible call of the function, it
1762 won't be NULL.)
1763
1764 3. The memmove() emulation function in pcre_internal.h, which is used on
1765 systems that lack both memmove() and bcopy() - that is, hardly ever -
1766 was missing a "static" storage class specifier.
1767
1768 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1769 containing an extended class (one that cannot be represented by a bitmap
1770 because it contains high-valued characters or Unicode property items, e.g.
1771 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1772 pattern, but PCRE should not loop if you do not (it no longer does).
1773 [Detail: two cases were found: (a) a repeated subpattern containing an
1774 extended class; (b) a recursive reference to a subpattern that followed a
1775 previous extended class. It wasn't skipping over the extended class
1776 correctly when UTF-8 mode was not set.]
1777
1778 5. A negated single-character class was not being recognized as fixed-length
1779 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1780 compile error "lookbehind assertion is not fixed length".
1781
1782 6. The RunPerlTest auxiliary script was showing an unexpected difference
1783 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1784 write a Perl script that can interpret lines of an input file either as
1785 byte characters or as UTF-8, which is what "perltest" was being required to
1786 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1787 can't do is switch easily at run time between having the "use utf8;" pragma
1788 or not. In the end, I fudged it by using the RunPerlTest script to insert
1789 "use utf8;" explicitly for the UTF-8 tests.
1790
1791 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1792 the end of the subject string, contrary to the documentation and to what
1793 Perl does. This was true of both matching functions. Now it matches only at
1794 the start of the subject and immediately after *internal* newlines.
1795
1796 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1797 a pointer to an int instead of a pointer to an unsigned long int. This
1798 caused problems on 64-bit systems.
1799
1800 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1801 instance of the 'standard' template library not being so standard".
1802
1803 10. There was no check on the number of named subpatterns nor the maximum
1804 length of a subpattern name. The product of these values is used to compute
1805 the size of the memory block for a compiled pattern. By supplying a very
1806 long subpattern name and a large number of named subpatterns, the size
1807 computation could be caused to overflow. This is now prevented by limiting
1808 the length of names to 32 characters, and the number of named subpatterns
1809 to 10,000.
1810
1811 11. Subpatterns that are repeated with specific counts have to be replicated in
1812 the compiled pattern. The size of memory for this was computed from the
1813 length of the subpattern and the repeat count. The latter is limited to
1814 65535, but there was no limit on the former, meaning that integer overflow
1815 could in principle occur. The compiled length of a repeated subpattern is
1816 now limited to 30,000 bytes in order to prevent this.
1817
1818 12. Added the optional facility to have named substrings with the same name.
1819
1820 13. Added the ability to use a named substring as a condition, using the
1821 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1822 are numbers (not recommended). Forward references are permitted.
1823
1824 14. Added forward references in named backreferences (if you see what I mean).
1825
1826 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1827 pattern could run off the end of the subject. For example, the pattern
1828 "(?s)(.{1,5})"8 did this with the subject "ab".
1829
1830 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1831 PCRE_CASELESS was set when matching characters that were quantified with ?
1832 or *.
1833
1834 17. A character class other than a single negated character that had a minimum
1835 but no maximum quantifier - for example [ab]{6,} - was not handled
1836 correctly by pce_dfa_exec(). It would match only one character.
1837
1838 18. A valid (though odd) pattern that looked like a POSIX character
1839 class but used an invalid character after [ (for example [[,abc,]]) caused
1840 pcre_compile() to give the error "Failed: internal error: code overflow" or
1841 in some cases to crash with a glibc free() error. This could even happen if
1842 the pattern terminated after [[ but there just happened to be a sequence of
1843 letters, a binary zero, and a closing ] in the memory that followed.
1844
1845 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1846 over the years. Originally (before any Unicode support), just the bottom 8
1847 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1848 output from "man perlunicode" includes this:
1849
1850 The regular expression compiler produces polymorphic opcodes. That
1851 is, the pattern adapts to the data and automatically switches to
1852 the Unicode character scheme when presented with Unicode data--or
1853 instead uses a traditional byte scheme when presented with byte
1854 data.
1855
1856 Sadly, a wide octal escape does not cause a switch, and in a string with
1857 no other multibyte characters, these octal escapes are treated as before.
1858 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1859 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1860 Unicode string.
1861
1862 I have not perpetrated such confusion in PCRE. Up till now, it took just
1863 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1864 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1865 translate to the appropriate multibyte character.
1866
1867 29. Applied some refactoring to reduce the number of warnings from Microsoft
1868 and Borland compilers. This has included removing the fudge introduced
1869 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1870 a warning about an unused variable.
1871
1872 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1873 characters since release 4.0, because Perl (from release 5.004) does not.
1874 [Or at least, is documented not to: some releases seem to be in conflict
1875 with the documentation.] However, when a pattern was studied with
1876 pcre_study() and all its branches started with \s, PCRE still included VT
1877 as a possible starting character. Of course, this did no harm; it just
1878 caused an unnecessary match attempt.
1879
1880 22. Removed a now-redundant internal flag bit that recorded the fact that case
1881 dependency changed within the pattern. This was once needed for "required
1882 byte" processing, but is no longer used. This recovers a now-scarce options
1883 bit. Also moved the least significant internal flag bit to the most-
1884 significant bit of the word, which was not previously used (hangover from
1885 the days when it was an int rather than a uint) to free up another bit for
1886 the future.
1887
1888 23. Added support for CRLF line endings as well as CR and LF. As well as the
1889 default being selectable at build time, it can now be changed at runtime
1890 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1891 specify that it is scanning data with non-default line endings.
1892
1893 24. Changed the definition of CXXLINK to make it agree with the definition of
1894 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1895
1896 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1897 recursions. This makes a big different to stack usage for some patterns.
1898
1899 26. If a subpattern containing a named recursion or subroutine reference such
1900 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1901 the space required for the compiled pattern went wrong and gave too small a
1902 value. Depending on the environment, this could lead to "Failed: internal
1903 error: code overflow at offset 49" or "glibc detected double free or
1904 corruption" errors.
1905
1906 27. Applied patches from Google (a) to support the new newline modes and (b) to
1907 advance over multibyte UTF-8 characters in GlobalReplace.
1908
1909 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1910 difference for some implementation of PCRE in some Windows version.
1911
1912 29. Added some extra testing facilities to pcretest:
1913
1914 \q<number> in a data line sets the "match limit" value
1915 \Q<number> in a data line sets the "match recursion limt" value
1916 -S <number> sets the stack size, where <number> is in megabytes
1917
1918 The -S option isn't available for Windows.
1919
1920
1921 Version 6.6 06-Feb-06
1922 ---------------------
1923
1924 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1925 in pcreposix.h. I have copied the definition from pcre.h.
1926
1927 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1928 because pcre.h is no longer a built file.
1929
1930 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1931 not normally included in the compiled code.
1932
1933
1934 Version 6.5 01-Feb-06
1935 ---------------------
1936
1937 1. When using the partial match feature with pcre_dfa_exec(), it was not
1938 anchoring the second and subsequent partial matches at the new starting
1939 point. This could lead to incorrect results. For example, with the pattern
1940 /1234/, partially matching against "123" and then "a4" gave a match.
1941
1942 2. Changes to pcregrep:
1943
1944 (a) All non-match returns from pcre_exec() were being treated as failures
1945 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1946 error message is output. Some extra information is given for the
1947 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1948 probably the only errors that are likely to be caused by users (by
1949 specifying a regex that has nested indefinite repeats, for instance).
1950 If there are more than 20 of these errors, pcregrep is abandoned.
1951
1952 (b) A binary zero was treated as data while matching, but terminated the
1953 output line if it was written out. This has been fixed: binary zeroes
1954 are now no different to any other data bytes.
1955
1956 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1957 used to set a locale for matching. The --locale=xxxx long option has
1958 been added (no short equivalent) to specify a locale explicitly on the
1959 pcregrep command, overriding the environment variables.
1960
1961 (d) When -B was used with -n, some line numbers in the output were one less
1962 than they should have been.
1963
1964 (e) Added the -o (--only-matching) option.
1965
1966 (f) If -A or -C was used with -c (count only), some lines of context were
1967 accidentally printed for the final match.
1968
1969 (g) Added the -H (--with-filename) option.
1970
1971 (h) The combination of options -rh failed to suppress file names for files
1972 that were found from directory arguments.
1973
1974 (i) Added the -D (--devices) and -d (--directories) options.
1975
1976 (j) Added the -F (--fixed-strings) option.
1977
1978 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1979
1980 (l) Added the --colo(u)r option.
1981
1982 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1983 is not present by default.
1984
1985 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1986 items such as (?R) or (?1), when the recursion could match a number of
1987 alternatives. If it matched one of the alternatives, but subsequently,
1988 outside the recursion, there was a failure, the code tried to back up into
1989 the recursion. However, because of the way PCRE is implemented, this is not
1990 possible, and the result was an incorrect result from the match.
1991
1992 In order to prevent this happening, the specification of recursion has
1993 been changed so that all such subpatterns are automatically treated as
1994 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1995
1996 4. I had overlooked the fact that, in some locales, there are characters for
1997 which isalpha() is true but neither isupper() nor islower() are true. In
1998 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1999 and ordfeminine) are like this. This affected the treatment of \w and \W
2000 when they appeared in character classes, but not when they appeared outside
2001 a character class. The bit map for "word" characters is now created
2002 separately from the results of isalnum() instead of just taking it from the
2003 upper, lower, and digit maps. (Plus the underscore character, of course.)
2004
2005 5. The above bug also affected the handling of POSIX character classes such as
2006 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2007 permanent tables. Instead, the bit maps for such a class were previously
2008 created as the appropriate unions of the upper, lower, and digit bitmaps.
2009 Now they are created by subtraction from the [[:word:]] class, which has
2010 its own bitmap.
2011
2012 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2013 It is created by subtracting the vertical space characters (\x09, \x0a,
2014 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2015 subtraction was done in the overall bitmap for a character class, meaning
2016 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2017 be recognized. This bug has been fixed.
2018
2019 7. Patches from the folks at Google:
2020
2021 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2022 real life, but is still worth protecting against".
2023
2024 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2025 regular expressions".
2026
2027 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2028 have it.
2029
2030 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2031 "configure" and the latter not, in order to fix a problem somebody had
2032 with compiling the Arg class on HP-UX.
2033
2034 (e) Improve the error-handling of the C++ wrapper a little bit.
2035
2036 (f) New tests for checking recursion limiting.
2037
2038 8. The pcre_memmove() function, which is used only if the environment does not
2039 have a standard memmove() function (and is therefore rarely compiled),
2040 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2041 returning a result (though PCRE never actually uses the result).
2042
2043 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2044 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2045 returned instead of calling malloc() with an overflowing number that would
2046 most likely cause subsequent chaos.
2047
2048 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2049
2050 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2051 with this option is matched, the nmatch and pmatch options of regexec() are
2052 ignored.
2053
2054 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2055 provided in case anyone wants to the the POSIX interface with UTF-8
2056 strings.
2057
2058 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2059 C++ linking (needed for some HP-UX environments).
2060
2061 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2062 (unused parameter) and in the pcre_printint() function (omitted "default"
2063 switch label when the default is to do nothing).
2064
2065 15. Added some code to make it possible, when PCRE is compiled as a C++
2066 library, to replace subject pointers for pcre_exec() with a smart pointer
2067 class, thus making it possible to process discontinuous strings.
2068
2069 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2070 much the same function. They were added by different people who were trying
2071 to make PCRE easy to compile on non-Unix systems. It has been suggested
2072 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2073 for compiling on Windows systems. I have therefore replaced it with
2074 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2075 defaults to "extern" for C or "extern C" for C++, which works fine on
2076 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2077 SCOPE with something explicit in config.h. In addition:
2078
2079 (a) pcreposix.h still had just "extern" instead of either of these macros;
2080 I have replaced it with PCRE_DATA_SCOPE.
2081
2082 (b) Functions such as _pcre_xclass(), which are internal to the library,
2083 but external in the C sense, all had PCRE_EXPORT in their definitions.
2084 This is apparently wrong for the Windows case, so I have removed it.
2085 (It makes no difference on Unix-like systems.)
2086
2087 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2088 of recursive calls to match(). This is different to MATCH_LIMIT because
2089 that limits the total number of calls to match(), not all of which increase
2090 the depth of recursion. Limiting the recursion depth limits the amount of
2091 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2092 when PCRE is compiled, and changed at run time. A patch from Google adds
2093 this functionality to the C++ interface.
2094
2095 18. Changes to the handling of Unicode character properties:
2096
2097 (a) Updated the table to Unicode 4.1.0.
2098
2099 (b) Recognize characters that are not in the table as "Cn" (undefined).
2100
2101 (c) I revised the way the table is implemented to a much improved format
2102 which includes recognition of ranges. It now supports the ranges that
2103 are defined in UnicodeData.txt, and it also amalgamates other
2104 characters into ranges. This has reduced the number of entries in the
2105 table from around 16,000 to around 3,000, thus reducing its size
2106 considerably. I realized I did not need to use a tree structure after
2107 all - a binary chop search is just as efficient. Having reduced the
2108 number of entries, I extended their size from 6 bytes to 8 bytes to
2109 allow for more data.
2110
2111 (d) Added support for Unicode script names via properties such as \p{Han}.
2112
2113 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2114 matching that character.
2115
2116 20. When matching a repeated Unicode property with a minimum greater than zero,
2117 (for example \pL{2,}), PCRE could look past the end of the subject if it
2118 reached it while seeking the minimum number of characters. This could
2119 happen only if some of the characters were more than one byte long, because
2120 there is a check for at least the minimum number of bytes.
2121
2122 21. Refactored the implementation of \p and \P so as to be more general, to
2123 allow for more different types of property in future. This has changed the
2124 compiled form incompatibly. Anybody with saved compiled patterns that use
2125 \p or \P will have to recompile them.
2126
2127 22. Added "Any" and "L&" to the supported property types.
2128
2129 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2130 but give a compile time error if the value is greater than 0xff.
2131
2132 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2133 accidentally not being installed or uninstalled.
2134
2135 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2136 made were to insert the current release number. This seemed silly, because
2137 it made things harder for people building PCRE on systems that don't run
2138 "configure". I have turned pcre.h into a distributed file, no longer built
2139 by "configure", with the version identification directly included. There is
2140 no longer a pcre.h.in file.
2141
2142 However, this change necessitated a change to the pcre-config script as
2143 well. It is built from pcre-config.in, and one of the substitutions was the
2144 release number. I have updated configure.ac so that ./configure now finds
2145 the release number by grepping pcre.h.
2146
2147 26. Added the ability to run the tests under valgrind.
2148
2149
2150 Version 6.4 05-Sep-05
2151 ---------------------
2152
2153 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2154 "--" to be printed when multiple files were scanned, even when none of the
2155 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2156 consider it to be a bug, and have restored the previous behaviour.
2157
2158 2. A couple of code tidies to get rid of compiler warnings.
2159
2160 3. The pcretest program used to cheat by referring to symbols in the library
2161 whose names begin with _pcre_. These are internal symbols that are not
2162 really supposed to be visible externally, and in some environments it is
2163 possible to suppress them. The cheating is now confined to including
2164 certain files from the library's source, which is a bit cleaner.
2165
2166 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2167 file's purpose clearer.
2168
2169 5. Reorganized pcre_ucp_findchar().
2170
2171
2172 Version 6.3 15-Aug-05
2173 ---------------------
2174
2175 1. The file libpcre.pc.in did not have general read permission in the tarball.
2176
2177 2. There were some problems when building without C++ support:
2178
2179 (a) If C++ support was not built, "make install" and "make test" still
2180 tried to test it.
2181
2182 (b) There were problems when the value of CXX was explicitly set. Some
2183 changes have been made to try to fix these, and ...
2184
2185 (c) --disable-cpp can now be used to explicitly disable C++ support.
2186
2187 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2188 backslash in a target when C++ was disabled. This confuses some
2189 versions of "make", apparently. Using an intermediate variable solves
2190 this. (Same for CPP_LOBJ.)
2191
2192 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2193 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2194 necessary on certain architectures.
2195
2196 4. Added a setting of -export-symbols-regex to the link command to remove
2197 those symbols that are exported in the C sense, but actually are local
2198 within the library, and not documented. Their names all begin with
2199 "_pcre_". This is not a perfect job, because (a) we have to except some
2200 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2201 available (and never for static libraries). I have made a note to try to
2202 find a way round (a) in the future.
2203
2204
2205 Version 6.2 01-Aug-05
2206 ---------------------
2207
2208 1. There was no test for integer overflow of quantifier values. A construction
2209 such as {1111111111111111} would give undefined results. What is worse, if
2210 a minimum quantifier for a parenthesized subpattern overflowed and became
2211 negative, the calculation of the memory size went wrong. This could have
2212 led to memory overwriting.
2213
2214 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2215
2216 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2217 operating environments where this matters.
2218
2219 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2220 PCRE options from within the C++ wrapper.
2221
2222 5. Named capturing subpatterns were not being correctly counted when a pattern
2223 was compiled. This caused two problems: (a) If there were more than 100
2224 such subpatterns, the calculation of the memory needed for the whole
2225 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2226 back references of the form \12, where the number was greater than 9, were
2227 not recognized as back references, even though there were sufficient
2228 previous subpatterns.
2229
2230 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2231 versions of gcc, e.g. 2.95.4.
2232
2233
2234 Version 6.1 21-Jun-05
2235 ---------------------
2236
2237 1. There was one reference to the variable "posix" in pcretest.c that was not
2238 surrounded by "#if !defined NOPOSIX".
2239
2240 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2241 the cross-check on the old pcre_info() function, for the benefit of the
2242 cut-down version of PCRE that is currently imported into Exim.
2243
2244 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2245 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2246 patterns that start (?i) but I don't think that matters. The use of (?i) is
2247 just an example; this all applies to the other options as well.
2248
2249 4. Since libtool seems to echo the compile commands it is issuing, the output
2250 from "make" can be reduced a bit by putting "@" in front of each libtool
2251 compile command.
2252
2253 5. Patch from the folks at Google for configure.in to be a bit more thorough
2254 in checking for a suitable C++ installation before trying to compile the
2255 C++ stuff. This should fix a reported problem when a compiler was present,
2256 but no suitable headers.
2257
2258 6. The man pages all had just "PCRE" as their title. I have changed them to
2259 be the relevant file name. I have also arranged that these names are
2260 retained in the file doc/pcre.txt, which is a concatenation in text format
2261 of all the man pages except the little individual ones for each function.
2262
2263 7. The NON-UNIX-USE file had not been updated for the different set of source
2264 files that come with release 6. I also added a few comments about the C++
2265 wrapper.
2266
2267
2268 Version 6.0 07-Jun-05
2269 ---------------------
2270
2271 1. Some minor internal re-organization to help with my DFA experiments.
2272
2273 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2274 didn't matter for the library itself when fully configured, but did matter
2275 when compiling without UCP support, or within Exim, where the ucp files are
2276 not imported.
2277
2278 3. Refactoring of the library code to split up the various functions into
2279 different source modules. The addition of the new DFA matching code (see
2280 below) to a single monolithic source would have made it really too
2281 unwieldy, quite apart from causing all the code to be include in a
2282 statically linked application, when only some functions are used. This is
2283 relevant even without the DFA addition now that patterns can be compiled in
2284 one application and matched in another.
2285
2286 The downside of splitting up is that there have to be some external
2287 functions and data tables that are used internally in different modules of
2288 the library but which are not part of the API. These have all had their
2289 names changed to start with "_pcre_" so that they are unlikely to clash
2290 with other external names.
2291
2292 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2293 a different (DFA) algorithm. Although it is slower than the original
2294 function, it does have some advantages for certain types of matching
2295 problem.
2296
2297 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2298 including restarting after a partial match.
2299
2300 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2301 defined when compiling for Windows was sent to me. I have put it into the
2302 code, though I have no means of testing or verifying it.
2303
2304 7. Added the pcre_refcount() auxiliary function.
2305
2306 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2307 match before or at the first newline in the subject string. In pcretest,
2308 the /f option on a pattern can be used to set this.
2309
2310 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2311 would behave wrongly. This has been present in PCRE since release 4.0.
2312
2313 10. A number of changes to the pcregrep command:
2314
2315 (a) Refactored how -x works; insert ^(...)$ instead of setting
2316 PCRE_ANCHORED and checking the length, in preparation for adding
2317 something similar for -w.
2318
2319 (b) Added the -w (match as a word) option.
2320
2321 (c) Refactored the way lines are read and buffered so as to have more
2322 than one at a time available.
2323
2324 (d) Implemented a pcregrep test script.
2325
2326 (e) Added the -M (multiline match) option. This allows patterns to match
2327 over several lines of the subject. The buffering ensures that at least
2328 8K, or the rest of the document (whichever is the shorter) is available
2329 for matching (and similarly the previous 8K for lookbehind assertions).
2330
2331 (f) Changed the --help output so that it now says
2332
2333 -w, --word-regex(p)
2334
2335 instead of two lines, one with "regex" and the other with "regexp"
2336 because that confused at least one person since the short forms are the
2337 same. (This required a bit of code, as the output is generated
2338 automatically from a table. It wasn't just a text change.)
2339
2340 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2341 option but starts with a hyphen. Could be a pattern or a path name
2342 starting with a hyphen, for instance.
2343
2344 (h) "-" can be given as a file name to represent stdin.
2345
2346 (i) When file names are being printed, "(standard input)" is used for
2347 the standard input, for compatibility with GNU grep. Previously
2348 "<stdin>" was used.
2349
2350 (j) The option --label=xxx can be used to supply a name to be used for
2351 stdin when file names are being printed. There is no short form.
2352
2353 (k) Re-factored the options decoding logic because we are going to add
2354 two more options that take data. Such options can now be given in four
2355 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2356
2357 (l) Added the -A, -B, and -C options for requesting that lines of context
2358 around matches be printed.
2359
2360 (m) Added the -L option to print the names of files that do not contain
2361 any matching lines, that is, the complement of -l.
2362
2363 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2364 continue to scan other files.
2365
2366 (o) The -s option was incorrectly implemented. For compatibility with other
2367 greps, it now suppresses the error message for a non-existent or non-
2368 accessible file (but not the return code). There is a new option called
2369 -q that suppresses the output of matching lines, which was what -s was
2370 previously doing.
2371
2372 (p) Added --include and --exclude options to specify files for inclusion
2373 and exclusion when recursing.
2374
2375 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2376 Hopefully, it now does.
2377
2378 12. Missing cast in pcre_study().
2379
2380 13. Added an "uninstall" target to the makefile.
2381
2382 14. Replaced "extern" in the function prototypes in Makefile.in with
2383 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2384 world, but is set differently for Windows.
2385
2386 15. Added a second compiling function called pcre_compile2(). The only
2387 difference is that it has an extra argument, which is a pointer to an
2388 integer error code. When there is a compile-time failure, this is set
2389 non-zero, in addition to the error test pointer being set to point to an
2390 error message. The new argument may be NULL if no error number is required
2391 (but then you may as well call pcre_compile(), which is now just a
2392 wrapper). This facility is provided because some applications need a
2393 numeric error indication, but it has also enabled me to tidy up the way
2394 compile-time errors are handled in the POSIX wrapper.
2395
2396 16. Added VPATH=.libs to the makefile; this should help when building with one
2397 prefix path and installing with another. (Or so I'm told by someone who
2398 knows more about this stuff than I do.)
2399
2400 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2401 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2402 match everything, including newlines. This is not POSIX-compatible, but
2403 somebody wanted the feature. From pcretest it can be activated by using
2404 both the P and the s flags.
2405
2406 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2407
2408 19. libpcre.pc was being incorrectly installed as executable.
2409
2410 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2411 it now also looks for '\r' so that it will work unmodified on Windows.
2412
2413 21. Added Google's contributed C++ wrapper to the distribution.
2414
2415 22. Added some untidy missing memory free() calls in pcretest, to keep
2416 Electric Fence happy when testing.
2417
2418
2419
2420 Version 5.0 13-Sep-04
2421 ---------------------
2422
2423 1. Internal change: literal characters are no longer packed up into items
2424 containing multiple characters in a single byte-string. Each character
2425 is now matched using a separate opcode. However, there may be more than one
2426 byte in the character in UTF-8 mode.
2427
2428 2. The pcre_callout_block structure has two new fields: pattern_position and
2429 next_item_length. These contain the offset in the pattern to the next match
2430 item, and its length, respectively.
2431
2432 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2433 insertion of callouts before each pattern item. Added the /C option to
2434 pcretest to make use of this.
2435
2436 4. On the advice of a Windows user, the lines
2437
2438 #if defined(_WIN32) || defined(WIN32)
2439 _setmode( _fileno( stdout ), 0x8000 );
2440 #endif /* defined(_WIN32) || defined(WIN32) */
2441
2442 have been added to the source of pcretest. This apparently does useful
2443 magic in relation to line terminators.
2444
2445 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2446 for the benefit of those environments where the "b" makes a difference.
2447
2448 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2449 to know about it. I have put a hack into configure.in that adds in code
2450 to set GCC=yes if CC=icc. This seems to end up at a point in the
2451 generated configure script that is early enough to affect the setting of
2452 compiler options, which is what is needed, but I have no means of testing
2453 whether it really works. (The user who reported this had patched the
2454 generated configure script, which of course I cannot do.)
2455
2456 LATER: After change 22 below (new libtool files), the configure script
2457 seems to know about icc (and also ecc). Therefore, I have commented out
2458 this hack in configure.in.
2459
2460 7. Added support for pkg-config (2 patches were sent in).
2461
2462 8. Negated POSIX character classes that used a combination of internal tables
2463 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2464 [[:^ascii]]. Typically, they would match almost any characters. The other
2465 POSIX classes were not broken in this way.
2466
2467 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2468 to find the match, as PCRE was deluded into thinking that the match had to
2469 start at the start point or following a newline. The same bug applied to
2470 patterns with negative forward assertions or any backward assertions
2471 preceding ".*" at the start, unless the pattern required a fixed first
2472 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2473
2474 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2475 starting at the last subject character, bytes beyond the end of the subject
2476 string were read.
2477
2478 11. Renamed the variable "class" as "classbits" to make life easier for C++
2479 users. (Previously there was a macro definition, but it apparently wasn't
2480 enough.)
2481
2482 12. Added the new field "tables" to the extra data so that tables can be passed
2483 in at exec time, or the internal tables can be re-selected. This allows
2484 a compiled regex to be saved and re-used at a later time by a different
2485 program that might have everything at different addresses.
2486
2487 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2488 -R library as well as a -L library.
2489
2490 14. The debugging options of pcretest (-d on the command line or D on a
2491 pattern) showed incorrect output for anything following an extended class
2492 that contained multibyte characters and which was followed by a quantifier.
2493
2494 15. Added optional support for general category Unicode character properties
2495 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2496 support. It adds about 90K to the size of the library. The meanings of the
2497 inbuilt class escapes such as \d and \s have NOT been changed.
2498
2499 16. Updated pcredemo.c to include calls to free() to release the memory for the
2500 compiled pattern.
2501
2502 17. The generated file chartables.c was being created in the source directory
2503 instead of in the building directory. This caused the build to fail if the
2504 source directory was different from the building directory, and was
2505 read-only.
2506
2507 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2508 file. No doubt somebody will tell me if they don't make sense... Also added
2509 Dan Mooney's comments about building on OpenVMS.
2510
2511 19. Added support for partial matching via the PCRE_PARTIAL option for
2512 pcre_exec() and the \P data escape in pcretest.
2513
2514 20. Extended pcretest with 3 new pattern features:
2515
2516 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2517 write the compiled pattern to the file whose name is "rest-of-line".
2518 This is a straight binary dump of the data, with the saved pointer to
2519 the character tables forced to be NULL. The study data, if any, is
2520 written too. After writing, pcretest reads a new pattern.
2521
2522 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2523 compiled pattern from the given file. There must not be any
2524 occurrences of "<" in the file name (pretty unlikely); if there are,
2525 pcretest will instead treat the initial "<" as a pattern delimiter.
2526 After reading in the pattern, pcretest goes on to read data lines as
2527 usual.
2528
2529 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2530 and 16-bit fields in a compiled pattern, to simulate a pattern that
2531 was compiled on a host of opposite endianness.
2532
2533 21. The pcre-exec() function can now cope with patterns that were compiled on
2534 hosts of opposite endianness, with this restriction:
2535
2536 As for any compiled expression that is saved and used later, the tables
2537 pointer field cannot be preserved; the extra_data field in the arguments
2538 to pcre_exec() should be used to pass in a tables address if a value
2539 other than the default internal tables were used at compile time.
2540
2541 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2542 now diagnosed as an error. Previously, most of the time, a negative number
2543 would have been treated as zero, but if in addition "ovector" was passed as
2544 NULL, a crash could occur.
2545
2546 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2547 new versions from the libtool 1.5 distribution (the last one is a copy of
2548 a file called libtool.m4). This seems to have fixed the need to patch
2549 "configure" to support Darwin 1.3 (which I used to do). However, I still
2550 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2551 workstation).
2552
2553 24. Changed the PCRE licence to be the more standard "BSD" licence.
2554
2555
2556 Version 4.5 01-Dec-03
2557 ---------------------
2558
2559 1. There has been some re-arrangement of the code for the match() function so
2560 that it can be compiled in a version that does not call itself recursively.
2561 Instead, it keeps those local variables that need separate instances for
2562 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2563 needs to "recurse". Keeping track of where control must go is done by means
2564 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2565 hide most of the details from the main code, and operates only if
2566 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2567 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2568 operating.
2569
2570 To make it easier for callers to provide specially tailored get/free
2571 functions for this usage, two new functions, pcre_stack_malloc, and
2572 pcre_stack_free, are used. They are always called in strict stacking order,
2573 and the size of block requested is always the same.
2574
2575 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2576 PCRE has been compiled to use the stack or the heap for recursion. The
2577 -C option of pcretest uses this to show which version is compiled.
2578
2579 A new data escape \S, is added to pcretest; it causes the amounts of store
2580 obtained and freed by both kinds of malloc/free at match time to be added
2581 to the output.
2582
2583 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2584 what's available on my current Linux desktop machine.
2585
2586 3. When matching a UTF-8 string, the test for a valid string at the start has
2587 been extended. If start_offset is not zero, PCRE now checks that it points
2588 to a byte that is the start of a UTF-8 character. If not, it returns
2589 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2590 this is necessary because there may be backward assertions in the pattern.
2591 When matching the same subject several times, it may save resources to use
2592 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2593
2594 4. The code for checking the validity of UTF-8 strings has been tightened so
2595 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2596 containing "overlong sequences".
2597
2598 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2599 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2600 should have been "&", but it just so happened that all the cases this let
2601 through by mistake were picked up later in the function.
2602
2603 6. I had used a variable called "isblank" - this is a C99 function, causing
2604 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2605
2606 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2607 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2608 so the version is shown; (c) stop "make test" echoing "./RunTest".
2609
2610 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2611
2612 9. The prototype for memmove() for systems that don't have it was using
2613 size_t, but the inclusion of the header that defines size_t was later. I've
2614 moved the #includes for the C headers earlier to avoid this.
2615
2616 10. Added some adjustments to the code to make it easier to compiler on certain
2617 special systems:
2618
2619 (a) Some "const" qualifiers were missing.
2620 (b) Added the macro EXPORT before all exported functions; by default this
2621 is defined to be empty.
2622 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2623 that it reads its output file name as an argument instead of writing
2624 to the standard output and assuming this can be redirected.
2625
2626 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2627 class containing characters with values greater than 255, PCRE compilation
2628 went into a loop.
2629
2630 12. A recursive reference to a subpattern that was within another subpattern
2631 that had a minimum quantifier of zero caused PCRE to crash. For example,
2632 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2633 recursion. If the recursively-called subpattern itself had a zero repeat,
2634 that was OK.
2635
2636 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2637 buffer into which it was copied (for escape processing) was still set at
2638 1024, so long lines caused crashes.
2639
2640 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2641 "internal error: code overflow...". This applied to any character class
2642 that was followed by a possessive quantifier.
2643
2644 15. Modified the Makefile to add libpcre.la as a prerequisite for
2645 libpcreposix.la because I was told this is needed for a parallel build to
2646 work.
2647
2648 16. If a pattern that contained .* following optional items at the start was
2649 studied, the wrong optimizing data was generated, leading to matching
2650 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2651 matching string must start with a or b or c. The correct conclusion for
2652 this pattern is that a match can start with any character.
2653
2654
2655 Version 4.4 13-Aug-03
2656 ---------------------
2657
2658 1. In UTF-8 mode, a character class containing characters with values between
2659 127 and 255 was not handled correctly if the compiled pattern was studied.
2660 In fixing this, I have also improved the studying algorithm for such
2661 classes (slightly).
2662
2663 2. Three internal functions had redundant arguments passed to them. Removal
2664 might give a very teeny performance improvement.
2665
2666 3. Documentation bug: the value of the capture_top field in a callout is *one
2667 more than* the number of the hightest numbered captured substring.
2668
2669 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2670 in incorrectly linking with a previously installed version. They now link
2671 explicitly with libpcre.la.
2672
2673 5. configure.in no longer needs to recognize Cygwin specially.
2674
2675 6. A problem in pcre.in for Windows platforms is fixed.
2676
2677 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2678 pcretest, it used to include the size of the study block as part of its
2679 output. Unfortunately, the structure contains a field that has a different
2680 size on different hardware architectures. This meant that the tests that
2681 showed this size failed. As the block is currently always of a fixed size,
2682 this information isn't actually particularly useful in pcretest output, so
2683 I have just removed it.
2684
2685 8. Three pre-processor statements accidentally did not start in column 1.
2686 Sadly, there are *still* compilers around that complain, even though
2687 standard C has not required this for well over a decade. Sigh.
2688
2689 9. In pcretest, the code for checking callouts passed small integers in the
2690 callout_data field, which is a void * field. However, some picky compilers
2691 complained about the casts involved for this on 64-bit systems. Now
2692 pcretest passes the address of the small integer instead, which should get
2693 rid of the warnings.
2694
2695 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2696 both compile and run time, and gives an error if an invalid UTF-8 sequence
2697 is found. There is a option for disabling this check in cases where the
2698 string is known to be correct and/or the maximum performance is wanted.
2699
2700 11. In response to a bug report, I changed one line in Makefile.in from
2701
2702 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2703 to
2704 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2705
2706 to look similar to other lines, but I have no way of telling whether this
2707 is the right thing to do, as I do not use Windows. No doubt I'll get told
2708 if it's wrong...
2709
2710
2711 Version 4.3 21-May-03
2712 ---------------------
2713
2714 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2715 Makefile.
2716
2717 2. Some refactoring to improve the quality of the code:
2718
2719 (i) The utf8_table... variables are now declared "const".
2720
2721 (ii) The code for \cx, which used the "case flipping" table to upper case
2722 lower case letters, now just substracts 32. This is ASCII-specific,
2723 but the whole concept of \cx is ASCII-specific, so it seems
2724 reasonable.
2725
2726 (iii) PCRE was using its character types table to recognize decimal and
2727 hexadecimal digits in the pattern. This is silly, because it handles
2728 only 0-9, a-f, and A-F, but the character types table is locale-
2729 specific, which means strange things might happen. A private
2730 table is now used for this - though it costs 256 bytes, a table is
2731 much faster than multiple explicit tests. Of course, the standard
2732 character types table is still used for matching digits in subject
2733 strings against \d.
2734
2735 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2736 ending in _t are). So I've renamed it as ESC_tee.
2737
2738 3. The first argument for regexec() in the POSIX wrapper should have been
2739 defined as "const".
2740
2741 4. Changed pcretest to use malloc() for its buffers so that they can be
2742 Electric Fenced for debugging.
2743
2744 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2745 to read one or more bytes before the start of the subject string. Often this
2746 had no effect on PCRE's behaviour, but in some circumstances it could
2747 provoke a segmentation fault.
2748
2749 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2750 to try to read one or more bytes before the start of the subject string.
2751
2752 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2753 UTF-8 support could misbehave in various ways if the subject string
2754 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2755 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2756 back over UTF-8 characters.)
2757
2758
2759 Version 4.2 14-Apr-03
2760 ---------------------
2761
2762 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2763
2764 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2765 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2766 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2767 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2768 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2769 and BUILD_EXEEXT
2770 Note: automatic setting of the BUILD variables is not yet working
2771 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2772 compile-time but not at link-time
2773 [LINK]: use for linking executables only
2774 make different versions for Windows and non-Windows
2775 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2776 libraries
2777 [LINK_FOR_BUILD]: new variable
2778 [OBJEXT]: use throughout
2779 [EXEEXT]: use throughout
2780 <winshared>: new target
2781 <wininstall>: new target
2782 <dftables.o>: use native compiler
2783 <dftables>: use native linker
2784 <install>: handle Windows platform correctly
2785 <clean>: ditto
2786 <check>: ditto
2787 copy DLL to top builddir before testing
2788
2789 As part of these changes, -no-undefined was removed again. This was reported
2790 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2791 in any case.
2792
2793 3. Some tidies to get rid of compiler warnings:
2794
2795 . In the match_data structure, match_limit was an unsigned long int, whereas
2796 match_call_count was an int. I've made them both unsigned long ints.
2797
2798 . In pcretest the fact that a const uschar * doesn't automatically cast to
2799 a void * provoked a warning.
2800
2801 . Turning on some more compiler warnings threw up some "shadow" variables
2802 and a few more missing casts.
2803
2804 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2805 option, a class that contained a single character with a value between 128
2806 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2807
2808 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2809 option, a class that contained several characters, but with at least one
2810 whose value was between 128 and 255 caused PCRE to crash.
2811
2812
2813 Version 4.1 12-Mar-03
2814 ---------------------
2815
2816 1. Compiling with gcc -pedantic found a couple of places where casts were
2817 needed, and a string in dftables.c that was longer than standard compilers are
2818 required to support.
2819
2820 2. Compiling with Sun's compiler found a few more places where the code could
2821 be tidied up in order to avoid warnings.
2822
2823 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2824 first of these names is deprecated in the latest Autoconf in favour of the name
2825 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2826 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2827 analogy I have changed it to CFLAGS_FOR_BUILD.
2828
2829 4. Added -no-undefined to the linking command in the Makefile, because this is
2830 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2831 linking step for the pcreposix library.
2832
2833 5. PCRE was failing to diagnose the case of two named groups with the same
2834 name.
2835
2836 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2837 literal character that is needed in the subject for a match, and scans along to
2838 ensure that it is present before embarking on the full matching process. This
2839 saves time in cases of nested unlimited repeats that are never going to match.
2840 Problem: the scan can take a lot of time if the subject is very long (e.g.
2841 megabytes), thus penalizing straightforward matches. It is now done only if the
2842 amount of subject to be scanned is less than 1000 bytes.
2843
2844 7. A lesser problem with the same optimization is that it was recording the
2845 first character of an anchored pattern as "needed", thus provoking a search
2846 right along the subject, even when the first match of the pattern was going to
2847 fail. The "needed" character is now not set for anchored patterns, unless it
2848 follows something in the pattern that is of non-fixed length. Thus, it still
2849 fulfils its original purpose of finding quick non-matches in cases of nested
2850 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2851
2852
2853 Version 4.0 17-Feb-03
2854 ---------------------
2855
2856 1. If a comment in an extended regex that started immediately after a meta-item
2857 extended to the end of string, PCRE compiled incorrect data. This could lead to
2858 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2859
2860 2. Moved to autoconf 2.53 and libtool 1.4.2.
2861
2862 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2863 the special perltest8 script is no longer needed - all the tests can be run
2864 from a single perltest script.
2865
2866 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2867 by \s. It has now been removed in PCRE. This means it isn't recognized as
2868 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2869 class [:space:] *does* include VT, thereby creating a mess.
2870
2871 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2872 space and tab.
2873
2874 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2875 its new features into the main test script, reducing the number of scripts.
2876
2877 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2878 were backward compatible, and made the (?i) apply to the whole pattern, as if
2879 /i were given. Now it behaves more logically, and applies the option setting
2880 only to what follows. PCRE has been changed to follow suit. However, if it
2881 finds options settings right at the start of the pattern, it extracts them into
2882 the global options, as before. Thus, they show up in the info data.
2883
2884 8. Added support for the \Q...\E escape sequence. Characters in between are
2885 treated as literals. This is slightly different from Perl in that $ and @ are
2886 also handled as literals inside the quotes. In Perl, they will cause variable
2887 interpolation. Note the following examples:
2888
2889 Pattern PCRE matches Perl matches
2890
2891 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2892 \Qabc\$xyz\E abc\$xyz abc\$xyz
2893 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2894
2895 For compatibility with Perl, \Q...\E sequences are recognized inside character
2896 classes as well as outside them.
2897
2898 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2899 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2900 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2901 signed/unsigned warnings.
2902
2903 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2904 option for pcretest, so I've replaced it by a simple function that does just
2905 that job.
2906
2907 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2908 "pcregrep -".
2909
2910 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2911 Java package. This provides some syntactic sugar for simple cases of what my
2912 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2913 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2914 item, you can use this simplified notation. Note that only makes sense with
2915 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2916 greediness, whatever the setting of the PCRE_UNGREEDY option.
2917
2918 13. A change of greediness default within a pattern was not taking effect at
2919 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2920 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2921 was abstracted outside.
2922
2923 14. PCRE now supports the \G assertion. It is true when the current matching
2924 position is at the start point of the match. This differs from \A when the
2925 starting offset is non-zero. Used with the /g option of pcretest (or similar
2926 code), it works in the same way as it does for Perl's /g option. If all
2927 alternatives of a regex begin with \G, the expression is anchored to the start
2928 match position, and the "anchored" flag is set in the compiled expression.
2929
2930 15. Some bugs concerning the handling of certain option changes within patterns
2931 have been fixed. These applied to options other than (?ims). For example,
2932 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2933 been the other way round. Some of this was related to change 7 above.
2934
2935 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2936 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2937 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2938 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2939
2940 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2941 mode. Unlike ".", it always matches newline, whatever the setting of
2942 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2943 assertions. Perl allows it, but it doesn't (in general) work because it can't
2944 calculate the length of the lookbehind. At least, that's the case for Perl
2945 5.8.0 - I've been told they are going to document that it doesn't work in
2946 future.
2947
2948 18. Added an error diagnosis for escapes that PCRE does not support: these are
2949 \L, \l, \N, \P, \p, \U, \u, and \X.
2950
2951 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2952 reading past the end of the pattern in cases such as /[abcd/.
2953
2954 20. PCRE was getting more memory than necessary for patterns with classes that
2955 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2956
2957 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2958 compiling PCRE for use with Virtual Pascal.
2959
2960 22. Small fix to the Makefile to make it work properly if the build is done
2961 outside the source tree.
2962
2963 23. Added a new extension: a condition to go with recursion. If a conditional
2964 subpattern starts with (?(R) the "true" branch is used if recursion has
2965 happened, whereas the "false" branch is used only at the top level.
2966
2967 24. When there was a very long string of literal characters (over 255 bytes
2968 without UTF support, over 250 bytes with UTF support), the computation of how
2969 much memory was required could be incorrect, leading to segfaults or other
2970 strange effects.
2971
2972 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2973 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2974 there was a subsequent back reference to those brackets. This meant that, for
2975 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2976 possible to check for precisely this case. All we can do is abandon the
2977 optimization if .* occurs inside capturing brackets when there are any back
2978 references whatsoever. (See below for a better fix that came later.)
2979
2980 26. The handling of the optimization for finding the first character of a
2981 non-anchored pattern, and for finding a character that is required later in the
2982 match were failing in some cases. This didn't break the matching; it just
2983 failed to optimize when it could. The way this is done has been re-implemented.
2984
2985 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2986
2987 28. Added a new feature that provides some of the functionality that Perl
2988 provides with (?{...}). The facility is termed a "callout". The way it is done
2989 in PCRE is for the caller to provide an optional function, by setting
2990 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2991 global variable. By default it is unset, which disables all calling out. To get
2992 the function called, the regex must include (?C) at appropriate points. This
2993 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2994 This provides a means of identifying different callout points. When PCRE
2995 reaches such a point in the regex, if pcre_callout has been set, the external
2996 function is called. It is provided with data in a structure called
2997 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2998 matching continues; if it returns a non-zero value, the match at the current
2999 point fails. However, backtracking will occur if possible. [This was changed
3000 later and other features added - see item 49 below.]
3001
3002 29. pcretest is upgraded to test the callout functionality. It provides a
3003 callout function that displays information. By default, it shows the start of
3004 the match and the current position in the text. There are some new data escapes
3005 to vary what happens:
3006
3007 \C+ in addition, show current contents of captured substrings
3008 \C- do not supply a callout function
3009 \C!n return 1 when callout number n is reached
3010 \C!n!m return 1 when callout number n is reached for the mth time
3011
3012 30. If pcregrep was called with the -l option and just a single file name, it
3013 output "<stdin>" if a match was found, instead of the file name.
3014
3015 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3016 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3017 pcre_exec(). This saves a malloc/free per call. The default value of
3018 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3019 when configuring.
3020
3021 32. The default maximum size of a compiled pattern is 64K. There have been a
3022 few cases of people hitting this limit. The code now uses macros to handle the
3023 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3024 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3025 configuring. Tests 2 and 5 work only with 2-byte links because they output
3026 debugging information about compiled patterns.
3027
3028 33. Internal code re-arrangements:
3029
3030 (a) Moved the debugging function for printing out a compiled regex into
3031 its own source file (printint.c) and used #include to pull it into
3032 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3033 separate copies.
3034
3035 (b) Defined the list of op-code names for debugging as a macro in
3036 internal.h so that it is next to the definition of the opcodes.
3037
3038 (c) Defined a table of op-code lengths for simpler skipping along compiled
3039 code. This is again a macro in internal.h so that it is next to the
3040 definition of the opcodes.
3041
3042 34. Added support for recursive calls to individual subpatterns, along the
3043 lines of Robin Houston's patch (but implemented somewhat differently).
3044
3045 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3046 allow it to read and process whole directories in Win32. This code was
3047 contributed by Lionel Fourquaux; it has not been tested by me.
3048
3049 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3050 used to name a group. Names consist of alphanumerics and underscores, and must
3051 be unique. Back references use the syntax (?P=name) and recursive calls use
3052 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3053 numbers. The function pcre_fullinfo() can be used after compilation to extract
3054 a name/number map. There are three relevant calls:
3055
3056 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3057 PCRE_INFO_NAMECOUNT yields the number of entries
3058 PCRE_INFO_NAMETABLE yields a pointer to the map.
3059
3060 The map is a vector of fixed-size entries. The size of each entry depends on
3061 the length of the longest name used. The first two bytes of each entry are the
3062 group number, most significant byte first. There follows the corresponding
3063 name, zero terminated. The names are in alphabetical order.
3064
3065 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3066 case instead of 255. Making it the same both with and without UTF-8 support
3067 means that the same test output works with both.
3068
3069 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3070 calling malloc() with a zero argument.
3071
3072 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3073 optimization. I've improved things by keeping a bitmap of backreferences with
3074 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3075 fact referenced, the optimization can be applied. It is unlikely that a
3076 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3077 the match to follow \n) will appear inside brackets with a number greater than
3078 31, but if it does, any back reference > 31 suppresses the optimization.
3079
3080 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3081 of disabling numbered capturing parentheses. Any opening parenthesis that is
3082 not followed by ? behaves as if it were followed by ?: but named parentheses
3083 can still be used for capturing (and they will acquire numbers in the usual
3084 way).
3085
3086 41. Redesigned the return codes from the match() function into yes/no/error so
3087 that errors can be passed back from deep inside the nested calls. A malloc
3088 failure while inside a recursive subpattern call now causes the
3089 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3090
3091 42. It is now possible to set a limit on the number of times the match()
3092 function is called in a call to pcre_exec(). This facility makes it possible to
3093 limit the amount of recursion and backtracking, though not in a directly
3094 obvious way, because the match() function is used in a number of different
3095 circumstances. The count starts from zero for each position in the subject
3096 string (for non-anchored patterns). The default limit is, for compatibility, a
3097 large number, namely 10 000 000. You can change this in two ways:
3098
3099 (a) When configuring PCRE before making, you can use --with-match-limit=n
3100 to set a default value for the compiled library.
3101
3102 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3103 a different value is set. See 45 below.
3104
3105 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3106
3107 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3108 of things that can be changed at compile time. The first argument specifies
3109 what is wanted and the second points to where the information is to be placed.
3110 The current list of available information is:
3111
3112 PCRE_CONFIG_UTF8
3113
3114 The output is an integer that is set to one if UTF-8 support is available;
3115 otherwise it is set to zero.
3116
3117 PCRE_CONFIG_NEWLINE
3118
3119 The output is an integer that it set to the value of the code that is used for
3120 newline. It is either LF (10) or CR (13).
3121
3122 PCRE_CONFIG_LINK_SIZE
3123
3124 The output is an integer that contains the number of bytes used for internal
3125 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3126
3127 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3128
3129 The output is an integer that contains the threshold above which the POSIX
3130 interface uses malloc() for output vectors. See item 31 above.
3131
3132 PCRE_CONFIG_MATCH_LIMIT
3133
3134 The output is an unsigned integer that contains the default limit of the number
3135 of match() calls in a pcre_exec() execution. See 42 above.
3136
3137 44. pcretest has been upgraded by the addition of the -C option. This causes it
3138 to extract all the available output from the new pcre_config() function, and to
3139 output it. The program then exits immediately.
3140
3141 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3142 order to support additional features. One way would have been to define
3143 pcre_exec2() (for example) with extra arguments, but this would not have been
3144 extensible, and would also have required all calls to the original function to
3145 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3146 is used for passing in "extra" data from pcre_study().
3147
3148 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3149 contains the following fields:
3150
3151 flags a bitmap indicating which of the following fields are set
3152 study_data opaque data from pcre_study()
3153 match_limit a way of specifying a limit on match() calls for a specific
3154 call to pcre_exec()
3155 callout_data data for callouts (see 49 below)
3156
3157 The flag bits are also defined in pcre.h, and are
3158
3159 PCRE_EXTRA_STUDY_DATA
3160 PCRE_EXTRA_MATCH_LIMIT
3161 PCRE_EXTRA_CALLOUT_DATA
3162
3163 The pcre_study() function now returns one of these new pcre_extra blocks, with
3164 the actual study data pointed to by the study_data field, and the
3165 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3166 before. That is, this change is entirely upwards-compatible and requires no
3167 change to existing code.
3168
3169 If you want to pass in additional data to pcre_exec(), you can either place it
3170 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3171 block.
3172
3173 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3174 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3175 times with different match limits, until it finds the minimum value needed for
3176 pcre_exec() to complete. The value is then output. This can be instructive; for
3177 most simple matches the number is quite small, but for pathological cases it
3178 gets very large very quickly.
3179
3180 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3181 returns the size of the data block pointed to by the study_data field in a
3182 pcre_extra block, that is, the value that was passed as the argument to
3183 pcre_malloc() when PCRE was getting memory in which to place the information
3184 created by pcre_study(). The fourth argument should point to a size_t variable.
3185 pcretest has been extended so that this information is shown after a successful
3186 pcre_study() call when information about the compiled regex is being displayed.
3187
3188 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3189 because what follows is always an absolute path. (Later: it turns out that this
3190 is more than cosmetic for MinGW, because it doesn't like empty path
3191 components.)
3192
3193 49. Some changes have been made to the callout feature (see 28 above):
3194
3195 (i) A callout function now has three choices for what it returns:
3196
3197 0 => success, carry on matching
3198 > 0 => failure at this point, but backtrack if possible
3199 < 0 => serious error, return this value from pcre_exec()
3200
3201 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3202 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3203 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3204 use by callout functions. It will never be used by PCRE itself.
3205
3206 (ii) The pcre_extra structure (see 45 above) has a void * field called
3207 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3208 pcre_callout_block structure has a field of the same name. The contents of
3209 the field passed in the pcre_extra structure are passed to the callout
3210 function in the corresponding field in the callout block. This makes it
3211 easier to use the same callout-containing regex from multiple threads. For
3212 testing, the pcretest program has a new data escape
3213
3214 \C*n pass the number n (may be negative) as callout_data
3215
3216 If the callout function in pcretest receives a non-zero value as
3217 callout_data, it returns that value.
3218
3219 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3220 there were some redundant $(CFLAGS) in commands that are now specified as
3221 $(LINK), which already includes $(CFLAGS).
3222
3223 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3224 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3225 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3226 one-byte characters throughout. Note that case-insensitive matching applies
3227 only to characters whose values are less than 256. PCRE doesn't support the
3228 notion of cases for higher-valued characters.
3229
3230 (i) A character class whose characters are all within 0-255 is handled as
3231 a bit map, and the map is inverted for negative classes. Previously, a
3232 character > 255 always failed to match such a class; however it should
3233 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3234
3235 (ii) A negated character class with a single character < 255 is coded as
3236 "not this character" (OP_NOT). This wasn't working properly when the test
3237 character was multibyte, either singly or repeated.
3238
3239 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3240 mode, for example: \x{100}{2,3}.
3241
3242 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3243 singly or repeated) now correctly test multibyte characters. However,
3244 PCRE doesn't recognize any characters with values greater than 255 as
3245 digits, spaces, or word characters. Such characters always match \D, \S,
3246 and \W, and never match \d, \s, or \w.
3247
3248 (v) Classes may now contain characters and character ranges with values
3249 greater than 255. For example: [ab\x{100}-\x{400}].
3250
3251 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3252 PCRE in UTF-8 mode.
3253
3254 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3255 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3256 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3257 value.)
3258
3259 53. The single man page has become too large. I have therefore split it up into
3260 a number of separate man pages. These also give rise to individual HTML pages;
3261 these are now put in a separate directory, and there is an index.html page that
3262 lists them all. Some hyperlinking between the pages has been installed.
3263
3264 54. Added convenience functions for handling named capturing parentheses.
3265
3266 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3267 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3268 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3269 are faulted.
3270
3271 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3272 calling configure. These values are used when compiling the dftables.c program
3273 which is run to generate the source of the default character tables. They
3274 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3275 you will need to set these values.
3276
3277 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3278
3279
3280 Version 3.9 02-Jan-02
3281 ---------------------
3282
3283 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3284
3285 2. If --disable-static was given, the building process failed when trying to
3286 build pcretest and pcregrep. (For some reason it was using libtool to compile
3287 them, which is not right, as they aren't part of the library.)
3288
3289
3290 Version 3.8 18-Dec-01
3291 ---------------------
3292
3293 1. The experimental UTF-8 code was completely screwed up. It was packing the
3294 bytes in the wrong order. How dumb can you get?
3295
3296
3297 Version 3.7 29-Oct-01
3298 ---------------------
3299
3300 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3301 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3302 this didn't happen under Solaris 8, where I normally test things.
3303
3304 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3305 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3306 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3307 this fix an hour or so after the initial 3.7 release.)
3308
3309
3310 Version 3.6 23-Oct-01
3311 ---------------------
3312
3313 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3314 offsets passed as NULL with zero offset count.
3315
3316 2. The config.guess and config.sub files had not been updated when I moved to
3317 the latest autoconf.
3318
3319
3320 Version 3.5 15-Aug-01
3321 ---------------------
3322
3323 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3324 had been forgotten.
3325
3326 2. By using declared but undefined structures, we can avoid using "void"
3327 definitions in pcre.h while keeping the internal definitions of the structures
3328 private.
3329
3330 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3331 user point of view, this means that both static and shared libraries are built
3332 by default, but this can be individually controlled. More of the work of
3333 handling this static/shared cases is now inside libtool instead of PCRE's make
3334 file.
3335
3336 4. The pcretest utility is now installed along with pcregrep because it is
3337 useful for users (to test regexs) and by doing this, it automatically gets
3338 relinked by libtool. The documentation has been turned into a man page, so
3339 there are now .1, .txt, and .html versions in /doc.
3340
3341 5. Upgrades to pcregrep:
3342 (i) Added long-form option names like gnu grep.
3343 (ii) Added --help to list all options with an explanatory phrase.
3344 (iii) Added -r, --recursive to recurse into sub-directories.
3345 (iv) Added -f, --file to read patterns from a file.
3346
3347 6. pcre_exec() was referring to its "code" argument before testing that
3348 argument for NULL (and giving an error if it was NULL).
3349
3350 7. Upgraded Makefile.in to allow for compiling in a different directory from
3351 the source directory.
3352
3353 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3354 options bits, the pointer it was passed was to an int instead of to an unsigned
3355 long int. This mattered only on 64-bit systems.
3356
3357 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3358 generated) instead of pcre.in, which it its source. Also made the same change
3359 in several of the .c files.
3360
3361 10. A new release of gcc defines printf() as a macro, which broke pcretest
3362 because it had an ifdef in the middle of a string argument for printf(). Fixed
3363 by using separate calls to printf().
3364
3365 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3366 script, to force use of CR or LF instead of \n in the source. On non-Unix
3367 systems, the value can be set in config.h.
3368
3369 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3370 absolute limit. Changed the text of the error message to make this clear, and
3371 likewise updated the man page.
3372
3373 13. The limit of 99 on the number of capturing subpatterns has been removed.
3374 The new limit is 65535, which I hope will not be a "real" limit.
3375
3376
3377 Version 3.4 22-Aug-00
3378 ---------------------
3379
3380 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3381
3382 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3383
3384
3385 Version 3.3 01-Aug-00
3386 ---------------------
3387
3388 1. If an octal character was given, but the value was greater than \377, it
3389 was not getting masked to the least significant bits, as documented. This could
3390 lead to crashes in some systems.
3391
3392 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3393 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3394
3395 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3396 These just pass their arguments on to (pcre_free)(), but they are provided
3397 because some uses of PCRE bind it to non-C systems that can call its functions,
3398 but cannot call free() or pcre_free() directly.
3399
3400 4. Add "make test" as a synonym for "make check". Corrected some comments in
3401 the Makefile.
3402
3403 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3404 Makefile.
3405
3406 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3407 command called pgrep for grepping around the active processes.
3408
3409 7. Added the beginnings of support for UTF-8 character strings.
3410
3411 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3412 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3413 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3414 out for the ar command.)
3415
3416
3417 Version 3.2 12-May-00
3418 ---------------------
3419
3420 This is purely a bug fixing release.
3421
3422 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3423 of ZA. This was just one example of several cases that could provoke this bug,
3424 which was introduced by change 9 of version 2.00. The code for breaking
3425 infinite loops after an iteration that matches an empty string was't working
3426 correctly.
3427
3428 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3429 when matched against abbab (for example). After matching an empty string, it
3430 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3431 caused it to match further down the string than it should.
3432
3433 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3434 was there because it doesn't seem to be needed, and it causes trouble on some
3435 systems, as it is not a Standard C header. It has been removed.
3436
3437 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3438 were reported on the Macintosh. The changes were from
3439
3440 while ((c = *(++ptr)) != 0 && c != '\n');
3441 to
3442 while ((c = *(++ptr)) != 0 && c != '\n') ;
3443
3444 Totally extraordinary, but if that's what it takes...
3445
3446 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3447 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3448 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3449 assumes the way PCRE uses memmove() (always moving upwards).
3450
3451 6. PCRE is being used in one environment where strchr() is not available. There
3452 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3453 faster code anyway.
3454
3455
3456 Version 3.1 09-Feb-00
3457 ---------------------
3458
3459 The only change in this release is the fixing of some bugs in Makefile.in for
3460 the "install" target:
3461
3462 (1) It was failing to install pcreposix.h.
3463
3464 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3465
3466
3467 Version 3.0 01-Feb-00
3468 ---------------------
3469
3470 1. Add support for the /+ modifier to perltest (to output $` like it does in
3471 pcretest).
3472
3473 2. Add support for the /g modifier to perltest.
3474
3475 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3476 matches null strings.
3477
3478 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3479 pattern. Perl treats empty patterns specially - it reuses the most recent
3480 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3481 effect.
3482
3483 5. The POSIX interface was broken in that it was just handing over the POSIX
3484 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3485 required a bigger vector, with some working space on the end. This means that
3486 the POSIX wrapper now has to get and free some memory, and copy the results.
3487
3488 6. Added some simple autoconf support, placing the test data and the
3489 documentation in separate directories, re-organizing some of the
3490 information files, and making it build pcre-config (a GNU standard). Also added
3491 libtool support for building PCRE as a shared library, which is now the
3492 default.
3493
3494 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3495 09 are not valid octal constants. Single digits will be used for minor values
3496 less than 10.
3497
3498 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3499 existing programs that set these in the POSIX interface can use PCRE without
3500 modification.
3501
3502 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3503 return all that pcre_info() returns, plus additional data. The pcre_info()
3504 function is retained for compatibility, but is considered to be obsolete.
3505
3506 10. Added experimental recursion feature (?R) to handle one common case that
3507 Perl 5.6 will be able to do with (?p{...}).
3508
3509 11. Added support for POSIX character classes like [:alpha:], which Perl is
3510 adopting.
3511
3512
3513 Version 2.08 31-Aug-99
3514 ----------------------
3515
3516 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3517 trying to match at the startoffset position, but instead was moving forward to
3518 the next newline as if a previous match had failed.
3519
3520 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3521 and could get into a loop if a null string was matched other than at the start
3522 of the subject.
3523
3524 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3525 be distinguished at compile time, and for completeness also added PCRE_DATE.
3526
3527 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3528 in GnuWin32 environments.
3529
3530
3531 Version 2.07 29-Jul-99
3532 ----------------------
3533
3534 1. The documentation is now supplied in plain text form and HTML as well as in
3535 the form of man page sources.
3536
3537 2. C++ compilers don't like assigning (void *) values to other pointer types.
3538 In particular this affects malloc(). Although there is no problem in Standard
3539 C, I've put in casts to keep C++ compilers happy.
3540
3541 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3542 should be (const char *).
3543
3544 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3545 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3546 However, I haven't made this a standard facility. The documentation doesn't
3547 mention it, and the Makefile doesn't support it.
3548
3549 5. The Makefile now contains an "install" target, with editable destinations at
3550 the top of the file. The pcretest program is not installed.
3551
3552 6. pgrep -V now gives the PCRE version number and date.
3553
3554 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3555 causing the entire string to be ignored, instead of just the last character.
3556
3557 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3558 non-matching string, it can take a very, very long time, even for strings of
3559 quite modest length, because of the nested recursion. PCRE now does better in
3560 some of these cases. It does this by remembering the last required literal
3561 character in the pattern, and pre-searching the subject to ensure it is present
3562 before running the real match. In other words, it applies a heuristic to detect
3563 some types of certain failure quickly, and in the above example, if presented
3564 with a string that has no trailing " it gives "no match" very quickly.
3565
3566 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3567 other alternatives are tried instead.
3568
3569
3570 Version 2.06 09-Jun-99
3571 ----------------------
3572
3573 1. Change pcretest's output for amount of store used to show just the code
3574 space, because the remainder (the data block) varies in size between 32-bit and
3575 64-bit systems.
3576
3577 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3578 start matching at. This allows lookbehinds to work when searching for multiple
3579 occurrences in a string.
3580
3581 3. Added additional options to pcretest for testing multiple occurrences:
3582
3583 /+ outputs the rest of the string that follows a match
3584 /g loops for multiple occurrences, using the new startoffset argument
3585 /G loops for multiple occurrences by passing an incremented pointer
3586
3587 4. PCRE wasn't doing the "first character" optimization for patterns starting
3588 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3589 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3590 the letter 'x'. On long subject strings, this gives a significant speed-up.
3591
3592
3593 Version 2.05 21-Apr-99
3594 ----------------------
3595
3596 1. Changed the type of magic_number from int to long int so that it works
3597 properly on 16-bit systems.
3598
3599 2. Fixed a bug which caused patterns starting with .* not to work correctly
3600 when the subject string contained newline characters. PCRE was assuming
3601 anchoring for such patterns in all cases, which is not correct because .* will
3602 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3603 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3604 must be retried after every newline in the subject.
3605
3606
3607 Version 2.04 18-Feb-99
3608 ----------------------
3609
3610 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3611 computation of the store needed to hold the pattern was incorrect (too large).
3612 If such patterns were nested a few deep, this could multiply and become a real
3613 problem.
3614
3615 2. Added /M option to pcretest to show the memory requirement of a specific
3616 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3617
3618 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3619 compiled in such a way that the backtracking after subsequent failure was
3620 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3621 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3622
3623
3624 Version 2.03 02-Feb-99
3625 ----------------------
3626
3627 1. Fixed typo and small mistake in man page.
3628
3629 2. Added 4th condition (GPL supersedes if conflict) and created separate
3630 LICENCE file containing the conditions.
3631
3632 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3633 Perl, that is the internal \ allows the delimiter to be included in the
3634 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3635 the final delimiter, add \ to the end of the pattern (to test the error).
3636
3637 4. Added the convenience functions for extracting substrings after a successful
3638 match. Updated pcretest to make it able to test these functions.
3639
3640
3641 Version 2.02 14-Jan-99
3642 ----------------------
3643
3644 1. Initialized the working variables associated with each extraction so that
3645 their saving and restoring doesn't refer to uninitialized store.
3646
3647 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3648 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3649 fix the problem.
3650
3651 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3652 calls, and wasn't printing the correct value for compiling calls. Increased the
3653 default value of LOOPREPEAT, and the number of significant figures in the
3654 times.
3655
3656 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3657
3658 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3659 a building problem on Windows NT with a FAT file system.
3660
3661
3662 Version 2.01 21-Oct-98
3663 ----------------------
3664
3665 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3666 to character tables built by pcre_maketables() in the current locale. If NULL
3667 is passed, the default tables are used.
3668
3669
3670 Version 2.00 24-Sep-98
3671 ----------------------
3672
3673 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3674 it any more.
3675
3676 2. Allow quantification of (?>) groups, and make it work correctly.
3677
3678 3. The first character computation wasn't working for (?>) groups.
3679
3680 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3681 end of the subject) and add 5.005's \z, which really does match only at the
3682 very end of the subject.
3683
3684 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3685
3686 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3687 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3688 localized options. All options to pcre_study() were also removed.
3689
3690 7. Add other new features from 5.005:
3691
3692 $(?<= positive lookbehind
3693 $(?<! negative lookbehind
3694 (?imsx-imsx) added the unsetting capability
3695 such a setting is global if at outer level; local otherwise
3696 (?imsx-imsx:) non-capturing groups with option setting
3697 (?(cond)re|re) conditional pattern matching
3698
3699 A backreference to itself in a repeated group matches the previous
3700 captured string.
3701
3702 8. General tidying up of studying (both automatic and via "study")
3703 consequential on the addition of new assertions.
3704
3705 9. As in 5.005, unlimited repeated groups that could match an empty substring
3706 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3707 runtime if any iteration does actually match an empty substring.
3708
3709 10. Include the RunTest script in the distribution.
3710
3711 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3712 discrepancies, some of which were old and were also with respect to 5.004. They
3713 have now been fixed.
3714
3715
3716 Version 1.09 28-Apr-98
3717 ----------------------
3718
3719 1. A negated single character class followed by a quantifier with a minimum
3720 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3721 program crashes, or just wrong answers. This did not apply to negated classes
3722 containing more than one character, or to minima other than one.
3723
3724
3725 Version 1.08 27-Mar-98
3726 ----------------------
3727
3728 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3729
3730 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3731 latter must appear before anything that relies on it in the pattern.
3732
3733
3734 Version 1.07 16-Feb-98
3735 ----------------------
3736
3737 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3738 repeat of a potentially empty string).
3739
3740
3741 Version 1.06 23-Jan-98
3742 ----------------------
3743
3744 1. Added Markus Oberhumer's little patches for C++.
3745
3746 2. Literal strings longer than 255 characters were broken.
3747
3748
3749 Version 1.05 23-Dec-97
3750 ----------------------
3751
3752 1. Negated character classes containing more than one character were failing if
3753 PCRE_CASELESS was set at run time.
3754
3755
3756 Version 1.04 19-Dec-97
3757 ----------------------
3758
3759 1. Corrected the man page, where some "const" qualifiers had been omitted.
3760
3761 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3762 input syntax.
3763
3764 3. Fixed memory leak which occurred when a regex with back references was
3765 matched with an offsets vector that wasn't big enough. The temporary memory
3766 that is used in this case wasn't being freed if the match failed.
3767
3768 4. Tidied pcretest to ensure it frees memory that it gets.
3769
3770 5. Temporary memory was being obtained in the case where the passed offsets
3771 vector was exactly big enough.
3772
3773 6. Corrected definition of offsetof() from change 5 below.
3774
3775 7. I had screwed up change 6 below and broken the rules for the use of
3776 setjmp(). Now fixed.
3777
3778
3779 Version 1.03 18-Dec-97
3780 ----------------------
3781
3782 1. A erroneous regex with a missing opening parenthesis was correctly
3783 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3784 on some systems.
3785
3786 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3787 it was reported that one broken compiler failed on the former because "code" is
3788 also an independent variable.
3789
3790 3. The erroneous regex a[]b caused an array overrun reference.
3791
3792 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3793 fail on data ending with that character. (It was going on too far, and checking
3794 the next character, typically a binary zero.) This was specific to the
3795 optimized code for single-character negative classes.
3796
3797 5. Added a contributed patch from the TIN world which does the following:
3798
3799 + Add an undef for memmove, in case the the system defines a macro for it.
3800
3801 + Add a definition of offsetof(), in case there isn't one. (I don't know
3802 the reason behind this - offsetof() is part of the ANSI standard - but
3803 it does no harm).
3804
3805 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3806 most of the places where whitespace preceded '#'. I have given up and
3807 allowed the remaining 2 cases to be at the margin.
3808
3809 + Rename some variables in pcre to eliminate shadowing. This seems very
3810 pedantic, but does no harm, of course.
3811
3812 6. Moved the call to setjmp() into its own function, to get rid of warnings
3813 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3814
3815 7. Constructs such as \d{8,} were compiling into the equivalent of
3816 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3817 outcome, but in this particular case used more store than had been allocated,
3818 which caused the bug to be discovered because it threw up an internal error.
3819
3820 8. The debugging code in both pcre and pcretest for outputting the compiled
3821 form of a regex was going wrong in the case of back references followed by
3822 curly-bracketed repeats.
3823
3824
3825 Version 1.02 12-Dec-97
3826 ----------------------
3827
3828 1. Typos in pcre.3 and comments in the source fixed.
3829
3830 2. Applied a contributed patch to get rid of places where it used to remove
3831 'const' from variables, and fixed some signed/unsigned and uninitialized
3832 variable warnings.
3833
3834 3. Added the "runtest" target to Makefile.
3835
3836 4. Set default compiler flag to -O2 rather than just -O.
3837
3838
3839 Version 1.01 19-Nov-97
3840 ----------------------
3841
3842 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3843 like /([ab]*)*/, that is, for classes with more than one character in them.
3844
3845 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3846 as /((?>a*))*/ (a PCRE_EXTRA facility).
3847
3848
3849 Version 1.00 18-Nov-97
3850 ----------------------
3851
3852 1. Added compile-time macros to support systems such as SunOS4 which don't have
3853 memmove() or strerror() but have other things that can be used instead.
3854
3855 2. Arranged that "make clean" removes the executables.
3856
3857
3858 Version 0.99 27-Oct-97
3859 ----------------------
3860
3861 1. Fixed bug in code for optimizing classes with only one character. It was
3862 initializing a 32-byte map regardless, which could cause it to run off the end
3863 of the memory it had got.
3864
3865 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3866
3867
3868 Version 0.98 22-Oct-97
3869 ----------------------
3870
3871 1. Fixed bug in code for handling temporary memory usage when there are more
3872 back references than supplied space in the ovector. This could cause segfaults.
3873
3874
3875 Version 0.97 21-Oct-97
3876 ----------------------
3877
3878 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3879
3880 2. Optimized negated single characters not to use a bit map.
3881
3882 3. Brought error texts together as macro definitions; clarified some of them;
3883 fixed one that was wrong - it said "range out of order" when it meant "invalid
3884 escape sequence".
3885
3886 4. Changed some char * arguments to const char *.
3887
3888 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3889
3890 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3891 pcretest.
3892
3893
3894 Version 0.96 16-Oct-97
3895 ----------------------
3896
3897 1. Added a simple "pgrep" utility to the distribution.
3898
3899 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3900 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3901 where "ddd" means "one or more decimal digits".
3902
3903 3. Fixed serious bug. If a pattern had a back reference, but the call to
3904 pcre_exec() didn't supply a large enough ovector to record the related
3905 identifying subpattern, the match always failed. PCRE now remembers the number
3906 of the largest back reference, and gets some temporary memory in which to save
3907 the offsets during matching if necessary, in order to ensure that
3908 backreferences always work.
3909
3910 4. Increased the compatibility with Perl in a number of ways:
3911
3912 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3913 to request this handling. The option can be set at compile or exec time.
3914
3915 (b) $ matches before a terminating newline by default; an option
3916 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3917 mode). The option can be set at compile or exec time.
3918
3919 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3920 the same as Perl's. If the decimal number it represents is less than 10
3921 or there aren't that many previous left capturing parentheses, an octal
3922 escape is read. Inside a character class, it's always an octal escape,
3923 even if it is a single digit.
3924
3925 (d) An escaped but undefined alphabetic character is taken as a literal,
3926 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3927 escapes.
3928
3929 (e) {0} is now permitted. (The previous item is removed from the compiled
3930 pattern).
3931
3932 5. Changed all the names of code files so that the basic parts are no longer
3933 than 10 characters, and abolished the teeny "globals.c" file.
3934
3935 6. Changed the handling of character classes; they are now done with a 32-byte
3936 bit map always.
3937
3938 7. Added the -d and /D options to pcretest to make it possible to look at the
3939 internals of compilation without having to recompile pcre.
3940
3941
3942 Version 0.95 23-Sep-97
3943 ----------------------
3944
3945 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3946 \x20 at the start of a run of normal characters. These were being treated as
3947 real characters, instead of the source characters being re-checked.
3948
3949
3950 Version 0.94 18-Sep-97
3951 ----------------------
3952
3953 1. The functions are now thread-safe, with the caveat that the global variables
3954 containing pointers to malloc() and free() or alternative functions are the
3955 same for all threads.
3956
3957 2. Get pcre_study() to generate a bitmap of initial characters for non-
3958 anchored patterns when this is possible, and use it if passed to pcre_exec().
3959
3960
3961 Version 0.93 15-Sep-97
3962 ----------------------
3963
3964 1. /(b)|(:+)/ was computing an incorrect first character.
3965
3966 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3967 but not actually doing anything yet.
3968
3969 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3970 as Perl does (e.g. [-az] or [az-]).
3971
3972 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3973 all possible positions.
3974
3975 5. Split up into different modules to avoid including unneeded functions in a
3976 compiled binary. However, compile and exec are still in one module. The "study"
3977 function is split off.
3978
3979 6. The character tables are now in a separate module whose source is generated
3980 by an auxiliary program - but can then be edited by hand if required. There are
3981 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3982 toupper() in the code.
3983
3984 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3985 make them global. Abolish the function for setting them, as the caller can now
3986 set them directly.
3987
3988
3989 Version 0.92 11-Sep-97
3990 ----------------------
3991
3992 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3993 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3994
3995 2. Caseless matching was not working in character classes if the characters in
3996 the pattern were in upper case.
3997
3998 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3999
4000 4. Make PCRE_ANCHORED public and accept as a compile option.
4001
4002 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4003 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4004 pass them.
4005
4006 6. Give an error if bad option bits passed at compile or run time.
4007
4008 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4009 pcretest to cause it to pass that flag.
4010
4011 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4012 options, and the first character, if set.
4013
4014 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4015
4016
4017 Version 0.91 10-Sep-97
4018 ----------------------
4019
4020 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4021 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4022
4023 2. PCRE was looping on encountering an indefinitely repeated back reference to
4024 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4025 Perl does - treats the match as successful.
4026
4027 ****

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