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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 579 - (show annotations) (download)
Wed Nov 24 17:39:25 2010 UTC (4 years ago) by ph10
File size: 196454 byte(s)
Tidies of documenation and code while preparing for release.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.11 24-Nov-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 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
71 can match an empty string. The code to do it in pcretest and pcredemo
72 needed fixing:
73
74 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
75 only one byte after an empty string match just before CRLF (this case
76 just got forgotten; "any" and "anycrlf" were OK).
77
78 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
79 mode when an empty string match preceded an ASCII character followed by
80 a non-ASCII character. (The code for advancing by one character rather
81 than one byte was nonsense.)
82
83 (c) The pcredemo.c sample program did not have any code at all to handle
84 the cases when CRLF is a valid newline sequence.
85
86 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
87 as a starting offset was within the subject string. There is now a new
88 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
89 negative or greater than the length of the string. In order to test this,
90 pcretest is extended to allow the setting of negative starting offsets.
91
92 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
93 starting offset points to the beginning of a UTF-8 character was
94 unnecessarily clumsy. I tidied it up.
95
96 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
97 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
98
99 17. Nobody had reported that the --include_dir option, which was added in
100 release 7.7 should have been called --include-dir (hyphen, not underscore)
101 for compatibility with GNU grep. I have changed it to --include-dir, but
102 left --include_dir as an undocumented synonym, and the same for
103 --exclude-dir, though that is not available in GNU grep, at least as of
104 release 2.5.4.
105
106 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
107 characters from a string of bytes) have been redefined so as not to use
108 loops, in order to improve performance in some environments. At the same
109 time, I abstracted some of the common code into auxiliary macros to save
110 repetition (this should not affect the compiled code).
111
112 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
113 compile-time error is now given if \c is not followed by an ASCII
114 character, that is, a byte less than 128. (In EBCDIC mode, the code is
115 different, and any byte value is allowed.)
116
117 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
118 START_OPTIMIZE option, which is now allowed at compile time - but just
119 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
120 to pcregrep and other applications that have no direct access to PCRE
121 options. The new /Y option in pcretest sets this option when calling
122 pcre_compile().
123
124 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
125 back references. Groups containing recursive back references were forced to
126 be atomic by that change, but in the case of named groups, the amount of
127 memory required was incorrectly computed, leading to "Failed: internal
128 error: code overflow". This has been fixed.
129
130
131 Version 8.10 25-Jun-2010
132 ------------------------
133
134 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
135 THEN.
136
137 2. (*ACCEPT) was not working when inside an atomic group.
138
139 3. Inside a character class, \B is treated as a literal by default, but
140 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
141 causes the error). The code is unchanged, but I tidied the documentation.
142
143 4. Inside a character class, PCRE always treated \R and \X as literals,
144 whereas Perl faults them if its -w option is set. I have changed PCRE so
145 that it faults them when PCRE_EXTRA is set.
146
147 5. Added support for \N, which always matches any character other than
148 newline. (It is the same as "." when PCRE_DOTALL is not set.)
149
150 6. When compiling pcregrep with newer versions of gcc which may have
151 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
152 declared with attribute warn_unused_result" were given. Just casting the
153 result to (void) does not stop the warnings; a more elaborate fudge is
154 needed. I've used a macro to implement this.
155
156 7. Minor change to pcretest.c to avoid a compiler warning.
157
158 8. Added four artifical Unicode properties to help with an option to make
159 \s etc use properties (see next item). The new properties are: Xan
160 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
161
162 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
163 use Unicode properties. (*UCP) at the start of a pattern can be used to set
164 this option. Modified pcretest to add /W to test this facility. Added
165 REG_UCP to make it available via the POSIX interface.
166
167 10. Added --line-buffered to pcregrep.
168
169 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
170 studied, and the match started with a letter with a code point greater than
171 127 whose first byte was different to the first byte of the other case of
172 the letter, the other case of this starting letter was not recognized
173 (#976).
174
175 12. If a pattern that was studied started with a repeated Unicode property
176 test, for example, \p{Nd}+, there was the theoretical possibility of
177 setting up an incorrect bitmap of starting bytes, but fortunately it could
178 not have actually happened in practice until change 8 above was made (it
179 added property types that matched character-matching opcodes).
180
181 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
182 possible starting bytes for non-anchored patterns.
183
184 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
185 \R, and also a number of cases that involve Unicode properties, both
186 explicit and implicit when PCRE_UCP is set.
187
188 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
189 input, it could crash or give wrong results if characters with values
190 greater than 0xc0 were present in the subject string. (Detail: it assumed
191 UTF-8 input when processing these items.)
192
193 16. Added a lot of (int) casts to avoid compiler warnings in systems where
194 size_t is 64-bit (#991).
195
196 17. Added a check for running out of memory when PCRE is compiled with
197 --disable-stack-for-recursion (#990).
198
199 18. If the last data line in a file for pcretest does not have a newline on
200 the end, a newline was missing in the output.
201
202 19. The default pcre_chartables.c file recognizes only ASCII characters (values
203 less than 128) in its various bitmaps. However, there is a facility for
204 generating tables according to the current locale when PCRE is compiled. It
205 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
206 characters, are recognized by isspace() and therefore were getting set in
207 these tables, and indeed these tables seem to approximate to ISO 8859. This
208 caused a problem in UTF-8 mode when pcre_study() was used to create a list
209 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
210 which of course cannot start UTF-8 characters. I have changed the code so
211 that only real ASCII characters (less than 128) and the correct starting
212 bytes for UTF-8 encodings are set for characters greater than 127 when in
213 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
214 altogether.)
215
216 20. Added the /T option to pcretest so as to be able to run tests with non-
217 standard character tables, thus making it possible to include the tests
218 used for 19 above in the standard set of tests.
219
220 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
221 reference to a subpattern the other side of a comment that contains an
222 opening parenthesis caused either an internal compiling error, or a
223 reference to the wrong subpattern.
224
225
226 Version 8.02 19-Mar-2010
227 ------------------------
228
229 1. The Unicode data tables have been updated to Unicode 5.2.0.
230
231 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
232 configured.
233
234 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
235 original author of that file, following a query about its status.
236
237 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
238 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
239
240 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
241 quantifier applied to a forward-referencing subroutine call, could compile
242 incorrect code or give the error "internal error: previously-checked
243 referenced subpattern not found".
244
245 6. Both MS Visual Studio and Symbian OS have problems with initializing
246 variables to point to external functions. For these systems, therefore,
247 pcre_malloc etc. are now initialized to local functions that call the
248 relevant global functions.
249
250 7. There were two entries missing in the vectors called coptable and poptable
251 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
252 I've fixed the data, and added a kludgy way of testing at compile time that
253 the lengths are correct (equal to the number of opcodes).
254
255 8. Following on from 7, I added a similar kludge to check the length of the
256 eint vector in pcreposix.c.
257
258 9. Error texts for pcre_compile() are held as one long string to avoid too
259 much relocation at load time. To find a text, the string is searched,
260 counting zeros. There was no check for running off the end of the string,
261 which could happen if a new error number was added without updating the
262 string.
263
264 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
265
266 11. \K was not working if it appeared in an atomic group or in a group that
267 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
268 \K is "not well defined" if used in an assertion. PCRE now accepts it if
269 the assertion is positive, but not if it is negative.
270
271 12. Change 11 fortuitously reduced the size of the stack frame used in the
272 "match()" function of pcre_exec.c by one pointer. Forthcoming
273 implementation of support for (*MARK) will need an extra pointer on the
274 stack; I have reserved it now, so that the stack frame size does not
275 decrease.
276
277 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
278 item in branch that calls a recursion is a subroutine call - as in the
279 second branch in the above example - was incorrectly given the compile-
280 time error "recursive call could loop indefinitely" because pcre_compile()
281 was not correctly checking the subroutine for matching a non-empty string.
282
283 14. The checks for overrunning compiling workspace could trigger after an
284 overrun had occurred. This is a "should never occur" error, but it can be
285 triggered by pathological patterns such as hundreds of nested parentheses.
286 The checks now trigger 100 bytes before the end of the workspace.
287
288 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
289
290
291 Version 8.01 19-Jan-2010
292 ------------------------
293
294 1. If a pattern contained a conditional subpattern with only one branch (in
295 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
296 computed the wrong minimum data length (which is of course zero for such
297 subpatterns). This could cause incorrect "no match" results.
298
299 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
300 the pattern is reset in the first branch, pcre_compile() failed with
301 "internal error: code overflow at offset...". This happened only when
302 the reset was to the original external option setting. (An optimization
303 abstracts leading options settings into an external setting, which was the
304 cause of this.)
305
306 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
307 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
308 assertion pattern did not match (meaning that the assertion was true), it
309 was incorrectly treated as false if the SKIP had been reached during the
310 matching. This also applied to assertions used as conditions.
311
312 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
313 assertion subpattern, including such a pattern used as a condition,
314 unpredictable results occurred, instead of the error return
315 PCRE_ERROR_DFA_UITEM.
316
317 5. The C++ GlobalReplace function was not working like Perl for the special
318 situation when an empty string is matched. It now does the fancy magic
319 stuff that is necessary.
320
321 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
322 removed. (These were left over from very, very early versions of PCRE.)
323
324 7. Some cosmetic changes to the code to make life easier when compiling it
325 as part of something else:
326
327 (a) Change DEBUG to PCRE_DEBUG.
328
329 (b) In pcre_compile(), rename the member of the "branch_chain" structure
330 called "current" as "current_branch", to prevent a collision with the
331 Linux macro when compiled as a kernel module.
332
333 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
334 prevent a collision with the Linux macro when compiled as a kernel
335 module.
336
337 8. In pcre_compile() there are some checks for integer overflows that used to
338 cast potentially large values to (double). This has been changed to that
339 when building, a check for int64_t is made, and if it is found, it is used
340 instead, thus avoiding the use of floating point arithmetic. (There is no
341 other use of FP in PCRE.) If int64_t is not found, the fallback is to
342 double.
343
344 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
345 2005 (difference between two addresses compared to an unsigned value).
346
347 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
348 custom one, because of the following reported problem in Windows:
349
350 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
351 under Win32.
352 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
353 therefore missing the function definition.
354 - The compiler thus generates a "C" signature for the test function.
355 - The linker fails to find the "C" function.
356 - PCRE fails to configure if asked to do so against libbz2.
357
358 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
359 messages were output:
360
361 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
362 rerunning libtoolize, to keep the correct libtool macros in-tree.
363 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
364
365 I have done both of these things.
366
367 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
368 most of the time, it *can* run out if it is given a pattern that contains a
369 runaway infinite recursion. I updated the discussion in the pcrestack man
370 page.
371
372 13. Now that we have gone to the x.xx style of version numbers, the minor
373 version may start with zero. Using 08 or 09 is a bad idea because users
374 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
375 interpreted as invalid octal numbers. I've updated the previous comment in
376 configure.ac, and also added a check that gives an error if 08 or 09 are
377 used.
378
379 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
380 causing partial matching to fail when the end of the subject matched \W
381 in a UTF-8 pattern where \W was quantified with a minimum of 3.
382
383 15. There were some discrepancies between the declarations in pcre_internal.h
384 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
385 their definitions. The declarations used "const uschar *" and the
386 definitions used USPTR. Even though USPTR is normally defined as "const
387 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
388 reported that: "This difference in casting confuses some C++ compilers, for
389 example, SunCC recognizes above declarations as different functions and
390 generates broken code for hbpcre." I have changed the declarations to use
391 USPTR.
392
393 16. GNU libtool is named differently on some systems. The autogen.sh script now
394 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
395 (FreeBSD).
396
397 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
398 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
399 comment: "Figure out how to create a longlong from a string: strtoll and
400 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
401 instance, but it only takes 2 args instead of 3!"
402
403 18. A subtle bug concerned with back references has been fixed by a change of
404 specification, with a corresponding code fix. A pattern such as
405 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
406 refers, was giving matches when it shouldn't. For example, xa=xaaa would
407 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
408 same bug. Such groups have to be quantified to be useful, or contained
409 inside another quantified group. (If there's no repetition, the reference
410 can never match.) The problem arises because, having left the group and
411 moved on to the rest of the pattern, a later failure that backtracks into
412 the group uses the captured value from the final iteration of the group
413 rather than the correct earlier one. I have fixed this in PCRE by forcing
414 any group that contains a reference to itself to be an atomic group; that
415 is, there cannot be any backtracking into it once it has completed. This is
416 similar to recursive and subroutine calls.
417
418
419 Version 8.00 19-Oct-09
420 ----------------------
421
422 1. The table for translating pcre_compile() error codes into POSIX error codes
423 was out-of-date, and there was no check on the pcre_compile() error code
424 being within the table. This could lead to an OK return being given in
425 error.
426
427 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
428 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
429 in a Windows environment.
430
431 3. The pcregrep --count option prints the count for each file even when it is
432 zero, as does GNU grep. However, pcregrep was also printing all files when
433 --files-with-matches was added. Now, when both options are given, it prints
434 counts only for those files that have at least one match. (GNU grep just
435 prints the file name in this circumstance, but including the count seems
436 more useful - otherwise, why use --count?) Also ensured that the
437 combination -clh just lists non-zero counts, with no names.
438
439 4. The long form of the pcregrep -F option was incorrectly implemented as
440 --fixed_strings instead of --fixed-strings. This is an incompatible change,
441 but it seems right to fix it, and I didn't think it was worth preserving
442 the old behaviour.
443
444 5. The command line items --regex=pattern and --regexp=pattern were not
445 recognized by pcregrep, which required --regex pattern or --regexp pattern
446 (with a space rather than an '='). The man page documented the '=' forms,
447 which are compatible with GNU grep; these now work.
448
449 6. No libpcreposix.pc file was created for pkg-config; there was just
450 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
451
452 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
453 when UCP support is not needed, by modifying the Python script that
454 generates it from Unicode data files. This should not matter if the module
455 is correctly used as a library, but I received one complaint about 50K of
456 unwanted data. My guess is that the person linked everything into his
457 program rather than using a library. Anyway, it does no harm.
458
459 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
460 was a minimum greater than 1 for a wide character in a possessive
461 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
462 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
463 character. Chaos in the form of incorrect output or a compiling loop could
464 result.
465
466 9. The restrictions on what a pattern can contain when partial matching is
467 requested for pcre_exec() have been removed. All patterns can now be
468 partially matched by this function. In addition, if there are at least two
469 slots in the offset vector, the offset of the earliest inspected character
470 for the match and the offset of the end of the subject are set in them when
471 PCRE_ERROR_PARTIAL is returned.
472
473 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
474 synonymous with PCRE_PARTIAL, for backwards compatibility, and
475 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
476 and may be more useful for multi-segment matching.
477
478 11. Partial matching with pcre_exec() is now more intuitive. A partial match
479 used to be given if ever the end of the subject was reached; now it is
480 given only if matching could not proceed because another character was
481 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
482 string "Z", which now yields "no match" instead of "partial match". In the
483 case of pcre_dfa_exec(), "no match" is given if every matching path for the
484 final character ended with (*FAIL).
485
486 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
487 if the pattern had a "must contain" character that was already found in the
488 earlier partial match, unless partial matching was again requested. For
489 example, with the pattern /dog.(body)?/, the "must contain" character is
490 "g". If the first part-match was for the string "dog", restarting with
491 "sbody" failed. This bug has been fixed.
492
493 13. The string returned by pcre_dfa_exec() after a partial match has been
494 changed so that it starts at the first inspected character rather than the
495 first character of the match. This makes a difference only if the pattern
496 starts with a lookbehind assertion or \b or \B (\K is not supported by
497 pcre_dfa_exec()). It's an incompatible change, but it makes the two
498 matching functions compatible, and I think it's the right thing to do.
499
500 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
501 so that the demonstration program is easily available in environments where
502 PCRE has not been installed from source.
503
504 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
505 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
506 library.
507
508 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
509 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
510 is not the first non-POSIX option to be added. Clearly some people find
511 these options useful.
512
513 17. If a caller to the POSIX matching function regexec() passes a non-zero
514 value for nmatch with a NULL value for pmatch, the value of
515 nmatch is forced to zero.
516
517 18. RunGrepTest did not have a test for the availability of the -u option of
518 the diff command, as RunTest does. It now checks in the same way as
519 RunTest, and also checks for the -b option.
520
521 19. If an odd number of negated classes containing just a single character
522 interposed, within parentheses, between a forward reference to a named
523 subpattern and the definition of the subpattern, compilation crashed with
524 an internal error, complaining that it could not find the referenced
525 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
526 [The bug was that it was starting one character too far in when skipping
527 over the character class, thus treating the ] as data rather than
528 terminating the class. This meant it could skip too much.]
529
530 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
531 /g option in pcretest when the pattern contains \K, which makes it possible
532 to have an empty string match not at the start, even when the pattern is
533 anchored. Updated pcretest and pcredemo to use this option.
534
535 21. If the maximum number of capturing subpatterns in a recursion was greater
536 than the maximum at the outer level, the higher number was returned, but
537 with unset values at the outer level. The correct (outer level) value is
538 now given.
539
540 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
541 PCRE did not set those parentheses (unlike Perl). I have now found a way to
542 make it do so. The string so far is captured, making this feature
543 compatible with Perl.
544
545 23. The tests have been re-organized, adding tests 11 and 12, to make it
546 possible to check the Perl 5.10 features against Perl 5.10.
547
548 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
549 pattern matches a fixed length string. PCRE did not allow this; now it
550 does. Neither allows recursion.
551
552 25. I finally figured out how to implement a request to provide the minimum
553 length of subject string that was needed in order to match a given pattern.
554 (It was back references and recursion that I had previously got hung up
555 on.) This code has now been added to pcre_study(); it finds a lower bound
556 to the length of subject needed. It is not necessarily the greatest lower
557 bound, but using it to avoid searching strings that are too short does give
558 some useful speed-ups. The value is available to calling programs via
559 pcre_fullinfo().
560
561 26. While implementing 25, I discovered to my embarrassment that pcretest had
562 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
563 study optimizations had never been tested with that matching function.
564 Oops. What is worse, even when it was passed study data, there was a bug in
565 pcre_dfa_exec() that meant it never actually used it. Double oops. There
566 were also very few tests of studied patterns with pcre_dfa_exec().
567
568 27. If (?| is used to create subpatterns with duplicate numbers, they are now
569 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
570 on the other side of the coin, they are no longer allowed to have different
571 names, because these cannot be distinguished in PCRE, and this has caused
572 confusion. (This is a difference from Perl.)
573
574 28. When duplicate subpattern names are present (necessarily with different
575 numbers, as required by 27 above), and a test is made by name in a
576 conditional pattern, either for a subpattern having been matched, or for
577 recursion in such a pattern, all the associated numbered subpatterns are
578 tested, and the overall condition is true if the condition is true for any
579 one of them. This is the way Perl works, and is also more like the way
580 testing by number works.
581
582
583 Version 7.9 11-Apr-09
584 ---------------------
585
586 1. When building with support for bzlib/zlib (pcregrep) and/or readline
587 (pcretest), all targets were linked against these libraries. This included
588 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
589 libraries. This caused unwanted dependencies to be created. This problem
590 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
591 pcretest is linked with readline.
592
593 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
594 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
595 moved outside it again, because FALSE and TRUE are already defined in AIX,
596 but BOOL is not.
597
598 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
599 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
600
601 4. The pcregrep documentation said spaces were inserted as well as colons (or
602 hyphens) following file names and line numbers when outputting matching
603 lines. This is not true; no spaces are inserted. I have also clarified the
604 wording for the --colour (or --color) option.
605
606 5. In pcregrep, when --colour was used with -o, the list of matching strings
607 was not coloured; this is different to GNU grep, so I have changed it to be
608 the same.
609
610 6. When --colo(u)r was used in pcregrep, only the first matching substring in
611 each matching line was coloured. Now it goes on to look for further matches
612 of any of the test patterns, which is the same behaviour as GNU grep.
613
614 7. A pattern that could match an empty string could cause pcregrep to loop; it
615 doesn't make sense to accept an empty string match in pcregrep, so I have
616 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
617 seems to be how GNU grep behaves.
618
619 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
620 start or after a newline", because the conditional assertion was not being
621 correctly handled. The rule now is that both the assertion and what follows
622 in the first alternative must satisfy the test.
623
624 9. If auto-callout was enabled in a pattern with a conditional group whose
625 condition was an assertion, PCRE could crash during matching, both with
626 pcre_exec() and pcre_dfa_exec().
627
628 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
629 used for matching.
630
631 11. Unicode property support in character classes was not working for
632 characters (bytes) greater than 127 when not in UTF-8 mode.
633
634 12. Added the -M command line option to pcretest.
635
636 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
637
638 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
639
640 16. Added comments and documentation about mis-use of no_arg in the C++
641 wrapper.
642
643 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
644 from Martin Jerabek that uses macro names for all relevant character and
645 string constants.
646
647 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
648 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
649 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
650 these, but not everybody uses configure.
651
652 19. A conditional group that had only one branch was not being correctly
653 recognized as an item that could match an empty string. This meant that an
654 enclosing group might also not be so recognized, causing infinite looping
655 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
656 with the subject "ab", where knowledge that the repeated group can match
657 nothing is needed in order to break the loop.
658
659 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
660 exec(), but without supplying a callout function, matching went wrong.
661
662 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
663 leak if the size of the offset vector was greater than 30. When the vector
664 is smaller, the saved offsets during recursion go onto a local stack
665 vector, but for larger vectors malloc() is used. It was failing to free
666 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
667 error, in fact).
668
669 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
670 heapframe that is used only when UTF-8 support is enabled. This caused no
671 problem, but was untidy.
672
673 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
674 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
675 included within another project.
676
677 24. Steven Van Ingelgem's patches to add more options to the CMake support,
678 slightly modified by me:
679
680 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
681 not building pcregrep.
682
683 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
684 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
685
686 25. Forward references, both numeric and by name, in patterns that made use of
687 duplicate group numbers, could behave incorrectly or give incorrect errors,
688 because when scanning forward to find the reference group, PCRE was not
689 taking into account the duplicate group numbers. A pattern such as
690 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
691
692 26. Changed a few more instances of "const unsigned char *" to USPTR, making
693 the feature of a custom pointer more persuasive (as requested by a user).
694
695 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
696 pcretest.c, inside #ifndefs, because it seems they are sometimes already
697 pre-defined.
698
699 28. Added support for (*UTF8) at the start of a pattern.
700
701 29. Arrange for flags added by the "release type" setting in CMake to be shown
702 in the configuration summary.
703
704
705 Version 7.8 05-Sep-08
706 ---------------------
707
708 1. Replaced UCP searching code with optimized version as implemented for Ad
709 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
710 stage table and inline lookup instead of a function, giving speed ups of 2
711 to 5 times on some simple patterns that I tested. Permission was given to
712 distribute the MultiStage2.py script that generates the tables (it's not in
713 the tarball, but is in the Subversion repository).
714
715 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
716 scripts.
717
718 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
719 a group with a zero qualifier. The result of the study could be incorrect,
720 or the function might crash, depending on the pattern.
721
722 4. Caseless matching was not working for non-ASCII characters in back
723 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
724 It now works when Unicode Property Support is available.
725
726 5. In pcretest, an escape such as \x{de} in the data was always generating
727 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
728 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
729 truncation.
730
731 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
732
733 7. Added two (int) casts to pcregrep when printing the difference of two
734 pointers, in case they are 64-bit values.
735
736 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
737 test 2 if it fails.
738
739 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
740 and a #define of that name to empty if it is not externally set. This is to
741 allow users of MSVC to set it if necessary.
742
743 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
744 the convenience functions in the pcre_get.c source file.
745
746 11. An option change at the start of a pattern that had top-level alternatives
747 could cause overwriting and/or a crash. This command provoked a crash in
748 some environments:
749
750 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
751
752 This potential security problem was recorded as CVE-2008-2371.
753
754 12. For a pattern where the match had to start at the beginning or immediately
755 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
756 pcre_dfa_exec() could read past the end of the passed subject if there was
757 no match. To help with detecting such bugs (e.g. with valgrind), I modified
758 pcretest so that it places the subject at the end of its malloc-ed buffer.
759
760 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
761 exec() might read past the end of the data buffer in UTF-8 mode.
762
763 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
764 the data contained the byte 0x85 as part of a UTF-8 character within its
765 first line. This applied both to normal and DFA matching.
766
767 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
768 /^[^d]*?$/8 failed to match "abc".
769
770 16. Added a missing copyright notice to pcrecpp_internal.h.
771
772 17. Make it more clear in the documentation that values returned from
773 pcre_exec() in ovector are byte offsets, not character counts.
774
775 18. Tidied a few places to stop certain compilers from issuing warnings.
776
777 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
778 supplied by Stefan Weber. I made a further small update for 7.8 because
779 there is a change of source arrangements: the pcre_searchfuncs.c module is
780 replaced by pcre_ucd.c.
781
782
783 Version 7.7 07-May-08
784 ---------------------
785
786 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
787 a string to a long long, pretend we don't even have a long long." This is
788 done by checking for the strtoq, strtoll, and _strtoi64 functions.
789
790 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
791 pre-7.6 versions, which defined a global no_arg variable instead of putting
792 it in the RE class. (See also #8 below.)
793
794 3. Remove a line of dead code, identified by coverity and reported by Nuno
795 Lopes.
796
797 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
798
799 (1) The include/exclude patterns were being applied to the whole pathnames
800 of files, instead of just to the final components.
801
802 (2) If there was more than one level of directory, the subdirectories were
803 skipped unless they satisfied the include/exclude conditions. This is
804 inconsistent with GNU grep (and could even be seen as contrary to the
805 pcregrep specification - which I improved to make it absolutely clear).
806 The action now is always to scan all levels of directory, and just
807 apply the include/exclude patterns to regular files.
808
809 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
810 --exclude_dir in the tests to avoid scanning .svn directories.
811
812 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
813 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
814 doesn't support NULs in patterns.
815
816 7. Added some missing "const"s to declarations of static tables in
817 pcre_compile.c and pcre_dfa_exec.c.
818
819 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
820 caused by fix #2 above. (Subsequently also a second patch to fix the
821 first patch. And a third patch - this was a messy problem.)
822
823 9. Applied Craig's patch to remove the use of push_back().
824
825 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
826 matching function regexec().
827
828 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
829 which, however, unlike Perl's \g{...}, are subroutine calls, not back
830 references. PCRE supports relative numbers with this syntax (I don't think
831 Oniguruma does).
832
833 12. Previously, a group with a zero repeat such as (...){0} was completely
834 omitted from the compiled regex. However, this means that if the group
835 was called as a subroutine from elsewhere in the pattern, things went wrong
836 (an internal error was given). Such groups are now left in the compiled
837 pattern, with a new opcode that causes them to be skipped at execution
838 time.
839
840 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
841 to the way PCRE behaves:
842
843 (a) A lone ] character is dis-allowed (Perl treats it as data).
844
845 (b) A back reference to an unmatched subpattern matches an empty string
846 (Perl fails the current match path).
847
848 (c) A data ] in a character class must be notated as \] because if the
849 first data character in a class is ], it defines an empty class. (In
850 Perl it is not possible to have an empty class.) The empty class []
851 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
852 The negative empty class [^] matches any one character, independently
853 of the DOTALL setting.
854
855 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
856 non-existent subpattern following a character class starting with ']' and
857 containing () gave an internal compiling error instead of "reference to
858 non-existent subpattern". Fortunately, when the pattern did exist, the
859 compiled code was correct. (When scanning forwards to check for the
860 existencd of the subpattern, it was treating the data ']' as terminating
861 the class, so got the count wrong. When actually compiling, the reference
862 was subsequently set up correctly.)
863
864 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
865 it was being rejected as not supported by pcre_dfa_exec(), even though
866 other assertions are supported. I have made pcre_dfa_exec() support
867 (*FAIL).
868
869 16. The implementation of 13c above involved the invention of a new opcode,
870 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
871 cannot be changed at match time, I realized I could make a small
872 improvement to matching performance by compiling OP_ALLANY instead of
873 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
874 on the OP_ANY path.
875
876 17. Compiling pcretest on Windows with readline support failed without the
877 following two fixes: (1) Make the unistd.h include conditional on
878 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
879
880 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
881 ncurses library to be included for pcretest when ReadLine support is
882 requested, but also to allow for it to be overridden. This patch came from
883 Daniel Bergström.
884
885 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
886 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
887 any errors with the current Unicode tables. Thanks to Peter Kankowski for
888 spotting this.
889
890
891 Version 7.6 28-Jan-08
892 ---------------------
893
894 1. A character class containing a very large number of characters with
895 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
896 overflow.
897
898 2. Patch to cut out the "long long" test in pcrecpp_unittest when
899 HAVE_LONG_LONG is not defined.
900
901 3. Applied Christian Ehrlicher's patch to update the CMake build files to
902 bring them up to date and include new features. This patch includes:
903
904 - Fixed PH's badly added libz and libbz2 support.
905 - Fixed a problem with static linking.
906 - Added pcredemo. [But later removed - see 7 below.]
907 - Fixed dftables problem and added an option.
908 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
909 HAVE_LONG_LONG.
910 - Added readline support for pcretest.
911 - Added an listing of the option settings after cmake has run.
912
913 4. A user submitted a patch to Makefile that makes it easy to create
914 "pcre.dll" under mingw when using Configure/Make. I added stuff to
915 Makefile.am that cause it to include this special target, without
916 affecting anything else. Note that the same mingw target plus all
917 the other distribution libraries and programs are now supported
918 when configuring with CMake (see 6 below) instead of with
919 Configure/Make.
920
921 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
922 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
923 exported in the Windows port". It has not yet been confirmed that the patch
924 solves the problem, but it does no harm.
925
926 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
927 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
928 with CMake, and also correct the comment about stack recursion.
929
930 7. Remove the automatic building of pcredemo from the ./configure system and
931 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
932 of a program that users should build themselves after PCRE is installed, so
933 building it automatically is not really right. What is more, it gave
934 trouble in some build environments.
935
936 8. Further tidies to CMakeLists.txt from Sheri and Christian.
937
938
939 Version 7.5 10-Jan-08
940 ---------------------
941
942 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
943 values in parens when parsing an RE using the C++ wrapper."
944
945 2. Negative specials like \S did not work in character classes in UTF-8 mode.
946 Characters greater than 255 were excluded from the class instead of being
947 included.
948
949 3. The same bug as (2) above applied to negated POSIX classes such as
950 [:^space:].
951
952 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
953 defined or documented. It seems to have been a typo for PCRE_STATIC, so
954 I have changed it.
955
956 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
957 first named subpattern) and a construct such as (?&a) would reference the
958 first named subpattern whose name started with "a" (in other words, the
959 length check was missing). Both these problems are fixed. "Subpattern name
960 expected" is now given for (?&) (a zero-length name), and this patch also
961 makes it give the same error for \k'' (previously it complained that that
962 was a reference to a non-existent subpattern).
963
964 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
965 this is right because (?- can be followed by option settings as well as by
966 digits. I have, however, made the messages clearer.
967
968 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
969 than the number used in the conditional) now cause a compile-time error.
970 This is actually not compatible with Perl, which accepts such patterns, but
971 treats the conditional as always being FALSE (as PCRE used to), but it
972 seems to me that giving a diagnostic is better.
973
974 8. Change "alphameric" to the more common word "alphanumeric" in comments
975 and messages.
976
977 9. Fix two occurrences of "backslash" in comments that should have been
978 "backspace".
979
980 10. Remove two redundant lines of code that can never be obeyed (their function
981 was moved elsewhere).
982
983 11. The program that makes PCRE's Unicode character property table had a bug
984 which caused it to generate incorrect table entries for sequences of
985 characters that have the same character type, but are in different scripts.
986 It amalgamated them into a single range, with the script of the first of
987 them. In other words, some characters were in the wrong script. There were
988 thirteen such cases, affecting characters in the following ranges:
989
990 U+002b0 - U+002c1
991 U+0060c - U+0060d
992 U+0061e - U+00612
993 U+0064b - U+0065e
994 U+0074d - U+0076d
995 U+01800 - U+01805
996 U+01d00 - U+01d77
997 U+01d9b - U+01dbf
998 U+0200b - U+0200f
999 U+030fc - U+030fe
1000 U+03260 - U+0327f
1001 U+0fb46 - U+0fbb1
1002 U+10450 - U+1049d
1003
1004 12. The -o option (show only the matching part of a line) for pcregrep was not
1005 compatible with GNU grep in that, if there was more than one match in a
1006 line, it showed only the first of them. It now behaves in the same way as
1007 GNU grep.
1008
1009 13. If the -o and -v options were combined for pcregrep, it printed a blank
1010 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1011 does the same. The return code can be used to tell if there were any
1012 non-matching lines.
1013
1014 14. Added --file-offsets and --line-offsets to pcregrep.
1015
1016 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1017 infinitely looping recursion. The bug was that positive lookaheads were not
1018 being skipped when checking for a possible empty match (negative lookaheads
1019 and both kinds of lookbehind were skipped).
1020
1021 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1022 inclusion of <windows.h> to before rather than after the definition of
1023 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1024
1025 17. Specifying a possessive quantifier with a specific limit for a Unicode
1026 character property caused pcre_compile() to compile bad code, which led at
1027 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1028 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1029 caused the error; without that there was no problem.
1030
1031 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1032
1033 19. Added --enable-pcretest-libreadline.
1034
1035 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1036 RE::GlobalReplace(). As a result, the number of replacements returned was
1037 double what it should be. I removed one of the increments, but Craig sent a
1038 later patch that removed the other one (the right fix) and added unit tests
1039 that check the return values (which was not done before).
1040
1041 21. Several CMake things:
1042
1043 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1044 the names libpcre and libpcreposix, not just pcre and pcreposix.
1045
1046 (2) The above change means that pcretest and pcregrep are now correctly
1047 linked with the newly-built libraries, not previously installed ones.
1048
1049 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1050
1051 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1052 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1053 UTF-8 newline character). The key issue is that the pattern starts .*;
1054 this means that the match must be either at the beginning, or after a
1055 newline. The bug was in the code for advancing after a failed match and
1056 checking that the new position followed a newline. It was not taking
1057 account of UTF-8 characters correctly.
1058
1059 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1060 character classes. PCRE was not treating the sequence [:...:] as a
1061 character class unless the ... were all letters. Perl, however, seems to
1062 allow any characters between [: and :], though of course it rejects as
1063 unknown any "names" that contain non-letters, because all the known class
1064 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1065 for example, whereas PCRE did not - it did not recognize a POSIX character
1066 class. This seemed a bit dangerous, so the code has been changed to be
1067 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1068 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1069 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1070 Perl does, and where it didn't before.
1071
1072 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1073 Windows environments %n is disabled by default.
1074
1075
1076 Version 7.4 21-Sep-07
1077 ---------------------
1078
1079 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1080 means that a class such as [\s] counted as "explicit reference to CR or
1081 LF". That isn't really right - the whole point of the change was to try to
1082 help when there was an actual mention of one of the two characters. So now
1083 the change happens only if \r or \n (or a literal CR or LF) character is
1084 encountered.
1085
1086 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1087 of both had grown to the point where there were only 3 bits left.
1088 Fortunately, there was spare space in the data structure, and so I have
1089 moved the internal flags into a new 16-bit field to free up more option
1090 bits.
1091
1092 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1093 but did not set the internal JCHANGED flag - either of these is enough to
1094 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1095 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1096 start sets both bits.
1097
1098 4. Added options (at build time, compile time, exec time) to change \R from
1099 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1100
1101 5. doc/pcresyntax.html was missing from the distribution.
1102
1103 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1104 compatibility, even though it is no longer used.
1105
1106 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1107 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1108 windows.h file is present (where different names are used). [This was
1109 reversed later after testing - see 16 below.]
1110
1111 8. Changed all #include <config.h> to #include "config.h". There were also
1112 some further <pcre.h> cases that I changed to "pcre.h".
1113
1114 9. When pcregrep was used with the --colour option, it missed the line ending
1115 sequence off the lines that it output.
1116
1117 10. It was pointed out to me that arrays of string pointers cause lots of
1118 relocations when a shared library is dynamically loaded. A technique of
1119 using a single long string with a table of offsets can drastically reduce
1120 these. I have refactored PCRE in four places to do this. The result is
1121 dramatic:
1122
1123 Originally: 290
1124 After changing UCP table: 187
1125 After changing error message table: 43
1126 After changing table of "verbs" 36
1127 After changing table of Posix names 22
1128
1129 Thanks to the folks working on Gregex for glib for this insight.
1130
1131 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1132 unicode-properties was also set.
1133
1134 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1135
1136 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1137 checked only for CRLF.
1138
1139 14. Added casts to pcretest.c to avoid compiler warnings.
1140
1141 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1142
1143 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1144 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1145 entirely. This removes changes made in 7 above.
1146
1147 17. The CMake files have been updated, and there is now more information about
1148 building with CMake in the NON-UNIX-USE document.
1149
1150
1151 Version 7.3 28-Aug-07
1152 ---------------------
1153
1154 1. In the rejigging of the build system that eventually resulted in 7.1, the
1155 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1156 brackets there is not right, since it causes compilers to look for an
1157 installed pcre.h, not the version that is in the source that is being
1158 compiled (which of course may be different). I have changed it back to:
1159
1160 #include "pcre.h"
1161
1162 I have a vague recollection that the change was concerned with compiling in
1163 different directories, but in the new build system, that is taken care of
1164 by the VPATH setting the Makefile.
1165
1166 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1167 when the subject happened to end in the byte 0x85 (e.g. if the last
1168 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1169 characters but of course it shouldn't be taken as a newline when it is part
1170 of another character. The bug was that, for an unlimited repeat of . in
1171 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1172 characters when looking for a newline.
1173
1174 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1175
1176 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1177 in debug output.
1178
1179 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1180 long printing in the pcrecpp unittest when running under MinGW.
1181
1182 6. ESC_K was left out of the EBCDIC table.
1183
1184 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1185 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1186 limit also applies to "virtual nesting" when a pattern is recursive, and in
1187 this case 1000 isn't so big. I have been able to remove this limit at the
1188 expense of backing off one optimization in certain circumstances. Normally,
1189 when pcre_exec() would call its internal match() function recursively and
1190 immediately return the result unconditionally, it uses a "tail recursion"
1191 feature to save stack. However, when a subpattern that can match an empty
1192 string has an unlimited repetition quantifier, it no longer makes this
1193 optimization. That gives it a stack frame in which to save the data for
1194 checking that an empty string has been matched. Previously this was taken
1195 from the 1000-entry workspace that had been reserved. So now there is no
1196 explicit limit, but more stack is used.
1197
1198 8. Applied Daniel's patches to solve problems with the import/export magic
1199 syntax that is required for Windows, and which was going wrong for the
1200 pcreposix and pcrecpp parts of the library. These were overlooked when this
1201 problem was solved for the main library.
1202
1203 9. There were some crude static tests to avoid integer overflow when computing
1204 the size of patterns that contain repeated groups with explicit upper
1205 limits. As the maximum quantifier is 65535, the maximum group length was
1206 set at 30,000 so that the product of these two numbers did not overflow a
1207 32-bit integer. However, it turns out that people want to use groups that
1208 are longer than 30,000 bytes (though not repeat them that many times).
1209 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1210 made it possible to implement the integer overflow checks in a much more
1211 dynamic way, which I have now done. The artificial limitation on group
1212 length has been removed - we now have only the limit on the total length of
1213 the compiled pattern, which depends on the LINK_SIZE setting.
1214
1215 10. Fixed a bug in the documentation for get/copy named substring when
1216 duplicate names are permitted. If none of the named substrings are set, the
1217 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1218 empty string.
1219
1220 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1221 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1222 because the ] is interpreted as the first data character and the
1223 terminating ] is not found. PCRE has been made compatible with Perl in this
1224 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1225 cause memory overwriting.
1226
1227 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1228 string has been matched (to stop an infinite loop). It was not recognizing
1229 a conditional subpattern that could match an empty string if that
1230 subpattern was within another subpattern. For example, it looped when
1231 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1232 condition was not nested. This bug has been fixed.
1233
1234 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1235 past the start of the subject in the presence of bytes with the top bit
1236 set, for example "\x8aBCD".
1237
1238 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1239 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1240
1241 14. Optimized (?!) to (*FAIL).
1242
1243 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1244 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1245 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1246 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1247 does: it's just the validity check that is more restrictive.
1248
1249 16. Inserted checks for integer overflows during escape sequence (backslash)
1250 processing, and also fixed erroneous offset values for syntax errors during
1251 backslash processing.
1252
1253 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1254 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1255
1256 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1257 caused an overrun.
1258
1259 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1260 something other than just ASCII characters) inside a group that had an
1261 unlimited repeat caused a loop at compile time (while checking to see
1262 whether the group could match an empty string).
1263
1264 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1265 [\P{Any}] did so. (Error in the code for printing property names.)
1266
1267 21. An orphan \E inside a character class could cause a crash.
1268
1269 22. A repeated capturing bracket such as (A)? could cause a wild memory
1270 reference during compilation.
1271
1272 23. There are several functions in pcre_compile() that scan along a compiled
1273 expression for various reasons (e.g. to see if it's fixed length for look
1274 behind). There were bugs in these functions when a repeated \p or \P was
1275 present in the pattern. These operators have additional parameters compared
1276 with \d, etc, and these were not being taken into account when moving along
1277 the compiled data. Specifically:
1278
1279 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1280 length.
1281
1282 (b) An item such as \pL+ within a repeated group could cause crashes or
1283 loops.
1284
1285 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1286 "reference to non-existent subpattern" error.
1287
1288 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1289
1290 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1291 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1292
1293 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1294
1295 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1296 character were causing crashes (broken optimization).
1297
1298 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1299 \p or \P) caused a compile-time loop.
1300
1301 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1302 break. For example, the unstudied pattern [\r\n]A does not match the string
1303 "\r\nA" because change 7.0/46 below moves the current point on by two
1304 characters after failing to match at the start. However, the pattern \nA
1305 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1306 the same is true. There doesn't seem any very clean way out of this, but
1307 what I have chosen to do makes the common cases work: PCRE now takes note
1308 of whether there can be an explicit match for \r or \n anywhere in the
1309 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1310 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1311 pattern has explicit CR or LF references.
1312
1313 29. Added (*CR) etc for changing newline setting at start of pattern.
1314
1315
1316 Version 7.2 19-Jun-07
1317 ---------------------
1318
1319 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1320 which is apparently normally available under Windows.
1321
1322 2. Re-jig the pcregrep tests with different newline settings in an attempt
1323 to make them independent of the local environment's newline setting.
1324
1325 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1326
1327 4. Some of the "internals" tests were previously cut out when the link size
1328 was not 2, because the output contained actual offsets. The recent new
1329 "Z" feature of pcretest means that these can be cut out, making the tests
1330 usable with all link sizes.
1331
1332 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1333 stack recursion. This gives a massive performance boost under BSD, but just
1334 a small improvement under Linux. However, it saves one field in the frame
1335 in all cases.
1336
1337 6. Added more features from the forthcoming Perl 5.10:
1338
1339 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1340 recursion call. It refers to the nth most recently opened parentheses.
1341
1342 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1343 to be opened parentheses.
1344
1345 (c) Conditions that refer to capturing parentheses can be specified
1346 relatively, for example, (?(-2)... or (?(+3)...
1347
1348 (d) \K resets the start of the current match so that everything before
1349 is not part of it.
1350
1351 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1352
1353 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1354 reference syntax.
1355
1356 (g) (?| introduces a group in which the numbering of parentheses in each
1357 alternative starts with the same number.
1358
1359 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1360
1361 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1362 PCRE_INFO_JCHANGED.
1363
1364 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1365 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1366 for detecting groups that can match an empty string.
1367
1368 9. A pattern with a very large number of alternatives (more than several
1369 hundred) was running out of internal workspace during the pre-compile
1370 phase, where pcre_compile() figures out how much memory will be needed. A
1371 bit of new cunning has reduced the workspace needed for groups with
1372 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1373 workspace instead of running out of the 4096 that are available.
1374
1375 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1376
1377 11. Applied patch from Google to remove an optimization that didn't quite work.
1378 The report of the bug said:
1379
1380 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1381 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1382 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1383
1384 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1385 it matched the wrong number of bytes.
1386
1387
1388 Version 7.1 24-Apr-07
1389 ---------------------
1390
1391 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1392 that is more "standard", making use of automake and other Autotools. There
1393 is some re-arrangement of the files and adjustment of comments consequent
1394 on this.
1395
1396 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1397 for recursive directory scanning broke on some systems because the files
1398 are not scanned in any specific order and on different systems the order
1399 was different. A call to "sort" has been inserted into RunGrepTest for the
1400 approprate test as a short-term fix. In the longer term there may be an
1401 alternative.
1402
1403 3. I had an email from Eric Raymond about problems translating some of PCRE's
1404 man pages to HTML (despite the fact that I distribute HTML pages, some
1405 people do their own conversions for various reasons). The problems
1406 concerned the use of low-level troff macros .br and .in. I have therefore
1407 removed all such uses from the man pages (some were redundant, some could
1408 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1409 HTML has been updated to handle .nf/.fi and to complain if it encounters
1410 .br or .in.
1411
1412 4. Updated comments in configure.ac that get placed in config.h.in and also
1413 arranged for config.h to be included in the distribution, with the name
1414 config.h.generic, for the benefit of those who have to compile without
1415 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1416
1417 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1418 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1419 makevp.bat for the current PCRE, using the additional files
1420 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1421
1422 6. A Windows user reported a minor discrepancy with test 2, which turned out
1423 to be caused by a trailing space on an input line that had got lost in his
1424 copy. The trailing space was an accident, so I've just removed it.
1425
1426 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1427 that is needed.
1428
1429 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1430 as "const" (a) because they are and (b) because it helps the PHP
1431 maintainers who have recently made a script to detect big data structures
1432 in the php code that should be moved to the .rodata section. I remembered
1433 to update Builducptable as well, so it won't revert if ucptable.h is ever
1434 re-created.
1435
1436 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1437 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1438 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1439 support is not required. This saves 1.5-2K of code, which is important in
1440 some applications.
1441
1442 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1443 so as not to refer to the tables, even though these functions will never be
1444 called when UTF-8 support is disabled. Otherwise there are problems with a
1445 shared library.
1446
1447 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1448
1449 (a) It was defining its arguments as char * instead of void *.
1450
1451 (b) It was assuming that all moves were upwards in memory; this was true
1452 a long time ago when I wrote it, but is no longer the case.
1453
1454 The emulated memove() is provided for those environments that have neither
1455 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1456 is clearly not the case, as these two bugs were recently reported.
1457
1458 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1459 and Detrail to create the HTML documentation, the .txt form of the man
1460 pages, and it removes trailing spaces from listed files. It also creates
1461 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1462 case, it wraps all the #defines with #ifndefs. This script should be run
1463 before "make dist".
1464
1465 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1466 with Unicode property support.
1467
1468 (a) For a maximizing quantifier, if the two different cases of the
1469 character were of different lengths in their UTF-8 codings (there are
1470 some cases like this - I found 11), and the matching function had to
1471 back up over a mixture of the two cases, it incorrectly assumed they
1472 were both the same length.
1473
1474 (b) When PCRE was configured to use the heap rather than the stack for
1475 recursion during matching, it was not correctly preserving the data for
1476 the other case of a UTF-8 character when checking ahead for a match
1477 while processing a minimizing repeat. If the check also involved
1478 matching a wide character, but failed, corruption could cause an
1479 erroneous result when trying to check for a repeat of the original
1480 character.
1481
1482 13. Some tidying changes to the testing mechanism:
1483
1484 (a) The RunTest script now detects the internal link size and whether there
1485 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1486 values substituted by "configure". (The RunGrepTest script already did
1487 this for UTF-8.) The configure.ac script no longer substitutes the
1488 relevant variables.
1489
1490 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1491 with length and offset values. This means that the output is different
1492 for different internal link sizes. Test 2 is skipped for link sizes
1493 other than 2 because of this, bypassing the problem. Unfortunately,
1494 there was also a test in test 3 (the locale tests) that used /B and
1495 failed for link sizes other than 2. Rather than cut the whole test out,
1496 I have added a new /Z option to pcretest that replaces the length and
1497 offset values with spaces. This is now used to make test 3 independent
1498 of link size. (Test 2 will be tidied up later.)
1499
1500 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1501 segmentation fault instead of returning the appropriate error message.
1502
1503 15. In multiline mode when the newline sequence was set to "any", the pattern
1504 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1505 This doesn't seem right; it now treats the CRLF combination as the line
1506 ending, and so does not match in that case. It's only a pattern such as ^$
1507 that would hit this one: something like ^ABC$ would have failed after \r
1508 and then tried again after \r\n.
1509
1510 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1511 in an attempt to make files that differ only in their line terminators
1512 compare equal. This works on Linux.
1513
1514 17. Under certain error circumstances pcregrep might try to free random memory
1515 as it exited. This is now fixed, thanks to valgrind.
1516
1517 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1518 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1519 was because its rules for how to advance for /g after matching an empty
1520 string at the end of a line did not allow for this case. They now check for
1521 it specially.
1522
1523 20. pcretest is supposed to handle patterns and data of any length, by
1524 extending its buffers when necessary. It was getting this wrong when the
1525 buffer for a data line had to be extended.
1526
1527 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1528 CRLF as a newline sequence.
1529
1530 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1531 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1532 I have nevertheless tidied it up.
1533
1534 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1535
1536 24. Added a man page for pcre-config.
1537
1538
1539 Version 7.0 19-Dec-06
1540 ---------------------
1541
1542 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1543 moving to gcc 4.1.1.
1544
1545 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1546 sys/time.h, which is documented as needed for this function. It doesn't
1547 seem to matter on Linux, but it showed up on some releases of OS X.
1548
1549 3. It seems that there are systems where bytes whose values are greater than
1550 127 match isprint() in the "C" locale. The "C" locale should be the
1551 default when a C program starts up. In most systems, only ASCII printing
1552 characters match isprint(). This difference caused the output from pcretest
1553 to vary, making some of the tests fail. I have changed pcretest so that:
1554
1555 (a) When it is outputting text in the compiled version of a pattern, bytes
1556 other than 32-126 are always shown as hex escapes.
1557
1558 (b) When it is outputting text that is a matched part of a subject string,
1559 it does the same, unless a different locale has been set for the match
1560 (using the /L modifier). In this case, it uses isprint() to decide.
1561
1562 4. Fixed a major bug that caused incorrect computation of the amount of memory
1563 required for a compiled pattern when options that changed within the
1564 pattern affected the logic of the preliminary scan that determines the
1565 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1566 that the computed length was too small. The symptoms of this bug were
1567 either the PCRE error "internal error: code overflow" from pcre_compile(),
1568 or a glibc crash with a message such as "pcretest: free(): invalid next
1569 size (fast)". Examples of patterns that provoked this bug (shown in
1570 pcretest format) are:
1571
1572 /(?-x: )/x
1573 /(?x)(?-x: \s*#\s*)/
1574 /((?i)[\x{c0}])/8
1575 /(?i:[\x{c0}])/8
1576
1577 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1578 is now done differently.
1579
1580 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1581 wrapper classes; (b) implement a new function in the C++ scanner that is
1582 more efficient than the old way of doing things because it avoids levels of
1583 recursion in the regex matching; (c) add a paragraph to the documentation
1584 for the FullMatch() function.
1585
1586 6. The escape sequence \n was being treated as whatever was defined as
1587 "newline". Not only was this contrary to the documentation, which states
1588 that \n is character 10 (hex 0A), but it also went horribly wrong when
1589 "newline" was defined as CRLF. This has been fixed.
1590
1591 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1592 was being set to -1 for the "end of line" case (supposedly a value that no
1593 character can have). Though this value is never used (the check for end of
1594 line is "zero bytes in current character"), it caused compiler complaints.
1595 I've changed it to 0xffffffff.
1596
1597 8. In pcre_version.c, the version string was being built by a sequence of
1598 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1599 string (as it is for production releases) called a macro with an empty
1600 argument. The C standard says the result of this is undefined. The gcc
1601 compiler treats it as an empty string (which was what was wanted) but it is
1602 reported that Visual C gives an error. The source has been hacked around to
1603 avoid this problem.
1604
1605 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1606 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1607 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1608 of them did).
1609
1610 10. Originally, pcretest opened its input and output without "b"; then I was
1611 told that "b" was needed in some environments, so it was added for release
1612 5.0 to both the input and output. (It makes no difference on Unix-like
1613 systems.) Later I was told that it is wrong for the input on Windows. I've
1614 now abstracted the modes into two macros, to make it easier to fiddle with
1615 them, and removed "b" from the input mode under Windows.
1616
1617 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1618
1619 12. Added -help and --help to pcretest as an official way of being reminded
1620 of the options.
1621
1622 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1623 and pcrecpp.cc because they annoy compilers at high warning levels.
1624
1625 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1626
1627 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1628 scripts are not C programs :-) and which was not noticed because it works
1629 on Linux.
1630
1631 16. pcretest is supposed to handle any length of pattern and data line (as one
1632 line or as a continued sequence of lines) by extending its input buffer if
1633 necessary. This feature was broken for very long pattern lines, leading to
1634 a string of junk being passed to pcre_compile() if the pattern was longer
1635 than about 50K.
1636
1637 17. I have done a major re-factoring of the way pcre_compile() computes the
1638 amount of memory needed for a compiled pattern. Previously, there was code
1639 that made a preliminary scan of the pattern in order to do this. That was
1640 OK when PCRE was new, but as the facilities have expanded, it has become
1641 harder and harder to keep it in step with the real compile phase, and there
1642 have been a number of bugs (see for example, 4 above). I have now found a
1643 cunning way of running the real compile function in a "fake" mode that
1644 enables it to compute how much memory it would need, while actually only
1645 ever using a few hundred bytes of working memory and without too many
1646 tests of the mode. This should make future maintenance and development
1647 easier. A side effect of this work is that the limit of 200 on the nesting
1648 depth of parentheses has been removed (though this was never a serious
1649 limitation, I suspect). However, there is a downside: pcre_compile() now
1650 runs more slowly than before (30% or more, depending on the pattern). I
1651 hope this isn't a big issue. There is no effect on runtime performance.
1652
1653 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1654 newline (only possible for the last line of a file) and it was a
1655 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1656
1657 19. Added additional timing features to pcretest. (1) The -tm option now times
1658 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1659 separate command line item, by a number that specifies the number of
1660 repeats to use when timing. The default is 50000; this gives better
1661 precision, but takes uncomfortably long for very large patterns.
1662
1663 20. Extended pcre_study() to be more clever in cases where a branch of a
1664 subpattern has no definite first character. For example, (a*|b*)[cd] would
1665 previously give no result from pcre_study(). Now it recognizes that the
1666 first character must be a, b, c, or d.
1667
1668 21. There was an incorrect error "recursive call could loop indefinitely" if
1669 a subpattern (or the entire pattern) that was being tested for matching an
1670 empty string contained only one non-empty item after a nested subpattern.
1671 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1672 incorrectly, because the \d was being skipped in the check.
1673
1674 22. The pcretest program now has a new pattern option /B and a command line
1675 option -b, which is equivalent to adding /B to every pattern. This causes
1676 it to show the compiled bytecode, without the additional information that
1677 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1678 is the same as /B/I).
1679
1680 23. A new optimization is now able automatically to treat some sequences such
1681 as a*b as a*+b. More specifically, if something simple (such as a character
1682 or a simple class like \d) has an unlimited quantifier, and is followed by
1683 something that cannot possibly match the quantified thing, the quantifier
1684 is automatically "possessified".
1685
1686 24. A recursive reference to a subpattern whose number was greater than 39
1687 went wrong under certain circumstances in UTF-8 mode. This bug could also
1688 have affected the operation of pcre_study().
1689
1690 25. Realized that a little bit of performance could be had by replacing
1691 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1692
1693 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1694
1695 27. Possessive quantifiers such as a++ were previously implemented by turning
1696 them into atomic groups such as ($>a+). Now they have their own opcodes,
1697 which improves performance. This includes the automatically created ones
1698 from 23 above.
1699
1700 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1701 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1702 the first matched character to be a colon. This applied both to named and
1703 numbered groups.
1704
1705 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1706
1707 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1708 building PCRE on Windows easier, so I have included it in the distribution.
1709
1710 31. There is now a check in pcretest against a ridiculously large number being
1711 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1712 loop, the loop is abandoned.
1713
1714 32. Forward references to subpatterns in conditions such as (?(2)...) where
1715 subpattern 2 is defined later cause pcre_compile() to search forwards in
1716 the pattern for the relevant set of parentheses. This search went wrong
1717 when there were unescaped parentheses in a character class, parentheses
1718 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1719
1720 33. "Subroutine" calls and backreferences were previously restricted to
1721 referencing subpatterns earlier in the regex. This restriction has now
1722 been removed.
1723
1724 34. Added a number of extra features that are going to be in Perl 5.10. On the
1725 whole, these are just syntactic alternatives for features that PCRE had
1726 previously implemented using the Python syntax or my own invention. The
1727 other formats are all retained for compatibility.
1728
1729 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1730 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1731 also .NET compatible.
1732
1733 (b) A recursion or subroutine call to a named group can now be defined as
1734 (?&name) as well as (?P>name).
1735
1736 (c) A backreference to a named group can now be defined as \k<name> or
1737 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1738 5.10, are also .NET compatible.
1739
1740 (d) A conditional reference to a named group can now use the syntax
1741 (?(<name>) or (?('name') as well as (?(name).
1742
1743 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1744 groups (named and numbered) that are never evaluated inline, but can be
1745 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1746 is always false. There may be only one alternative in such a group.
1747
1748 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1749 as the simple (?(R). The condition is true only if the most recent
1750 recursion is that of the given number or name. It does not search out
1751 through the entire recursion stack.
1752
1753 (g) The escape \gN or \g{N} has been added, where N is a positive or
1754 negative number, specifying an absolute or relative reference.
1755
1756 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1757 some "unreachable code" warnings.
1758
1759 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1760 things, this adds five new scripts.
1761
1762 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1763 There were also incompatibilities regarding the handling of \Q..\E inside
1764 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1765 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1766
1767 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1768 matches an empty string, and forcibly breaks the loop. There were bugs in
1769 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1770 against aaaa the result was just "a" rather than "aaaa", for example. Two
1771 separate and independent bugs (that affected different cases) have been
1772 fixed.
1773
1774 39. Refactored the code to abolish the use of different opcodes for small
1775 capturing bracket numbers. This is a tidy that I avoided doing when I
1776 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1777 The new approach is not only tidier, it makes it possible to reduce the
1778 memory needed to fix the previous bug (38).
1779
1780 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1781 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1782 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1783 mode.
1784
1785 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1786 report.
1787
1788 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1789 copy construction and assignment in the C++ wrapper.
1790
1791 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1792 couple of bugs that could have given wrong results in the "--newline=crlf"
1793 case.
1794
1795 44. Added a number of casts and did some reorganization of signed/unsigned int
1796 variables following suggestions from Dair Grant. Also renamed the variable
1797 "this" as "item" because it is a C++ keyword.
1798
1799 45. Arranged for dftables to add
1800
1801 #include "pcre_internal.h"
1802
1803 to pcre_chartables.c because without it, gcc 4.x may remove the array
1804 definition from the final binary if PCRE is built into a static library and
1805 dead code stripping is activated.
1806
1807 46. For an unanchored pattern, if a match attempt fails at the start of a
1808 newline sequence, and the newline setting is CRLF or ANY, and the next two
1809 characters are CRLF, advance by two characters instead of one.
1810
1811
1812 Version 6.7 04-Jul-06
1813 ---------------------
1814
1815 1. In order to handle tests when input lines are enormously long, pcretest has
1816 been re-factored so that it automatically extends its buffers when
1817 necessary. The code is crude, but this _is_ just a test program. The
1818 default size has been increased from 32K to 50K.
1819
1820 2. The code in pcre_study() was using the value of the re argument before
1821 testing it for NULL. (Of course, in any sensible call of the function, it
1822 won't be NULL.)
1823
1824 3. The memmove() emulation function in pcre_internal.h, which is used on
1825 systems that lack both memmove() and bcopy() - that is, hardly ever -
1826 was missing a "static" storage class specifier.
1827
1828 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1829 containing an extended class (one that cannot be represented by a bitmap
1830 because it contains high-valued characters or Unicode property items, e.g.
1831 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1832 pattern, but PCRE should not loop if you do not (it no longer does).
1833 [Detail: two cases were found: (a) a repeated subpattern containing an
1834 extended class; (b) a recursive reference to a subpattern that followed a
1835 previous extended class. It wasn't skipping over the extended class
1836 correctly when UTF-8 mode was not set.]
1837
1838 5. A negated single-character class was not being recognized as fixed-length
1839 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1840 compile error "lookbehind assertion is not fixed length".
1841
1842 6. The RunPerlTest auxiliary script was showing an unexpected difference
1843 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1844 write a Perl script that can interpret lines of an input file either as
1845 byte characters or as UTF-8, which is what "perltest" was being required to
1846 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1847 can't do is switch easily at run time between having the "use utf8;" pragma
1848 or not. In the end, I fudged it by using the RunPerlTest script to insert
1849 "use utf8;" explicitly for the UTF-8 tests.
1850
1851 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1852 the end of the subject string, contrary to the documentation and to what
1853 Perl does. This was true of both matching functions. Now it matches only at
1854 the start of the subject and immediately after *internal* newlines.
1855
1856 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1857 a pointer to an int instead of a pointer to an unsigned long int. This
1858 caused problems on 64-bit systems.
1859
1860 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1861 instance of the 'standard' template library not being so standard".
1862
1863 10. There was no check on the number of named subpatterns nor the maximum
1864 length of a subpattern name. The product of these values is used to compute
1865 the size of the memory block for a compiled pattern. By supplying a very
1866 long subpattern name and a large number of named subpatterns, the size
1867 computation could be caused to overflow. This is now prevented by limiting
1868 the length of names to 32 characters, and the number of named subpatterns
1869 to 10,000.
1870
1871 11. Subpatterns that are repeated with specific counts have to be replicated in
1872 the compiled pattern. The size of memory for this was computed from the
1873 length of the subpattern and the repeat count. The latter is limited to
1874 65535, but there was no limit on the former, meaning that integer overflow
1875 could in principle occur. The compiled length of a repeated subpattern is
1876 now limited to 30,000 bytes in order to prevent this.
1877
1878 12. Added the optional facility to have named substrings with the same name.
1879
1880 13. Added the ability to use a named substring as a condition, using the
1881 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1882 are numbers (not recommended). Forward references are permitted.
1883
1884 14. Added forward references in named backreferences (if you see what I mean).
1885
1886 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1887 pattern could run off the end of the subject. For example, the pattern
1888 "(?s)(.{1,5})"8 did this with the subject "ab".
1889
1890 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1891 PCRE_CASELESS was set when matching characters that were quantified with ?
1892 or *.
1893
1894 17. A character class other than a single negated character that had a minimum
1895 but no maximum quantifier - for example [ab]{6,} - was not handled
1896 correctly by pce_dfa_exec(). It would match only one character.
1897
1898 18. A valid (though odd) pattern that looked like a POSIX character
1899 class but used an invalid character after [ (for example [[,abc,]]) caused
1900 pcre_compile() to give the error "Failed: internal error: code overflow" or
1901 in some cases to crash with a glibc free() error. This could even happen if
1902 the pattern terminated after [[ but there just happened to be a sequence of
1903 letters, a binary zero, and a closing ] in the memory that followed.
1904
1905 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1906 over the years. Originally (before any Unicode support), just the bottom 8
1907 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1908 output from "man perlunicode" includes this:
1909
1910 The regular expression compiler produces polymorphic opcodes. That
1911 is, the pattern adapts to the data and automatically switches to
1912 the Unicode character scheme when presented with Unicode data--or
1913 instead uses a traditional byte scheme when presented with byte
1914 data.
1915
1916 Sadly, a wide octal escape does not cause a switch, and in a string with
1917 no other multibyte characters, these octal escapes are treated as before.
1918 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1919 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1920 Unicode string.
1921
1922 I have not perpetrated such confusion in PCRE. Up till now, it took just
1923 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1924 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1925 translate to the appropriate multibyte character.
1926
1927 29. Applied some refactoring to reduce the number of warnings from Microsoft
1928 and Borland compilers. This has included removing the fudge introduced
1929 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1930 a warning about an unused variable.
1931
1932 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1933 characters since release 4.0, because Perl (from release 5.004) does not.
1934 [Or at least, is documented not to: some releases seem to be in conflict
1935 with the documentation.] However, when a pattern was studied with
1936 pcre_study() and all its branches started with \s, PCRE still included VT
1937 as a possible starting character. Of course, this did no harm; it just
1938 caused an unnecessary match attempt.
1939
1940 22. Removed a now-redundant internal flag bit that recorded the fact that case
1941 dependency changed within the pattern. This was once needed for "required
1942 byte" processing, but is no longer used. This recovers a now-scarce options
1943 bit. Also moved the least significant internal flag bit to the most-
1944 significant bit of the word, which was not previously used (hangover from
1945 the days when it was an int rather than a uint) to free up another bit for
1946 the future.
1947
1948 23. Added support for CRLF line endings as well as CR and LF. As well as the
1949 default being selectable at build time, it can now be changed at runtime
1950 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1951 specify that it is scanning data with non-default line endings.
1952
1953 24. Changed the definition of CXXLINK to make it agree with the definition of
1954 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1955
1956 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1957 recursions. This makes a big different to stack usage for some patterns.
1958
1959 26. If a subpattern containing a named recursion or subroutine reference such
1960 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1961 the space required for the compiled pattern went wrong and gave too small a
1962 value. Depending on the environment, this could lead to "Failed: internal
1963 error: code overflow at offset 49" or "glibc detected double free or
1964 corruption" errors.
1965
1966 27. Applied patches from Google (a) to support the new newline modes and (b) to
1967 advance over multibyte UTF-8 characters in GlobalReplace.
1968
1969 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1970 difference for some implementation of PCRE in some Windows version.
1971
1972 29. Added some extra testing facilities to pcretest:
1973
1974 \q<number> in a data line sets the "match limit" value
1975 \Q<number> in a data line sets the "match recursion limt" value
1976 -S <number> sets the stack size, where <number> is in megabytes
1977
1978 The -S option isn't available for Windows.
1979
1980
1981 Version 6.6 06-Feb-06
1982 ---------------------
1983
1984 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1985 in pcreposix.h. I have copied the definition from pcre.h.
1986
1987 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1988 because pcre.h is no longer a built file.
1989
1990 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1991 not normally included in the compiled code.
1992
1993
1994 Version 6.5 01-Feb-06
1995 ---------------------
1996
1997 1. When using the partial match feature with pcre_dfa_exec(), it was not
1998 anchoring the second and subsequent partial matches at the new starting
1999 point. This could lead to incorrect results. For example, with the pattern
2000 /1234/, partially matching against "123" and then "a4" gave a match.
2001
2002 2. Changes to pcregrep:
2003
2004 (a) All non-match returns from pcre_exec() were being treated as failures
2005 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2006 error message is output. Some extra information is given for the
2007 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2008 probably the only errors that are likely to be caused by users (by
2009 specifying a regex that has nested indefinite repeats, for instance).
2010 If there are more than 20 of these errors, pcregrep is abandoned.
2011
2012 (b) A binary zero was treated as data while matching, but terminated the
2013 output line if it was written out. This has been fixed: binary zeroes
2014 are now no different to any other data bytes.
2015
2016 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2017 used to set a locale for matching. The --locale=xxxx long option has
2018 been added (no short equivalent) to specify a locale explicitly on the
2019 pcregrep command, overriding the environment variables.
2020
2021 (d) When -B was used with -n, some line numbers in the output were one less
2022 than they should have been.
2023
2024 (e) Added the -o (--only-matching) option.
2025
2026 (f) If -A or -C was used with -c (count only), some lines of context were
2027 accidentally printed for the final match.
2028
2029 (g) Added the -H (--with-filename) option.
2030
2031 (h) The combination of options -rh failed to suppress file names for files
2032 that were found from directory arguments.
2033
2034 (i) Added the -D (--devices) and -d (--directories) options.
2035
2036 (j) Added the -F (--fixed-strings) option.
2037
2038 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2039
2040 (l) Added the --colo(u)r option.
2041
2042 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2043 is not present by default.
2044
2045 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2046 items such as (?R) or (?1), when the recursion could match a number of
2047 alternatives. If it matched one of the alternatives, but subsequently,
2048 outside the recursion, there was a failure, the code tried to back up into
2049 the recursion. However, because of the way PCRE is implemented, this is not
2050 possible, and the result was an incorrect result from the match.
2051
2052 In order to prevent this happening, the specification of recursion has
2053 been changed so that all such subpatterns are automatically treated as
2054 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2055
2056 4. I had overlooked the fact that, in some locales, there are characters for
2057 which isalpha() is true but neither isupper() nor islower() are true. In
2058 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2059 and ordfeminine) are like this. This affected the treatment of \w and \W
2060 when they appeared in character classes, but not when they appeared outside
2061 a character class. The bit map for "word" characters is now created
2062 separately from the results of isalnum() instead of just taking it from the
2063 upper, lower, and digit maps. (Plus the underscore character, of course.)
2064
2065 5. The above bug also affected the handling of POSIX character classes such as
2066 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2067 permanent tables. Instead, the bit maps for such a class were previously
2068 created as the appropriate unions of the upper, lower, and digit bitmaps.
2069 Now they are created by subtraction from the [[:word:]] class, which has
2070 its own bitmap.
2071
2072 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2073 It is created by subtracting the vertical space characters (\x09, \x0a,
2074 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2075 subtraction was done in the overall bitmap for a character class, meaning
2076 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2077 be recognized. This bug has been fixed.
2078
2079 7. Patches from the folks at Google:
2080
2081 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2082 real life, but is still worth protecting against".
2083
2084 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2085 regular expressions".
2086
2087 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2088 have it.
2089
2090 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2091 "configure" and the latter not, in order to fix a problem somebody had
2092 with compiling the Arg class on HP-UX.
2093
2094 (e) Improve the error-handling of the C++ wrapper a little bit.
2095
2096 (f) New tests for checking recursion limiting.
2097
2098 8. The pcre_memmove() function, which is used only if the environment does not
2099 have a standard memmove() function (and is therefore rarely compiled),
2100 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2101 returning a result (though PCRE never actually uses the result).
2102
2103 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2104 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2105 returned instead of calling malloc() with an overflowing number that would
2106 most likely cause subsequent chaos.
2107
2108 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2109
2110 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2111 with this option is matched, the nmatch and pmatch options of regexec() are
2112 ignored.
2113
2114 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2115 provided in case anyone wants to the the POSIX interface with UTF-8
2116 strings.
2117
2118 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2119 C++ linking (needed for some HP-UX environments).
2120
2121 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2122 (unused parameter) and in the pcre_printint() function (omitted "default"
2123 switch label when the default is to do nothing).
2124
2125 15. Added some code to make it possible, when PCRE is compiled as a C++
2126 library, to replace subject pointers for pcre_exec() with a smart pointer
2127 class, thus making it possible to process discontinuous strings.
2128
2129 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2130 much the same function. They were added by different people who were trying
2131 to make PCRE easy to compile on non-Unix systems. It has been suggested
2132 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2133 for compiling on Windows systems. I have therefore replaced it with
2134 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2135 defaults to "extern" for C or "extern C" for C++, which works fine on
2136 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2137 SCOPE with something explicit in config.h. In addition:
2138
2139 (a) pcreposix.h still had just "extern" instead of either of these macros;
2140 I have replaced it with PCRE_DATA_SCOPE.
2141
2142 (b) Functions such as _pcre_xclass(), which are internal to the library,
2143 but external in the C sense, all had PCRE_EXPORT in their definitions.
2144 This is apparently wrong for the Windows case, so I have removed it.
2145 (It makes no difference on Unix-like systems.)
2146
2147 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2148 of recursive calls to match(). This is different to MATCH_LIMIT because
2149 that limits the total number of calls to match(), not all of which increase
2150 the depth of recursion. Limiting the recursion depth limits the amount of
2151 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2152 when PCRE is compiled, and changed at run time. A patch from Google adds
2153 this functionality to the C++ interface.
2154
2155 18. Changes to the handling of Unicode character properties:
2156
2157 (a) Updated the table to Unicode 4.1.0.
2158
2159 (b) Recognize characters that are not in the table as "Cn" (undefined).
2160
2161 (c) I revised the way the table is implemented to a much improved format
2162 which includes recognition of ranges. It now supports the ranges that
2163 are defined in UnicodeData.txt, and it also amalgamates other
2164 characters into ranges. This has reduced the number of entries in the
2165 table from around 16,000 to around 3,000, thus reducing its size
2166 considerably. I realized I did not need to use a tree structure after
2167 all - a binary chop search is just as efficient. Having reduced the
2168 number of entries, I extended their size from 6 bytes to 8 bytes to
2169 allow for more data.
2170
2171 (d) Added support for Unicode script names via properties such as \p{Han}.
2172
2173 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2174 matching that character.
2175
2176 20. When matching a repeated Unicode property with a minimum greater than zero,
2177 (for example \pL{2,}), PCRE could look past the end of the subject if it
2178 reached it while seeking the minimum number of characters. This could
2179 happen only if some of the characters were more than one byte long, because
2180 there is a check for at least the minimum number of bytes.
2181
2182 21. Refactored the implementation of \p and \P so as to be more general, to
2183 allow for more different types of property in future. This has changed the
2184 compiled form incompatibly. Anybody with saved compiled patterns that use
2185 \p or \P will have to recompile them.
2186
2187 22. Added "Any" and "L&" to the supported property types.
2188
2189 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2190 but give a compile time error if the value is greater than 0xff.
2191
2192 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2193 accidentally not being installed or uninstalled.
2194
2195 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2196 made were to insert the current release number. This seemed silly, because
2197 it made things harder for people building PCRE on systems that don't run
2198 "configure". I have turned pcre.h into a distributed file, no longer built
2199 by "configure", with the version identification directly included. There is
2200 no longer a pcre.h.in file.
2201
2202 However, this change necessitated a change to the pcre-config script as
2203 well. It is built from pcre-config.in, and one of the substitutions was the
2204 release number. I have updated configure.ac so that ./configure now finds
2205 the release number by grepping pcre.h.
2206
2207 26. Added the ability to run the tests under valgrind.
2208
2209
2210 Version 6.4 05-Sep-05
2211 ---------------------
2212
2213 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2214 "--" to be printed when multiple files were scanned, even when none of the
2215 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2216 consider it to be a bug, and have restored the previous behaviour.
2217
2218 2. A couple of code tidies to get rid of compiler warnings.
2219
2220 3. The pcretest program used to cheat by referring to symbols in the library
2221 whose names begin with _pcre_. These are internal symbols that are not
2222 really supposed to be visible externally, and in some environments it is
2223 possible to suppress them. The cheating is now confined to including
2224 certain files from the library's source, which is a bit cleaner.
2225
2226 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2227 file's purpose clearer.
2228
2229 5. Reorganized pcre_ucp_findchar().
2230
2231
2232 Version 6.3 15-Aug-05
2233 ---------------------
2234
2235 1. The file libpcre.pc.in did not have general read permission in the tarball.
2236
2237 2. There were some problems when building without C++ support:
2238
2239 (a) If C++ support was not built, "make install" and "make test" still
2240 tried to test it.
2241
2242 (b) There were problems when the value of CXX was explicitly set. Some
2243 changes have been made to try to fix these, and ...
2244
2245 (c) --disable-cpp can now be used to explicitly disable C++ support.
2246
2247 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2248 backslash in a target when C++ was disabled. This confuses some
2249 versions of "make", apparently. Using an intermediate variable solves
2250 this. (Same for CPP_LOBJ.)
2251
2252 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2253 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2254 necessary on certain architectures.
2255
2256 4. Added a setting of -export-symbols-regex to the link command to remove
2257 those symbols that are exported in the C sense, but actually are local
2258 within the library, and not documented. Their names all begin with
2259 "_pcre_". This is not a perfect job, because (a) we have to except some
2260 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2261 available (and never for static libraries). I have made a note to try to
2262 find a way round (a) in the future.
2263
2264
2265 Version 6.2 01-Aug-05
2266 ---------------------
2267
2268 1. There was no test for integer overflow of quantifier values. A construction
2269 such as {1111111111111111} would give undefined results. What is worse, if
2270 a minimum quantifier for a parenthesized subpattern overflowed and became
2271 negative, the calculation of the memory size went wrong. This could have
2272 led to memory overwriting.
2273
2274 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2275
2276 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2277 operating environments where this matters.
2278
2279 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2280 PCRE options from within the C++ wrapper.
2281
2282 5. Named capturing subpatterns were not being correctly counted when a pattern
2283 was compiled. This caused two problems: (a) If there were more than 100
2284 such subpatterns, the calculation of the memory needed for the whole
2285 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2286 back references of the form \12, where the number was greater than 9, were
2287 not recognized as back references, even though there were sufficient
2288 previous subpatterns.
2289
2290 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2291 versions of gcc, e.g. 2.95.4.
2292
2293
2294 Version 6.1 21-Jun-05
2295 ---------------------
2296
2297 1. There was one reference to the variable "posix" in pcretest.c that was not
2298 surrounded by "#if !defined NOPOSIX".
2299
2300 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2301 the cross-check on the old pcre_info() function, for the benefit of the
2302 cut-down version of PCRE that is currently imported into Exim.
2303
2304 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2305 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2306 patterns that start (?i) but I don't think that matters. The use of (?i) is
2307 just an example; this all applies to the other options as well.
2308
2309 4. Since libtool seems to echo the compile commands it is issuing, the output
2310 from "make" can be reduced a bit by putting "@" in front of each libtool
2311 compile command.
2312
2313 5. Patch from the folks at Google for configure.in to be a bit more thorough
2314 in checking for a suitable C++ installation before trying to compile the
2315 C++ stuff. This should fix a reported problem when a compiler was present,
2316 but no suitable headers.
2317
2318 6. The man pages all had just "PCRE" as their title. I have changed them to
2319 be the relevant file name. I have also arranged that these names are
2320 retained in the file doc/pcre.txt, which is a concatenation in text format
2321 of all the man pages except the little individual ones for each function.
2322
2323 7. The NON-UNIX-USE file had not been updated for the different set of source
2324 files that come with release 6. I also added a few comments about the C++
2325 wrapper.
2326
2327
2328 Version 6.0 07-Jun-05
2329 ---------------------
2330
2331 1. Some minor internal re-organization to help with my DFA experiments.
2332
2333 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2334 didn't matter for the library itself when fully configured, but did matter
2335 when compiling without UCP support, or within Exim, where the ucp files are
2336 not imported.
2337
2338 3. Refactoring of the library code to split up the various functions into
2339 different source modules. The addition of the new DFA matching code (see
2340 below) to a single monolithic source would have made it really too
2341 unwieldy, quite apart from causing all the code to be include in a
2342 statically linked application, when only some functions are used. This is
2343 relevant even without the DFA addition now that patterns can be compiled in
2344 one application and matched in another.
2345
2346 The downside of splitting up is that there have to be some external
2347 functions and data tables that are used internally in different modules of
2348 the library but which are not part of the API. These have all had their
2349 names changed to start with "_pcre_" so that they are unlikely to clash
2350 with other external names.
2351
2352 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2353 a different (DFA) algorithm. Although it is slower than the original
2354 function, it does have some advantages for certain types of matching
2355 problem.
2356
2357 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2358 including restarting after a partial match.
2359
2360 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2361 defined when compiling for Windows was sent to me. I have put it into the
2362 code, though I have no means of testing or verifying it.
2363
2364 7. Added the pcre_refcount() auxiliary function.
2365
2366 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2367 match before or at the first newline in the subject string. In pcretest,
2368 the /f option on a pattern can be used to set this.
2369
2370 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2371 would behave wrongly. This has been present in PCRE since release 4.0.
2372
2373 10. A number of changes to the pcregrep command:
2374
2375 (a) Refactored how -x works; insert ^(...)$ instead of setting
2376 PCRE_ANCHORED and checking the length, in preparation for adding
2377 something similar for -w.
2378
2379 (b) Added the -w (match as a word) option.
2380
2381 (c) Refactored the way lines are read and buffered so as to have more
2382 than one at a time available.
2383
2384 (d) Implemented a pcregrep test script.
2385
2386 (e) Added the -M (multiline match) option. This allows patterns to match
2387 over several lines of the subject. The buffering ensures that at least
2388 8K, or the rest of the document (whichever is the shorter) is available
2389 for matching (and similarly the previous 8K for lookbehind assertions).
2390
2391 (f) Changed the --help output so that it now says
2392
2393 -w, --word-regex(p)
2394
2395 instead of two lines, one with "regex" and the other with "regexp"
2396 because that confused at least one person since the short forms are the
2397 same. (This required a bit of code, as the output is generated
2398 automatically from a table. It wasn't just a text change.)
2399
2400 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2401 option but starts with a hyphen. Could be a pattern or a path name
2402 starting with a hyphen, for instance.
2403
2404 (h) "-" can be given as a file name to represent stdin.
2405
2406 (i) When file names are being printed, "(standard input)" is used for
2407 the standard input, for compatibility with GNU grep. Previously
2408 "<stdin>" was used.
2409
2410 (j) The option --label=xxx can be used to supply a name to be used for
2411 stdin when file names are being printed. There is no short form.
2412
2413 (k) Re-factored the options decoding logic because we are going to add
2414 two more options that take data. Such options can now be given in four
2415 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2416
2417 (l) Added the -A, -B, and -C options for requesting that lines of context
2418 around matches be printed.
2419
2420 (m) Added the -L option to print the names of files that do not contain
2421 any matching lines, that is, the complement of -l.
2422
2423 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2424 continue to scan other files.
2425
2426 (o) The -s option was incorrectly implemented. For compatibility with other
2427 greps, it now suppresses the error message for a non-existent or non-
2428 accessible file (but not the return code). There is a new option called
2429 -q that suppresses the output of matching lines, which was what -s was
2430 previously doing.
2431
2432 (p) Added --include and --exclude options to specify files for inclusion
2433 and exclusion when recursing.
2434
2435 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2436 Hopefully, it now does.
2437
2438 12. Missing cast in pcre_study().
2439
2440 13. Added an "uninstall" target to the makefile.
2441
2442 14. Replaced "extern" in the function prototypes in Makefile.in with
2443 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2444 world, but is set differently for Windows.
2445
2446 15. Added a second compiling function called pcre_compile2(). The only
2447 difference is that it has an extra argument, which is a pointer to an
2448 integer error code. When there is a compile-time failure, this is set
2449 non-zero, in addition to the error test pointer being set to point to an
2450 error message. The new argument may be NULL if no error number is required
2451 (but then you may as well call pcre_compile(), which is now just a
2452 wrapper). This facility is provided because some applications need a
2453 numeric error indication, but it has also enabled me to tidy up the way
2454 compile-time errors are handled in the POSIX wrapper.
2455
2456 16. Added VPATH=.libs to the makefile; this should help when building with one
2457 prefix path and installing with another. (Or so I'm told by someone who
2458 knows more about this stuff than I do.)
2459
2460 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2461 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2462 match everything, including newlines. This is not POSIX-compatible, but
2463 somebody wanted the feature. From pcretest it can be activated by using
2464 both the P and the s flags.
2465
2466 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2467
2468 19. libpcre.pc was being incorrectly installed as executable.
2469
2470 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2471 it now also looks for '\r' so that it will work unmodified on Windows.
2472
2473 21. Added Google's contributed C++ wrapper to the distribution.
2474
2475 22. Added some untidy missing memory free() calls in pcretest, to keep
2476 Electric Fence happy when testing.
2477
2478
2479
2480 Version 5.0 13-Sep-04
2481 ---------------------
2482
2483 1. Internal change: literal characters are no longer packed up into items
2484 containing multiple characters in a single byte-string. Each character
2485 is now matched using a separate opcode. However, there may be more than one
2486 byte in the character in UTF-8 mode.
2487
2488 2. The pcre_callout_block structure has two new fields: pattern_position and
2489 next_item_length. These contain the offset in the pattern to the next match
2490 item, and its length, respectively.
2491
2492 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2493 insertion of callouts before each pattern item. Added the /C option to
2494 pcretest to make use of this.
2495
2496 4. On the advice of a Windows user, the lines
2497
2498 #if defined(_WIN32) || defined(WIN32)
2499 _setmode( _fileno( stdout ), 0x8000 );
2500 #endif /* defined(_WIN32) || defined(WIN32) */
2501
2502 have been added to the source of pcretest. This apparently does useful
2503 magic in relation to line terminators.
2504
2505 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2506 for the benefit of those environments where the "b" makes a difference.
2507
2508 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2509 to know about it. I have put a hack into configure.in that adds in code
2510 to set GCC=yes if CC=icc. This seems to end up at a point in the
2511 generated configure script that is early enough to affect the setting of
2512 compiler options, which is what is needed, but I have no means of testing
2513 whether it really works. (The user who reported this had patched the
2514 generated configure script, which of course I cannot do.)
2515
2516 LATER: After change 22 below (new libtool files), the configure script
2517 seems to know about icc (and also ecc). Therefore, I have commented out
2518 this hack in configure.in.
2519
2520 7. Added support for pkg-config (2 patches were sent in).
2521
2522 8. Negated POSIX character classes that used a combination of internal tables
2523 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2524 [[:^ascii]]. Typically, they would match almost any characters. The other
2525 POSIX classes were not broken in this way.
2526
2527 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2528 to find the match, as PCRE was deluded into thinking that the match had to
2529 start at the start point or following a newline. The same bug applied to
2530 patterns with negative forward assertions or any backward assertions
2531 preceding ".*" at the start, unless the pattern required a fixed first
2532 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2533
2534 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2535 starting at the last subject character, bytes beyond the end of the subject
2536 string were read.
2537
2538 11. Renamed the variable "class" as "classbits" to make life easier for C++
2539 users. (Previously there was a macro definition, but it apparently wasn't
2540 enough.)
2541
2542 12. Added the new field "tables" to the extra data so that tables can be passed
2543 in at exec time, or the internal tables can be re-selected. This allows
2544 a compiled regex to be saved and re-used at a later time by a different
2545 program that might have everything at different addresses.
2546
2547 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2548 -R library as well as a -L library.
2549
2550 14. The debugging options of pcretest (-d on the command line or D on a
2551 pattern) showed incorrect output for anything following an extended class
2552 that contained multibyte characters and which was followed by a quantifier.
2553
2554 15. Added optional support for general category Unicode character properties
2555 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2556 support. It adds about 90K to the size of the library. The meanings of the
2557 inbuilt class escapes such as \d and \s have NOT been changed.
2558
2559 16. Updated pcredemo.c to include calls to free() to release the memory for the
2560 compiled pattern.
2561
2562 17. The generated file chartables.c was being created in the source directory
2563 instead of in the building directory. This caused the build to fail if the
2564 source directory was different from the building directory, and was
2565 read-only.
2566
2567 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2568 file. No doubt somebody will tell me if they don't make sense... Also added
2569 Dan Mooney's comments about building on OpenVMS.
2570
2571 19. Added support for partial matching via the PCRE_PARTIAL option for
2572 pcre_exec() and the \P data escape in pcretest.
2573
2574 20. Extended pcretest with 3 new pattern features:
2575
2576 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2577 write the compiled pattern to the file whose name is "rest-of-line".
2578 This is a straight binary dump of the data, with the saved pointer to
2579 the character tables forced to be NULL. The study data, if any, is
2580 written too. After writing, pcretest reads a new pattern.
2581
2582 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2583 compiled pattern from the given file. There must not be any
2584 occurrences of "<" in the file name (pretty unlikely); if there are,
2585 pcretest will instead treat the initial "<" as a pattern delimiter.
2586 After reading in the pattern, pcretest goes on to read data lines as
2587 usual.
2588
2589 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2590 and 16-bit fields in a compiled pattern, to simulate a pattern that
2591 was compiled on a host of opposite endianness.
2592
2593 21. The pcre-exec() function can now cope with patterns that were compiled on
2594 hosts of opposite endianness, with this restriction:
2595
2596 As for any compiled expression that is saved and used later, the tables
2597 pointer field cannot be preserved; the extra_data field in the arguments
2598 to pcre_exec() should be used to pass in a tables address if a value
2599 other than the default internal tables were used at compile time.
2600
2601 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2602 now diagnosed as an error. Previously, most of the time, a negative number
2603 would have been treated as zero, but if in addition "ovector" was passed as
2604 NULL, a crash could occur.
2605
2606 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2607 new versions from the libtool 1.5 distribution (the last one is a copy of
2608 a file called libtool.m4). This seems to have fixed the need to patch
2609 "configure" to support Darwin 1.3 (which I used to do). However, I still
2610 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2611 workstation).
2612
2613 24. Changed the PCRE licence to be the more standard "BSD" licence.
2614
2615
2616 Version 4.5 01-Dec-03
2617 ---------------------
2618
2619 1. There has been some re-arrangement of the code for the match() function so
2620 that it can be compiled in a version that does not call itself recursively.
2621 Instead, it keeps those local variables that need separate instances for
2622 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2623 needs to "recurse". Keeping track of where control must go is done by means
2624 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2625 hide most of the details from the main code, and operates only if
2626 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2627 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2628 operating.
2629
2630 To make it easier for callers to provide specially tailored get/free
2631 functions for this usage, two new functions, pcre_stack_malloc, and
2632 pcre_stack_free, are used. They are always called in strict stacking order,
2633 and the size of block requested is always the same.
2634
2635 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2636 PCRE has been compiled to use the stack or the heap for recursion. The
2637 -C option of pcretest uses this to show which version is compiled.
2638
2639 A new data escape \S, is added to pcretest; it causes the amounts of store
2640 obtained and freed by both kinds of malloc/free at match time to be added
2641 to the output.
2642
2643 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2644 what's available on my current Linux desktop machine.
2645
2646 3. When matching a UTF-8 string, the test for a valid string at the start has
2647 been extended. If start_offset is not zero, PCRE now checks that it points
2648 to a byte that is the start of a UTF-8 character. If not, it returns
2649 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2650 this is necessary because there may be backward assertions in the pattern.
2651 When matching the same subject several times, it may save resources to use
2652 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2653
2654 4. The code for checking the validity of UTF-8 strings has been tightened so
2655 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2656 containing "overlong sequences".
2657
2658 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2659 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2660 should have been "&", but it just so happened that all the cases this let
2661 through by mistake were picked up later in the function.
2662
2663 6. I had used a variable called "isblank" - this is a C99 function, causing
2664 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2665
2666 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2667 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2668 so the version is shown; (c) stop "make test" echoing "./RunTest".
2669
2670 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2671
2672 9. The prototype for memmove() for systems that don't have it was using
2673 size_t, but the inclusion of the header that defines size_t was later. I've
2674 moved the #includes for the C headers earlier to avoid this.
2675
2676 10. Added some adjustments to the code to make it easier to compiler on certain
2677 special systems:
2678
2679 (a) Some "const" qualifiers were missing.
2680 (b) Added the macro EXPORT before all exported functions; by default this
2681 is defined to be empty.
2682 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2683 that it reads its output file name as an argument instead of writing
2684 to the standard output and assuming this can be redirected.
2685
2686 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2687 class containing characters with values greater than 255, PCRE compilation
2688 went into a loop.
2689
2690 12. A recursive reference to a subpattern that was within another subpattern
2691 that had a minimum quantifier of zero caused PCRE to crash. For example,
2692 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2693 recursion. If the recursively-called subpattern itself had a zero repeat,
2694 that was OK.
2695
2696 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2697 buffer into which it was copied (for escape processing) was still set at
2698 1024, so long lines caused crashes.
2699
2700 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2701 "internal error: code overflow...". This applied to any character class
2702 that was followed by a possessive quantifier.
2703
2704 15. Modified the Makefile to add libpcre.la as a prerequisite for
2705 libpcreposix.la because I was told this is needed for a parallel build to
2706 work.
2707
2708 16. If a pattern that contained .* following optional items at the start was
2709 studied, the wrong optimizing data was generated, leading to matching
2710 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2711 matching string must start with a or b or c. The correct conclusion for
2712 this pattern is that a match can start with any character.
2713
2714
2715 Version 4.4 13-Aug-03
2716 ---------------------
2717
2718 1. In UTF-8 mode, a character class containing characters with values between
2719 127 and 255 was not handled correctly if the compiled pattern was studied.
2720 In fixing this, I have also improved the studying algorithm for such
2721 classes (slightly).
2722
2723 2. Three internal functions had redundant arguments passed to them. Removal
2724 might give a very teeny performance improvement.
2725
2726 3. Documentation bug: the value of the capture_top field in a callout is *one
2727 more than* the number of the hightest numbered captured substring.
2728
2729 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2730 in incorrectly linking with a previously installed version. They now link
2731 explicitly with libpcre.la.
2732
2733 5. configure.in no longer needs to recognize Cygwin specially.
2734
2735 6. A problem in pcre.in for Windows platforms is fixed.
2736
2737 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2738 pcretest, it used to include the size of the study block as part of its
2739 output. Unfortunately, the structure contains a field that has a different
2740 size on different hardware architectures. This meant that the tests that
2741 showed this size failed. As the block is currently always of a fixed size,
2742 this information isn't actually particularly useful in pcretest output, so
2743 I have just removed it.
2744
2745 8. Three pre-processor statements accidentally did not start in column 1.
2746 Sadly, there are *still* compilers around that complain, even though
2747 standard C has not required this for well over a decade. Sigh.
2748
2749 9. In pcretest, the code for checking callouts passed small integers in the
2750 callout_data field, which is a void * field. However, some picky compilers
2751 complained about the casts involved for this on 64-bit systems. Now
2752 pcretest passes the address of the small integer instead, which should get
2753 rid of the warnings.
2754
2755 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2756 both compile and run time, and gives an error if an invalid UTF-8 sequence
2757 is found. There is a option for disabling this check in cases where the
2758 string is known to be correct and/or the maximum performance is wanted.
2759
2760 11. In response to a bug report, I changed one line in Makefile.in from
2761
2762 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2763 to
2764 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2765
2766 to look similar to other lines, but I have no way of telling whether this
2767 is the right thing to do, as I do not use Windows. No doubt I'll get told
2768 if it's wrong...
2769
2770
2771 Version 4.3 21-May-03
2772 ---------------------
2773
2774 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2775 Makefile.
2776
2777 2. Some refactoring to improve the quality of the code:
2778
2779 (i) The utf8_table... variables are now declared "const".
2780
2781 (ii) The code for \cx, which used the "case flipping" table to upper case
2782 lower case letters, now just substracts 32. This is ASCII-specific,
2783 but the whole concept of \cx is ASCII-specific, so it seems
2784 reasonable.
2785
2786 (iii) PCRE was using its character types table to recognize decimal and
2787 hexadecimal digits in the pattern. This is silly, because it handles
2788 only 0-9, a-f, and A-F, but the character types table is locale-
2789 specific, which means strange things might happen. A private
2790 table is now used for this - though it costs 256 bytes, a table is
2791 much faster than multiple explicit tests. Of course, the standard
2792 character types table is still used for matching digits in subject
2793 strings against \d.
2794
2795 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2796 ending in _t are). So I've renamed it as ESC_tee.
2797
2798 3. The first argument for regexec() in the POSIX wrapper should have been
2799 defined as "const".
2800
2801 4. Changed pcretest to use malloc() for its buffers so that they can be
2802 Electric Fenced for debugging.
2803
2804 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2805 to read one or more bytes before the start of the subject string. Often this
2806 had no effect on PCRE's behaviour, but in some circumstances it could
2807 provoke a segmentation fault.
2808
2809 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2810 to try to read one or more bytes before the start of the subject string.
2811
2812 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2813 UTF-8 support could misbehave in various ways if the subject string
2814 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2815 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2816 back over UTF-8 characters.)
2817
2818
2819 Version 4.2 14-Apr-03
2820 ---------------------
2821
2822 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2823
2824 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2825 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2826 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2827 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2828 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2829 and BUILD_EXEEXT
2830 Note: automatic setting of the BUILD variables is not yet working
2831 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2832 compile-time but not at link-time
2833 [LINK]: use for linking executables only
2834 make different versions for Windows and non-Windows
2835 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2836 libraries
2837 [LINK_FOR_BUILD]: new variable
2838 [OBJEXT]: use throughout
2839 [EXEEXT]: use throughout
2840 <winshared>: new target
2841 <wininstall>: new target
2842 <dftables.o>: use native compiler
2843 <dftables>: use native linker
2844 <install>: handle Windows platform correctly
2845 <clean>: ditto
2846 <check>: ditto
2847 copy DLL to top builddir before testing
2848
2849 As part of these changes, -no-undefined was removed again. This was reported
2850 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2851 in any case.
2852
2853 3. Some tidies to get rid of compiler warnings:
2854
2855 . In the match_data structure, match_limit was an unsigned long int, whereas
2856 match_call_count was an int. I've made them both unsigned long ints.
2857
2858 . In pcretest the fact that a const uschar * doesn't automatically cast to
2859 a void * provoked a warning.
2860
2861 . Turning on some more compiler warnings threw up some "shadow" variables
2862 and a few more missing casts.
2863
2864 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2865 option, a class that contained a single character with a value between 128
2866 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2867
2868 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2869 option, a class that contained several characters, but with at least one
2870 whose value was between 128 and 255 caused PCRE to crash.
2871
2872
2873 Version 4.1 12-Mar-03
2874 ---------------------
2875
2876 1. Compiling with gcc -pedantic found a couple of places where casts were
2877 needed, and a string in dftables.c that was longer than standard compilers are
2878 required to support.
2879
2880 2. Compiling with Sun's compiler found a few more places where the code could
2881 be tidied up in order to avoid warnings.
2882
2883 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2884 first of these names is deprecated in the latest Autoconf in favour of the name
2885 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2886 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2887 analogy I have changed it to CFLAGS_FOR_BUILD.
2888
2889 4. Added -no-undefined to the linking command in the Makefile, because this is
2890 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2891 linking step for the pcreposix library.
2892
2893 5. PCRE was failing to diagnose the case of two named groups with the same
2894 name.
2895
2896 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2897 literal character that is needed in the subject for a match, and scans along to
2898 ensure that it is present before embarking on the full matching process. This
2899 saves time in cases of nested unlimited repeats that are never going to match.
2900 Problem: the scan can take a lot of time if the subject is very long (e.g.
2901 megabytes), thus penalizing straightforward matches. It is now done only if the
2902 amount of subject to be scanned is less than 1000 bytes.
2903
2904 7. A lesser problem with the same optimization is that it was recording the
2905 first character of an anchored pattern as "needed", thus provoking a search
2906 right along the subject, even when the first match of the pattern was going to
2907 fail. The "needed" character is now not set for anchored patterns, unless it
2908 follows something in the pattern that is of non-fixed length. Thus, it still
2909 fulfils its original purpose of finding quick non-matches in cases of nested
2910 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2911
2912
2913 Version 4.0 17-Feb-03
2914 ---------------------
2915
2916 1. If a comment in an extended regex that started immediately after a meta-item
2917 extended to the end of string, PCRE compiled incorrect data. This could lead to
2918 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2919
2920 2. Moved to autoconf 2.53 and libtool 1.4.2.
2921
2922 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2923 the special perltest8 script is no longer needed - all the tests can be run
2924 from a single perltest script.
2925
2926 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2927 by \s. It has now been removed in PCRE. This means it isn't recognized as
2928 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2929 class [:space:] *does* include VT, thereby creating a mess.
2930
2931 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2932 space and tab.
2933
2934 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2935 its new features into the main test script, reducing the number of scripts.
2936
2937 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2938 were backward compatible, and made the (?i) apply to the whole pattern, as if
2939 /i were given. Now it behaves more logically, and applies the option setting
2940 only to what follows. PCRE has been changed to follow suit. However, if it
2941 finds options settings right at the start of the pattern, it extracts them into
2942 the global options, as before. Thus, they show up in the info data.
2943
2944 8. Added support for the \Q...\E escape sequence. Characters in between are
2945 treated as literals. This is slightly different from Perl in that $ and @ are
2946 also handled as literals inside the quotes. In Perl, they will cause variable
2947 interpolation. Note the following examples:
2948
2949 Pattern PCRE matches Perl matches
2950
2951 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2952 \Qabc\$xyz\E abc\$xyz abc\$xyz
2953 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2954
2955 For compatibility with Perl, \Q...\E sequences are recognized inside character
2956 classes as well as outside them.
2957
2958 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2959 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2960 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2961 signed/unsigned warnings.
2962
2963 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2964 option for pcretest, so I've replaced it by a simple function that does just
2965 that job.
2966
2967 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2968 "pcregrep -".
2969
2970 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2971 Java package. This provides some syntactic sugar for simple cases of what my
2972 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2973 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2974 item, you can use this simplified notation. Note that only makes sense with
2975 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2976 greediness, whatever the setting of the PCRE_UNGREEDY option.
2977
2978 13. A change of greediness default within a pattern was not taking effect at
2979 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2980 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2981 was abstracted outside.
2982
2983 14. PCRE now supports the \G assertion. It is true when the current matching
2984 position is at the start point of the match. This differs from \A when the
2985 starting offset is non-zero. Used with the /g option of pcretest (or similar
2986 code), it works in the same way as it does for Perl's /g option. If all
2987 alternatives of a regex begin with \G, the expression is anchored to the start
2988 match position, and the "anchored" flag is set in the compiled expression.
2989
2990 15. Some bugs concerning the handling of certain option changes within patterns
2991 have been fixed. These applied to options other than (?ims). For example,
2992 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2993 been the other way round. Some of this was related to change 7 above.
2994
2995 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2996 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2997 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2998 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2999
3000 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3001 mode. Unlike ".", it always matches newline, whatever the setting of
3002 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3003 assertions. Perl allows it, but it doesn't (in general) work because it can't
3004 calculate the length of the lookbehind. At least, that's the case for Perl
3005 5.8.0 - I've been told they are going to document that it doesn't work in
3006 future.
3007
3008 18. Added an error diagnosis for escapes that PCRE does not support: these are
3009 \L, \l, \N, \P, \p, \U, \u, and \X.
3010
3011 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3012 reading past the end of the pattern in cases such as /[abcd/.
3013
3014 20. PCRE was getting more memory than necessary for patterns with classes that
3015 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3016
3017 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3018 compiling PCRE for use with Virtual Pascal.
3019
3020 22. Small fix to the Makefile to make it work properly if the build is done
3021 outside the source tree.
3022
3023 23. Added a new extension: a condition to go with recursion. If a conditional
3024 subpattern starts with (?(R) the "true" branch is used if recursion has
3025 happened, whereas the "false" branch is used only at the top level.
3026
3027 24. When there was a very long string of literal characters (over 255 bytes
3028 without UTF support, over 250 bytes with UTF support), the computation of how
3029 much memory was required could be incorrect, leading to segfaults or other
3030 strange effects.
3031
3032 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3033 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3034 there was a subsequent back reference to those brackets. This meant that, for
3035 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3036 possible to check for precisely this case. All we can do is abandon the
3037 optimization if .* occurs inside capturing brackets when there are any back
3038 references whatsoever. (See below for a better fix that came later.)
3039
3040 26. The handling of the optimization for finding the first character of a
3041 non-anchored pattern, and for finding a character that is required later in the
3042 match were failing in some cases. This didn't break the matching; it just
3043 failed to optimize when it could. The way this is done has been re-implemented.
3044
3045 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3046
3047 28. Added a new feature that provides some of the functionality that Perl
3048 provides with (?{...}). The facility is termed a "callout". The way it is done
3049 in PCRE is for the caller to provide an optional function, by setting
3050 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3051 global variable. By default it is unset, which disables all calling out. To get
3052 the function called, the regex must include (?C) at appropriate points. This
3053 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3054 This provides a means of identifying different callout points. When PCRE
3055 reaches such a point in the regex, if pcre_callout has been set, the external
3056 function is called. It is provided with data in a structure called
3057 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3058 matching continues; if it returns a non-zero value, the match at the current
3059 point fails. However, backtracking will occur if possible. [This was changed
3060 later and other features added - see item 49 below.]
3061
3062 29. pcretest is upgraded to test the callout functionality. It provides a
3063 callout function that displays information. By default, it shows the start of
3064 the match and the current position in the text. There are some new data escapes
3065 to vary what happens:
3066
3067 \C+ in addition, show current contents of captured substrings
3068 \C- do not supply a callout function
3069 \C!n return 1 when callout number n is reached
3070 \C!n!m return 1 when callout number n is reached for the mth time
3071
3072 30. If pcregrep was called with the -l option and just a single file name, it
3073 output "<stdin>" if a match was found, instead of the file name.
3074
3075 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3076 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3077 pcre_exec(). This saves a malloc/free per call. The default value of
3078 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3079 when configuring.
3080
3081 32. The default maximum size of a compiled pattern is 64K. There have been a
3082 few cases of people hitting this limit. The code now uses macros to handle the
3083 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3084 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3085 configuring. Tests 2 and 5 work only with 2-byte links because they output
3086 debugging information about compiled patterns.
3087
3088 33. Internal code re-arrangements:
3089
3090 (a) Moved the debugging function for printing out a compiled regex into
3091 its own source file (printint.c) and used #include to pull it into
3092 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3093 separate copies.
3094
3095 (b) Defined the list of op-code names for debugging as a macro in
3096 internal.h so that it is next to the definition of the opcodes.
3097
3098 (c) Defined a table of op-code lengths for simpler skipping along compiled
3099 code. This is again a macro in internal.h so that it is next to the
3100 definition of the opcodes.
3101
3102 34. Added support for recursive calls to individual subpatterns, along the
3103 lines of Robin Houston's patch (but implemented somewhat differently).
3104
3105 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3106 allow it to read and process whole directories in Win32. This code was
3107 contributed by Lionel Fourquaux; it has not been tested by me.
3108
3109 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3110 used to name a group. Names consist of alphanumerics and underscores, and must
3111 be unique. Back references use the syntax (?P=name) and recursive calls use
3112 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3113 numbers. The function pcre_fullinfo() can be used after compilation to extract
3114 a name/number map. There are three relevant calls:
3115
3116 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3117 PCRE_INFO_NAMECOUNT yields the number of entries
3118 PCRE_INFO_NAMETABLE yields a pointer to the map.
3119
3120 The map is a vector of fixed-size entries. The size of each entry depends on
3121 the length of the longest name used. The first two bytes of each entry are the
3122 group number, most significant byte first. There follows the corresponding
3123 name, zero terminated. The names are in alphabetical order.
3124
3125 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3126 case instead of 255. Making it the same both with and without UTF-8 support
3127 means that the same test output works with both.
3128
3129 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3130 calling malloc() with a zero argument.
3131
3132 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3133 optimization. I've improved things by keeping a bitmap of backreferences with
3134 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3135 fact referenced, the optimization can be applied. It is unlikely that a
3136 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3137 the match to follow \n) will appear inside brackets with a number greater than
3138 31, but if it does, any back reference > 31 suppresses the optimization.
3139
3140 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3141 of disabling numbered capturing parentheses. Any opening parenthesis that is
3142 not followed by ? behaves as if it were followed by ?: but named parentheses
3143 can still be used for capturing (and they will acquire numbers in the usual
3144 way).
3145
3146 41. Redesigned the return codes from the match() function into yes/no/error so
3147 that errors can be passed back from deep inside the nested calls. A malloc
3148 failure while inside a recursive subpattern call now causes the
3149 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3150
3151 42. It is now possible to set a limit on the number of times the match()
3152 function is called in a call to pcre_exec(). This facility makes it possible to
3153 limit the amount of recursion and backtracking, though not in a directly
3154 obvious way, because the match() function is used in a number of different
3155 circumstances. The count starts from zero for each position in the subject
3156 string (for non-anchored patterns). The default limit is, for compatibility, a
3157 large number, namely 10 000 000. You can change this in two ways:
3158
3159 (a) When configuring PCRE before making, you can use --with-match-limit=n
3160 to set a default value for the compiled library.
3161
3162 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3163 a different value is set. See 45 below.
3164
3165 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3166
3167 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3168 of things that can be changed at compile time. The first argument specifies
3169 what is wanted and the second points to where the information is to be placed.
3170 The current list of available information is:
3171
3172 PCRE_CONFIG_UTF8
3173
3174 The output is an integer that is set to one if UTF-8 support is available;
3175 otherwise it is set to zero.
3176
3177 PCRE_CONFIG_NEWLINE
3178
3179 The output is an integer that it set to the value of the code that is used for
3180 newline. It is either LF (10) or CR (13).
3181
3182 PCRE_CONFIG_LINK_SIZE
3183
3184 The output is an integer that contains the number of bytes used for internal
3185 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3186
3187 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3188
3189 The output is an integer that contains the threshold above which the POSIX
3190 interface uses malloc() for output vectors. See item 31 above.
3191
3192 PCRE_CONFIG_MATCH_LIMIT
3193
3194 The output is an unsigned integer that contains the default limit of the number
3195 of match() calls in a pcre_exec() execution. See 42 above.
3196
3197 44. pcretest has been upgraded by the addition of the -C option. This causes it
3198 to extract all the available output from the new pcre_config() function, and to
3199 output it. The program then exits immediately.
3200
3201 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3202 order to support additional features. One way would have been to define
3203 pcre_exec2() (for example) with extra arguments, but this would not have been
3204 extensible, and would also have required all calls to the original function to
3205 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3206 is used for passing in "extra" data from pcre_study().
3207
3208 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3209 contains the following fields:
3210
3211 flags a bitmap indicating which of the following fields are set
3212 study_data opaque data from pcre_study()
3213 match_limit a way of specifying a limit on match() calls for a specific
3214 call to pcre_exec()
3215 callout_data data for callouts (see 49 below)
3216
3217 The flag bits are also defined in pcre.h, and are
3218
3219 PCRE_EXTRA_STUDY_DATA
3220 PCRE_EXTRA_MATCH_LIMIT
3221 PCRE_EXTRA_CALLOUT_DATA
3222
3223 The pcre_study() function now returns one of these new pcre_extra blocks, with
3224 the actual study data pointed to by the study_data field, and the
3225 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3226 before. That is, this change is entirely upwards-compatible and requires no
3227 change to existing code.
3228
3229 If you want to pass in additional data to pcre_exec(), you can either place it
3230 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3231 block.
3232
3233 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3234 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3235 times with different match limits, until it finds the minimum value needed for
3236 pcre_exec() to complete. The value is then output. This can be instructive; for
3237 most simple matches the number is quite small, but for pathological cases it
3238 gets very large very quickly.
3239
3240 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3241 returns the size of the data block pointed to by the study_data field in a
3242 pcre_extra block, that is, the value that was passed as the argument to
3243 pcre_malloc() when PCRE was getting memory in which to place the information
3244 created by pcre_study(). The fourth argument should point to a size_t variable.
3245 pcretest has been extended so that this information is shown after a successful
3246 pcre_study() call when information about the compiled regex is being displayed.
3247
3248 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3249 because what follows is always an absolute path. (Later: it turns out that this
3250 is more than cosmetic for MinGW, because it doesn't like empty path
3251 components.)
3252
3253 49. Some changes have been made to the callout feature (see 28 above):
3254
3255 (i) A callout function now has three choices for what it returns:
3256
3257 0 => success, carry on matching
3258 > 0 => failure at this point, but backtrack if possible
3259 < 0 => serious error, return this value from pcre_exec()
3260
3261 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3262 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3263 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3264 use by callout functions. It will never be used by PCRE itself.
3265
3266 (ii) The pcre_extra structure (see 45 above) has a void * field called
3267 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3268 pcre_callout_block structure has a field of the same name. The contents of
3269 the field passed in the pcre_extra structure are passed to the callout
3270 function in the corresponding field in the callout block. This makes it
3271 easier to use the same callout-containing regex from multiple threads. For
3272 testing, the pcretest program has a new data escape
3273
3274 \C*n pass the number n (may be negative) as callout_data
3275
3276 If the callout function in pcretest receives a non-zero value as
3277 callout_data, it returns that value.
3278
3279 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3280 there were some redundant $(CFLAGS) in commands that are now specified as
3281 $(LINK), which already includes $(CFLAGS).
3282
3283 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3284 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3285 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3286 one-byte characters throughout. Note that case-insensitive matching applies
3287 only to characters whose values are less than 256. PCRE doesn't support the
3288 notion of cases for higher-valued characters.
3289
3290 (i) A character class whose characters are all within 0-255 is handled as
3291 a bit map, and the map is inverted for negative classes. Previously, a
3292 character > 255 always failed to match such a class; however it should
3293 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3294
3295 (ii) A negated character class with a single character < 255 is coded as
3296 "not this character" (OP_NOT). This wasn't working properly when the test
3297 character was multibyte, either singly or repeated.
3298
3299 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3300 mode, for example: \x{100}{2,3}.
3301
3302 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3303 singly or repeated) now correctly test multibyte characters. However,
3304 PCRE doesn't recognize any characters with values greater than 255 as
3305 digits, spaces, or word characters. Such characters always match \D, \S,
3306 and \W, and never match \d, \s, or \w.
3307
3308 (v) Classes may now contain characters and character ranges with values
3309 greater than 255. For example: [ab\x{100}-\x{400}].
3310
3311 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3312 PCRE in UTF-8 mode.
3313
3314 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3315 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3316 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3317 value.)
3318
3319 53. The single man page has become too large. I have therefore split it up into
3320 a number of separate man pages. These also give rise to individual HTML pages;
3321 these are now put in a separate directory, and there is an index.html page that
3322 lists them all. Some hyperlinking between the pages has been installed.
3323
3324 54. Added convenience functions for handling named capturing parentheses.
3325
3326 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3327 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3328 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3329 are faulted.
3330
3331 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3332 calling configure. These values are used when compiling the dftables.c program
3333 which is run to generate the source of the default character tables. They
3334 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3335 you will need to set these values.
3336
3337 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3338
3339
3340 Version 3.9 02-Jan-02
3341 ---------------------
3342
3343 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3344
3345 2. If --disable-static was given, the building process failed when trying to
3346 build pcretest and pcregrep. (For some reason it was using libtool to compile
3347 them, which is not right, as they aren't part of the library.)
3348
3349
3350 Version 3.8 18-Dec-01
3351 ---------------------
3352
3353 1. The experimental UTF-8 code was completely screwed up. It was packing the
3354 bytes in the wrong order. How dumb can you get?
3355
3356
3357 Version 3.7 29-Oct-01
3358 ---------------------
3359
3360 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3361 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3362 this didn't happen under Solaris 8, where I normally test things.
3363
3364 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3365 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3366 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3367 this fix an hour or so after the initial 3.7 release.)
3368
3369
3370 Version 3.6 23-Oct-01
3371 ---------------------
3372
3373 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3374 offsets passed as NULL with zero offset count.
3375
3376 2. The config.guess and config.sub files had not been updated when I moved to
3377 the latest autoconf.
3378
3379
3380 Version 3.5 15-Aug-01
3381 ---------------------
3382
3383 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3384 had been forgotten.
3385
3386 2. By using declared but undefined structures, we can avoid using "void"
3387 definitions in pcre.h while keeping the internal definitions of the structures
3388 private.
3389
3390 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3391 user point of view, this means that both static and shared libraries are built
3392 by default, but this can be individually controlled. More of the work of
3393 handling this static/shared cases is now inside libtool instead of PCRE's make
3394 file.
3395
3396 4. The pcretest utility is now installed along with pcregrep because it is
3397 useful for users (to test regexs) and by doing this, it automatically gets
3398 relinked by libtool. The documentation has been turned into a man page, so
3399 there are now .1, .txt, and .html versions in /doc.
3400
3401 5. Upgrades to pcregrep:
3402 (i) Added long-form option names like gnu grep.
3403 (ii) Added --help to list all options with an explanatory phrase.
3404 (iii) Added -r, --recursive to recurse into sub-directories.
3405 (iv) Added -f, --file to read patterns from a file.
3406
3407 6. pcre_exec() was referring to its "code" argument before testing that
3408 argument for NULL (and giving an error if it was NULL).
3409
3410 7. Upgraded Makefile.in to allow for compiling in a different directory from
3411 the source directory.
3412
3413 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3414 options bits, the pointer it was passed was to an int instead of to an unsigned
3415 long int. This mattered only on 64-bit systems.
3416
3417 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3418 generated) instead of pcre.in, which it its source. Also made the same change
3419 in several of the .c files.
3420
3421 10. A new release of gcc defines printf() as a macro, which broke pcretest
3422 because it had an ifdef in the middle of a string argument for printf(). Fixed
3423 by using separate calls to printf().
3424
3425 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3426 script, to force use of CR or LF instead of \n in the source. On non-Unix
3427 systems, the value can be set in config.h.
3428
3429 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3430 absolute limit. Changed the text of the error message to make this clear, and
3431 likewise updated the man page.
3432
3433 13. The limit of 99 on the number of capturing subpatterns has been removed.
3434 The new limit is 65535, which I hope will not be a "real" limit.
3435
3436
3437 Version 3.4 22-Aug-00
3438 ---------------------
3439
3440 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3441
3442 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3443
3444
3445 Version 3.3 01-Aug-00
3446 ---------------------
3447
3448 1. If an octal character was given, but the value was greater than \377, it
3449 was not getting masked to the least significant bits, as documented. This could
3450 lead to crashes in some systems.
3451
3452 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3453 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3454
3455 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3456 These just pass their arguments on to (pcre_free)(), but they are provided
3457 because some uses of PCRE bind it to non-C systems that can call its functions,
3458 but cannot call free() or pcre_free() directly.
3459
3460 4. Add "make test" as a synonym for "make check". Corrected some comments in
3461 the Makefile.
3462
3463 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3464 Makefile.
3465
3466 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3467 command called pgrep for grepping around the active processes.
3468
3469 7. Added the beginnings of support for UTF-8 character strings.
3470
3471 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3472 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3473 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3474 out for the ar command.)
3475
3476
3477 Version 3.2 12-May-00
3478 ---------------------
3479
3480 This is purely a bug fixing release.
3481
3482 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3483 of ZA. This was just one example of several cases that could provoke this bug,
3484 which was introduced by change 9 of version 2.00. The code for breaking
3485 infinite loops after an iteration that matches an empty string was't working
3486 correctly.
3487
3488 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3489 when matched against abbab (for example). After matching an empty string, it
3490 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3491 caused it to match further down the string than it should.
3492
3493 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3494 was there because it doesn't seem to be needed, and it causes trouble on some
3495 systems, as it is not a Standard C header. It has been removed.
3496
3497 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3498 were reported on the Macintosh. The changes were from
3499
3500 while ((c = *(++ptr)) != 0 && c != '\n');
3501 to
3502 while ((c = *(++ptr)) != 0 && c != '\n') ;
3503
3504 Totally extraordinary, but if that's what it takes...
3505
3506 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3507 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3508 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3509 assumes the way PCRE uses memmove() (always moving upwards).
3510
3511 6. PCRE is being used in one environment where strchr() is not available. There
3512 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3513 faster code anyway.
3514
3515
3516 Version 3.1 09-Feb-00
3517 ---------------------
3518
3519 The only change in this release is the fixing of some bugs in Makefile.in for
3520 the "install" target:
3521
3522 (1) It was failing to install pcreposix.h.
3523
3524 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3525
3526
3527 Version 3.0 01-Feb-00
3528 ---------------------
3529
3530 1. Add support for the /+ modifier to perltest (to output $` like it does in
3531 pcretest).
3532
3533 2. Add support for the /g modifier to perltest.
3534
3535 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3536 matches null strings.
3537
3538 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3539 pattern. Perl treats empty patterns specially - it reuses the most recent
3540 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3541 effect.
3542
3543 5. The POSIX interface was broken in that it was just handing over the POSIX
3544 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3545 required a bigger vector, with some working space on the end. This means that
3546 the POSIX wrapper now has to get and free some memory, and copy the results.
3547
3548 6. Added some simple autoconf support, placing the test data and the
3549 documentation in separate directories, re-organizing some of the
3550 information files, and making it build pcre-config (a GNU standard). Also added
3551 libtool support for building PCRE as a shared library, which is now the
3552 default.
3553
3554 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3555 09 are not valid octal constants. Single digits will be used for minor values
3556 less than 10.
3557
3558 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3559 existing programs that set these in the POSIX interface can use PCRE without
3560 modification.
3561
3562 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3563 return all that pcre_info() returns, plus additional data. The pcre_info()
3564 function is retained for compatibility, but is considered to be obsolete.
3565
3566 10. Added experimental recursion feature (?R) to handle one common case that
3567 Perl 5.6 will be able to do with (?p{...}).
3568
3569 11. Added support for POSIX character classes like [:alpha:], which Perl is
3570 adopting.
3571
3572
3573 Version 2.08 31-Aug-99
3574 ----------------------
3575
3576 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3577 trying to match at the startoffset position, but instead was moving forward to
3578 the next newline as if a previous match had failed.
3579
3580 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3581 and could get into a loop if a null string was matched other than at the start
3582 of the subject.
3583
3584 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3585 be distinguished at compile time, and for completeness also added PCRE_DATE.
3586
3587 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3588 in GnuWin32 environments.
3589
3590
3591 Version 2.07 29-Jul-99
3592 ----------------------
3593
3594 1. The documentation is now supplied in plain text form and HTML as well as in
3595 the form of man page sources.
3596
3597 2. C++ compilers don't like assigning (void *) values to other pointer types.
3598 In particular this affects malloc(). Although there is no problem in Standard
3599 C, I've put in casts to keep C++ compilers happy.
3600
3601 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3602 should be (const char *).
3603
3604 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3605 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3606 However, I haven't made this a standard facility. The documentation doesn't
3607 mention it, and the Makefile doesn't support it.
3608
3609 5. The Makefile now contains an "install" target, with editable destinations at
3610 the top of the file. The pcretest program is not installed.
3611
3612 6. pgrep -V now gives the PCRE version number and date.
3613
3614 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3615 causing the entire string to be ignored, instead of just the last character.
3616
3617 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3618 non-matching string, it can take a very, very long time, even for strings of
3619 quite modest length, because of the nested recursion. PCRE now does better in
3620 some of these cases. It does this by remembering the last required literal
3621 character in the pattern, and pre-searching the subject to ensure it is present
3622 before running the real match. In other words, it applies a heuristic to detect
3623 some types of certain failure quickly, and in the above example, if presented
3624 with a string that has no trailing " it gives "no match" very quickly.
3625
3626 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3627 other alternatives are tried instead.
3628
3629
3630 Version 2.06 09-Jun-99
3631 ----------------------
3632
3633 1. Change pcretest's output for amount of store used to show just the code
3634 space, because the remainder (the data block) varies in size between 32-bit and
3635 64-bit systems.
3636
3637 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3638 start matching at. This allows lookbehinds to work when searching for multiple
3639 occurrences in a string.
3640
3641 3. Added additional options to pcretest for testing multiple occurrences:
3642
3643 /+ outputs the rest of the string that follows a match
3644 /g loops for multiple occurrences, using the new startoffset argument
3645 /G loops for multiple occurrences by passing an incremented pointer
3646
3647 4. PCRE wasn't doing the "first character" optimization for patterns starting
3648 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3649 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3650 the letter 'x'. On long subject strings, this gives a significant speed-up.
3651
3652
3653 Version 2.05 21-Apr-99
3654 ----------------------
3655
3656 1. Changed the type of magic_number from int to long int so that it works
3657 properly on 16-bit systems.
3658
3659 2. Fixed a bug which caused patterns starting with .* not to work correctly
3660 when the subject string contained newline characters. PCRE was assuming
3661 anchoring for such patterns in all cases, which is not correct because .* will
3662 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3663 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3664 must be retried after every newline in the subject.
3665
3666
3667 Version 2.04 18-Feb-99
3668 ----------------------
3669
3670 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3671 computation of the store needed to hold the pattern was incorrect (too large).
3672 If such patterns were nested a few deep, this could multiply and become a real
3673 problem.
3674
3675 2. Added /M option to pcretest to show the memory requirement of a specific
3676 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3677
3678 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3679 compiled in such a way that the backtracking after subsequent failure was
3680 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3681 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3682
3683
3684 Version 2.03 02-Feb-99
3685 ----------------------
3686
3687 1. Fixed typo and small mistake in man page.
3688
3689 2. Added 4th condition (GPL supersedes if conflict) and created separate
3690 LICENCE file containing the conditions.
3691
3692 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3693 Perl, that is the internal \ allows the delimiter to be included in the
3694 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3695 the final delimiter, add \ to the end of the pattern (to test the error).
3696
3697 4. Added the convenience functions for extracting substrings after a successful
3698 match. Updated pcretest to make it able to test these functions.
3699
3700
3701 Version 2.02 14-Jan-99
3702 ----------------------
3703
3704 1. Initialized the working variables associated with each extraction so that
3705 their saving and restoring doesn't refer to uninitialized store.
3706
3707 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3708 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3709 fix the problem.
3710
3711 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3712 calls, and wasn't printing the correct value for compiling calls. Increased the
3713 default value of LOOPREPEAT, and the number of significant figures in the
3714 times.
3715
3716 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3717
3718 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3719 a building problem on Windows NT with a FAT file system.
3720
3721
3722 Version 2.01 21-Oct-98
3723 ----------------------
3724
3725 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3726 to character tables built by pcre_maketables() in the current locale. If NULL
3727 is passed, the default tables are used.
3728
3729
3730 Version 2.00 24-Sep-98
3731 ----------------------
3732
3733 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3734 it any more.
3735
3736 2. Allow quantification of (?>) groups, and make it work correctly.
3737
3738 3. The first character computation wasn't working for (?>) groups.
3739
3740 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3741 end of the subject) and add 5.005's \z, which really does match only at the
3742 very end of the subject.
3743
3744 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3745
3746 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3747 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3748 localized options. All options to pcre_study() were also removed.
3749
3750 7. Add other new features from 5.005:
3751
3752 $(?<= positive lookbehind
3753 $(?<! negative lookbehind
3754 (?imsx-imsx) added the unsetting capability
3755 such a setting is global if at outer level; local otherwise
3756 (?imsx-imsx:) non-capturing groups with option setting
3757 (?(cond)re|re) conditional pattern matching
3758
3759 A backreference to itself in a repeated group matches the previous
3760 captured string.
3761
3762 8. General tidying up of studying (both automatic and via "study")
3763 consequential on the addition of new assertions.
3764
3765 9. As in 5.005, unlimited repeated groups that could match an empty substring
3766 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3767 runtime if any iteration does actually match an empty substring.
3768
3769 10. Include the RunTest script in the distribution.
3770
3771 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3772 discrepancies, some of which were old and were also with respect to 5.004. They
3773 have now been fixed.
3774
3775
3776 Version 1.09 28-Apr-98
3777 ----------------------
3778
3779 1. A negated single character class followed by a quantifier with a minimum
3780 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3781 program crashes, or just wrong answers. This did not apply to negated classes
3782 containing more than one character, or to minima other than one.
3783
3784
3785 Version 1.08 27-Mar-98
3786 ----------------------
3787
3788 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3789
3790 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3791 latter must appear before anything that relies on it in the pattern.
3792
3793
3794 Version 1.07 16-Feb-98
3795 ----------------------
3796
3797 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3798 repeat of a potentially empty string).
3799
3800
3801 Version 1.06 23-Jan-98
3802 ----------------------
3803
3804 1. Added Markus Oberhumer's little patches for C++.
3805
3806 2. Literal strings longer than 255 characters were broken.
3807
3808
3809 Version 1.05 23-Dec-97
3810 ----------------------
3811
3812 1. Negated character classes containing more than one character were failing if
3813 PCRE_CASELESS was set at run time.
3814
3815
3816 Version 1.04 19-Dec-97
3817 ----------------------
3818
3819 1. Corrected the man page, where some "const" qualifiers had been omitted.
3820
3821 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3822 input syntax.
3823
3824 3. Fixed memory leak which occurred when a regex with back references was
3825 matched with an offsets vector that wasn't big enough. The temporary memory
3826 that is used in this case wasn't being freed if the match failed.
3827
3828 4. Tidied pcretest to ensure it frees memory that it gets.
3829
3830 5. Temporary memory was being obtained in the case where the passed offsets
3831 vector was exactly big enough.
3832
3833 6. Corrected definition of offsetof() from change 5 below.
3834
3835 7. I had screwed up change 6 below and broken the rules for the use of
3836 setjmp(). Now fixed.
3837
3838
3839 Version 1.03 18-Dec-97
3840 ----------------------
3841
3842 1. A erroneous regex with a missing opening parenthesis was correctly
3843 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3844 on some systems.
3845
3846 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3847 it was reported that one broken compiler failed on the former because "code" is
3848 also an independent variable.
3849
3850 3. The erroneous regex a[]b caused an array overrun reference.
3851
3852 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3853 fail on data ending with that character. (It was going on too far, and checking
3854 the next character, typically a binary zero.) This was specific to the
3855 optimized code for single-character negative classes.
3856
3857 5. Added a contributed patch from the TIN world which does the following:
3858
3859 + Add an undef for memmove, in case the the system defines a macro for it.
3860
3861 + Add a definition of offsetof(), in case there isn't one. (I don't know
3862 the reason behind this - offsetof() is part of the ANSI standard - but
3863 it does no harm).
3864
3865 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3866 most of the places where whitespace preceded '#'. I have given up and
3867 allowed the remaining 2 cases to be at the margin.
3868
3869 + Rename some variables in pcre to eliminate shadowing. This seems very
3870 pedantic, but does no harm, of course.
3871
3872 6. Moved the call to setjmp() into its own function, to get rid of warnings
3873 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3874
3875 7. Constructs such as \d{8,} were compiling into the equivalent of
3876 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3877 outcome, but in this particular case used more store than had been allocated,
3878 which caused the bug to be discovered because it threw up an internal error.
3879
3880 8. The debugging code in both pcre and pcretest for outputting the compiled
3881 form of a regex was going wrong in the case of back references followed by
3882 curly-bracketed repeats.
3883
3884
3885 Version 1.02 12-Dec-97
3886 ----------------------
3887
3888 1. Typos in pcre.3 and comments in the source fixed.
3889
3890 2. Applied a contributed patch to get rid of places where it used to remove
3891 'const' from variables, and fixed some signed/unsigned and uninitialized
3892 variable warnings.
3893
3894 3. Added the "runtest" target to Makefile.
3895
3896 4. Set default compiler flag to -O2 rather than just -O.
3897
3898
3899 Version 1.01 19-Nov-97
3900 ----------------------
3901
3902 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3903 like /([ab]*)*/, that is, for classes with more than one character in them.
3904
3905 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3906 as /((?>a*))*/ (a PCRE_EXTRA facility).
3907
3908
3909 Version 1.00 18-Nov-97
3910 ----------------------
3911
3912 1. Added compile-time macros to support systems such as SunOS4 which don't have
3913 memmove() or strerror() but have other things that can be used instead.
3914
3915 2. Arranged that "make clean" removes the executables.
3916
3917
3918 Version 0.99 27-Oct-97
3919 ----------------------
3920
3921 1. Fixed bug in code for optimizing classes with only one character. It was
3922 initializing a 32-byte map regardless, which could cause it to run off the end
3923 of the memory it had got.
3924
3925 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3926
3927
3928 Version 0.98 22-Oct-97
3929 ----------------------
3930
3931 1. Fixed bug in code for handling temporary memory usage when there are more
3932 back references than supplied space in the ovector. This could cause segfaults.
3933
3934
3935 Version 0.97 21-Oct-97
3936 ----------------------
3937
3938 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3939
3940 2. Optimized negated single characters not to use a bit map.
3941
3942 3. Brought error texts together as macro definitions; clarified some of them;
3943 fixed one that was wrong - it said "range out of order" when it meant "invalid
3944 escape sequence".
3945
3946 4. Changed some char * arguments to const char *.
3947
3948 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3949
3950 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3951 pcretest.
3952
3953
3954 Version 0.96 16-Oct-97
3955 ----------------------
3956
3957 1. Added a simple "pgrep" utility to the distribution.
3958
3959 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3960 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3961 where "ddd" means "one or more decimal digits".
3962
3963 3. Fixed serious bug. If a pattern had a back reference, but the call to
3964 pcre_exec() didn't supply a large enough ovector to record the related
3965 identifying subpattern, the match always failed. PCRE now remembers the number
3966 of the largest back reference, and gets some temporary memory in which to save
3967 the offsets during matching if necessary, in order to ensure that
3968 backreferences always work.
3969
3970 4. Increased the compatibility with Perl in a number of ways:
3971
3972 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3973 to request this handling. The option can be set at compile or exec time.
3974
3975 (b) $ matches before a terminating newline by default; an option
3976 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3977 mode). The option can be set at compile or exec time.
3978
3979 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3980 the same as Perl's. If the decimal number it represents is less than 10
3981 or there aren't that many previous left capturing parentheses, an octal
3982 escape is read. Inside a character class, it's always an octal escape,
3983 even if it is a single digit.
3984
3985 (d) An escaped but undefined alphabetic character is taken as a literal,
3986 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3987 escapes.
3988
3989 (e) {0} is now permitted. (The previous item is removed from the compiled
3990 pattern).
3991
3992 5. Changed all the names of code files so that the basic parts are no longer
3993 than 10 characters, and abolished the teeny "globals.c" file.
3994
3995 6. Changed the handling of character classes; they are now done with a 32-byte
3996 bit map always.
3997
3998 7. Added the -d and /D options to pcretest to make it possible to look at the
3999 internals of compilation without having to recompile pcre.
4000
4001
4002 Version 0.95 23-Sep-97
4003 ----------------------
4004
4005 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4006 \x20 at the start of a run of normal characters. These were being treated as
4007 real characters, instead of the source characters being re-checked.
4008
4009
4010 Version 0.94 18-Sep-97
4011 ----------------------
4012
4013 1. The functions are now thread-safe, with the caveat that the global variables
4014 containing pointers to malloc() and free() or alternative functions are the
4015 same for all threads.
4016
4017 2. Get pcre_study() to generate a bitmap of initial characters for non-
4018 anchored patterns when this is possible, and use it if passed to pcre_exec().
4019
4020
4021 Version 0.93 15-Sep-97
4022 ----------------------
4023
4024 1. /(b)|(:+)/ was computing an incorrect first character.
4025
4026 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4027 but not actually doing anything yet.
4028
4029 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4030 as Perl does (e.g. [-az] or [az-]).
4031
4032 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4033 all possible positions.
4034
4035 5. Split up into different modules to avoid including unneeded functions in a
4036 compiled binary. However, compile and exec are still in one module. The "study"
4037 function is split off.
4038
4039 6. The character tables are now in a separate module whose source is generated
4040 by an auxiliary program - but can then be edited by hand if required. There are
4041 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4042 toupper() in the code.
4043
4044 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4045 make them global. Abolish the function for setting them, as the caller can now
4046 set them directly.
4047
4048
4049 Version 0.92 11-Sep-97
4050 ----------------------
4051
4052 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4053 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4054
4055 2. Caseless matching was not working in character classes if the characters in
4056 the pattern were in upper case.
4057
4058 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4059
4060 4. Make PCRE_ANCHORED public and accept as a compile option.
4061
4062 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4063 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4064 pass them.
4065
4066 6. Give an error if bad option bits passed at compile or run time.
4067
4068 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4069 pcretest to cause it to pass that flag.
4070
4071 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4072 options, and the first character, if set.
4073
4074 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4075
4076
4077 Version 0.91 10-Sep-97
4078 ----------------------
4079
4080 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4081 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4082
4083 2. PCRE was looping on encountering an indefinitely repeated back reference to
4084 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4085 Perl does - treats the match as successful.
4086
4087 ****

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