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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 611 - (show annotations) (download)
Wed Jun 29 08:49:21 2011 UTC (3 years, 5 months ago) by ph10
File size: 203842 byte(s)
Fix \R problem with study: incorrect minimum subject length.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.13 30-Apr-2011
5 ------------------------
6
7 1. The Unicode data tables have been updated to Unicode 6.0.0.
8
9 2. Two minor typos in pcre_internal.h have been fixed.
10
11 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
12 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
13 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
14
15 4. There were a number of related bugs in the code for matching backrefences
16 caselessly in UTF-8 mode when codes for the characters concerned were
17 different numbers of bytes. For example, U+023A and U+2C65 are an upper
18 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
19 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
20 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
21 2-byte code at the end of the subject (it thought there wasn't enough data
22 left).
23
24 5. Comprehensive information about what went wrong is now returned by
25 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
26 as the output vector has at least 2 elements. The offset of the start of
27 the failing character and a reason code are placed in the vector.
28
29 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
30 now returned is for the first byte of the failing character, instead of the
31 last byte inspected. This is an incompatible change, but I hope it is small
32 enough not to be a problem. It makes the returned offset consistent with
33 pcre_exec() and pcre_dfa_exec().
34
35 7. pcretest now gives a text phrase as well as the error number when
36 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
37 failure, the offset and reason code are output.
38
39 8. When \R was used with a maximizing quantifier it failed to skip backwards
40 over a \r\n pair if the subsequent match failed. Instead, it just skipped
41 back over a single character (\n). This seems wrong (because it treated the
42 two characters as a single entity when going forwards), conflicts with the
43 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
44 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
45 has been changed.
46
47 9. Some internal refactoring has changed the processing so that the handling
48 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
49 time (the PCRE_DOTALL option was changed this way some time ago: version
50 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
51 which was always a bit of a fudge. It also means that there is one less
52 argument for the match() function, which reduces its stack requirements
53 slightly. This change also fixes an incompatibility with Perl: the pattern
54 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
55
56 10. More internal refactoring has drastically reduced the number of recursive
57 calls to match() for possessively repeated groups such as (abc)++ when
58 using pcre_exec().
59
60 11. While implementing 10, a number of bugs in the handling of groups were
61 discovered and fixed:
62
63 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
64 (a|)*(?1) gave a compile-time internal error.
65 ((a|)+)+ did not notice that the outer group could match an empty string.
66 (^a|^)+ was not marked as anchored.
67 (.*a|.*)+ was not marked as matching at start or after a newline.
68
69 12. Yet more internal refactoring has removed another argument from the match()
70 function. Special calls to this function are now indicated by setting a
71 value in a variable in the "match data" data block.
72
73 13. Be more explicit in pcre_study() instead of relying on "default" for
74 opcodes that mean there is no starting character; this means that when new
75 ones are added and accidentally left out of pcre_study(), testing should
76 pick them up.
77
78 14. The -s option of pcretest has been documented for ages as being an old
79 synonym of -m (show memory usage). I have changed it to mean "force study
80 for every regex", that is, assume /S for every regex. This is similar to -i
81 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
82 using it. It makes it easier to run collection of tests with study enabled,
83 and thereby test pcre_study() more easily.
84
85 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
86 restoration of the capturing data to the outer values was not happening
87 correctly.
88
89 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
90 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
91 pattern had matched an empty string, and so incorrectly returned a no
92 match.
93
94 17. There was optimizing code for the last branch of non-capturing parentheses,
95 and also for the obeyed branch of a conditional subexpression, which used
96 tail recursion to cut down on stack usage. Unfortunately, not that there is
97 the possibility of (*THEN) occurring in these branches, tail recursion is
98 no longer possible because the return has to be checked for (*THEN). These
99 two optimizations have therefore been removed.
100
101 18. If a pattern containing \R was studied, it was assumed that \R always
102 matched two bytes, thus causing the minimum subject length to be
103 incorrectly computed because \R can also match just one byte.
104
105
106 Version 8.12 15-Jan-2011
107 ------------------------
108
109 1. Fixed some typos in the markup of the man pages, and wrote a script that
110 checks for such things as part of the documentation building process.
111
112 2. On a big-endian 64-bit system, pcregrep did not correctly process the
113 --match-limit and --recursion-limit options (added for 8.11). In
114 particular, this made one of the standard tests fail. (The integer value
115 went into the wrong half of a long int.)
116
117 3. If the --colour option was given to pcregrep with -v (invert match), it
118 did strange things, either producing crazy output, or crashing. It should,
119 of course, ignore a request for colour when reporting lines that do not
120 match.
121
122 4. Another pcregrep bug caused similar problems if --colour was specified with
123 -M (multiline) and the pattern match finished with a line ending.
124
125 5. In pcregrep, when a pattern that ended with a literal newline sequence was
126 matched in multiline mode, the following line was shown as part of the
127 match. This seems wrong, so I have changed it.
128
129 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
130 the check for further matches in the same line (so they could be coloured)
131 to overrun the end of the current line. If another match was found, it was
132 incorrectly shown (and then shown again when found in the next line).
133
134 7. If pcregrep was compiled under Windows, there was a reference to the
135 function pcregrep_exit() before it was defined. I am assuming this was
136 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
137 reported by a user. I've moved the definition above the reference.
138
139
140 Version 8.11 10-Dec-2010
141 ------------------------
142
143 1. (*THEN) was not working properly if there were untried alternatives prior
144 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
145 backtracked to try for "b" instead of moving to the next alternative branch
146 at the same level (in this case, to look for "c"). The Perl documentation
147 is clear that when (*THEN) is backtracked onto, it goes to the "next
148 alternative in the innermost enclosing group".
149
150 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
151 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
152 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
153 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
154 (*THEN).
155
156 3. If \s appeared in a character class, it removed the VT character from
157 the class, even if it had been included by some previous item, for example
158 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
159 of \s, but is part of the POSIX "space" class.)
160
161 4. A partial match never returns an empty string (because you can always
162 match an empty string at the end of the subject); however the checking for
163 an empty string was starting at the "start of match" point. This has been
164 changed to the "earliest inspected character" point, because the returned
165 data for a partial match starts at this character. This means that, for
166 example, /(?<=abc)def/ gives a partial match for the subject "abc"
167 (previously it gave "no match").
168
169 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
170 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
171 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
172 has an implication that the given string is incomplete (because a partial
173 match is preferred over a full match). For this reason, these items now
174 give a partial match in this situation. [Aside: previously, the one case
175 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
176 match rather than a full match, which was wrong by the old rules, but is
177 now correct.]
178
179 6. There was a bug in the handling of #-introduced comments, recognized when
180 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
181 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
182 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
183 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
184 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
185 places in pcre_compile().
186
187 7. Related to (6) above, when pcre_compile() was skipping #-introduced
188 comments when looking ahead for named forward references to subpatterns,
189 the only newline sequence it recognized was NL. It now handles newlines
190 according to the set newline convention.
191
192 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
193 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
194 cater for a lack of strerror(). These oversights have been fixed.
195
196 9. Added --match-limit and --recursion-limit to pcregrep.
197
198 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
199
200 11. When the -o option was used, pcregrep was setting a return code of 1, even
201 when matches were found, and --line-buffered was not being honoured.
202
203 12. Added an optional parentheses number to the -o and --only-matching options
204 of pcregrep.
205
206 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
207 can match an empty string. The code to do it in pcretest and pcredemo
208 needed fixing:
209
210 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
211 only one byte after an empty string match just before CRLF (this case
212 just got forgotten; "any" and "anycrlf" were OK).
213
214 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
215 mode when an empty string match preceded an ASCII character followed by
216 a non-ASCII character. (The code for advancing by one character rather
217 than one byte was nonsense.)
218
219 (c) The pcredemo.c sample program did not have any code at all to handle
220 the cases when CRLF is a valid newline sequence.
221
222 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
223 as a starting offset was within the subject string. There is now a new
224 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
225 negative or greater than the length of the string. In order to test this,
226 pcretest is extended to allow the setting of negative starting offsets.
227
228 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
229 starting offset points to the beginning of a UTF-8 character was
230 unnecessarily clumsy. I tidied it up.
231
232 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
233 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
234
235 17. Nobody had reported that the --include_dir option, which was added in
236 release 7.7 should have been called --include-dir (hyphen, not underscore)
237 for compatibility with GNU grep. I have changed it to --include-dir, but
238 left --include_dir as an undocumented synonym, and the same for
239 --exclude-dir, though that is not available in GNU grep, at least as of
240 release 2.5.4.
241
242 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
243 characters from a string of bytes) have been redefined so as not to use
244 loops, in order to improve performance in some environments. At the same
245 time, I abstracted some of the common code into auxiliary macros to save
246 repetition (this should not affect the compiled code).
247
248 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
249 compile-time error is now given if \c is not followed by an ASCII
250 character, that is, a byte less than 128. (In EBCDIC mode, the code is
251 different, and any byte value is allowed.)
252
253 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
254 START_OPTIMIZE option, which is now allowed at compile time - but just
255 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
256 to pcregrep and other applications that have no direct access to PCRE
257 options. The new /Y option in pcretest sets this option when calling
258 pcre_compile().
259
260 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
261 back references. Groups containing recursive back references were forced to
262 be atomic by that change, but in the case of named groups, the amount of
263 memory required was incorrectly computed, leading to "Failed: internal
264 error: code overflow". This has been fixed.
265
266 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
267 pcretest.c, to avoid build problems in some Borland environments.
268
269
270 Version 8.10 25-Jun-2010
271 ------------------------
272
273 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
274 THEN.
275
276 2. (*ACCEPT) was not working when inside an atomic group.
277
278 3. Inside a character class, \B is treated as a literal by default, but
279 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
280 causes the error). The code is unchanged, but I tidied the documentation.
281
282 4. Inside a character class, PCRE always treated \R and \X as literals,
283 whereas Perl faults them if its -w option is set. I have changed PCRE so
284 that it faults them when PCRE_EXTRA is set.
285
286 5. Added support for \N, which always matches any character other than
287 newline. (It is the same as "." when PCRE_DOTALL is not set.)
288
289 6. When compiling pcregrep with newer versions of gcc which may have
290 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
291 declared with attribute warn_unused_result" were given. Just casting the
292 result to (void) does not stop the warnings; a more elaborate fudge is
293 needed. I've used a macro to implement this.
294
295 7. Minor change to pcretest.c to avoid a compiler warning.
296
297 8. Added four artifical Unicode properties to help with an option to make
298 \s etc use properties (see next item). The new properties are: Xan
299 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
300
301 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
302 use Unicode properties. (*UCP) at the start of a pattern can be used to set
303 this option. Modified pcretest to add /W to test this facility. Added
304 REG_UCP to make it available via the POSIX interface.
305
306 10. Added --line-buffered to pcregrep.
307
308 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
309 studied, and the match started with a letter with a code point greater than
310 127 whose first byte was different to the first byte of the other case of
311 the letter, the other case of this starting letter was not recognized
312 (#976).
313
314 12. If a pattern that was studied started with a repeated Unicode property
315 test, for example, \p{Nd}+, there was the theoretical possibility of
316 setting up an incorrect bitmap of starting bytes, but fortunately it could
317 not have actually happened in practice until change 8 above was made (it
318 added property types that matched character-matching opcodes).
319
320 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
321 possible starting bytes for non-anchored patterns.
322
323 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
324 \R, and also a number of cases that involve Unicode properties, both
325 explicit and implicit when PCRE_UCP is set.
326
327 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
328 input, it could crash or give wrong results if characters with values
329 greater than 0xc0 were present in the subject string. (Detail: it assumed
330 UTF-8 input when processing these items.)
331
332 16. Added a lot of (int) casts to avoid compiler warnings in systems where
333 size_t is 64-bit (#991).
334
335 17. Added a check for running out of memory when PCRE is compiled with
336 --disable-stack-for-recursion (#990).
337
338 18. If the last data line in a file for pcretest does not have a newline on
339 the end, a newline was missing in the output.
340
341 19. The default pcre_chartables.c file recognizes only ASCII characters (values
342 less than 128) in its various bitmaps. However, there is a facility for
343 generating tables according to the current locale when PCRE is compiled. It
344 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
345 characters, are recognized by isspace() and therefore were getting set in
346 these tables, and indeed these tables seem to approximate to ISO 8859. This
347 caused a problem in UTF-8 mode when pcre_study() was used to create a list
348 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
349 which of course cannot start UTF-8 characters. I have changed the code so
350 that only real ASCII characters (less than 128) and the correct starting
351 bytes for UTF-8 encodings are set for characters greater than 127 when in
352 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
353 altogether.)
354
355 20. Added the /T option to pcretest so as to be able to run tests with non-
356 standard character tables, thus making it possible to include the tests
357 used for 19 above in the standard set of tests.
358
359 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
360 reference to a subpattern the other side of a comment that contains an
361 opening parenthesis caused either an internal compiling error, or a
362 reference to the wrong subpattern.
363
364
365 Version 8.02 19-Mar-2010
366 ------------------------
367
368 1. The Unicode data tables have been updated to Unicode 5.2.0.
369
370 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
371 configured.
372
373 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
374 original author of that file, following a query about its status.
375
376 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
377 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
378
379 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
380 quantifier applied to a forward-referencing subroutine call, could compile
381 incorrect code or give the error "internal error: previously-checked
382 referenced subpattern not found".
383
384 6. Both MS Visual Studio and Symbian OS have problems with initializing
385 variables to point to external functions. For these systems, therefore,
386 pcre_malloc etc. are now initialized to local functions that call the
387 relevant global functions.
388
389 7. There were two entries missing in the vectors called coptable and poptable
390 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
391 I've fixed the data, and added a kludgy way of testing at compile time that
392 the lengths are correct (equal to the number of opcodes).
393
394 8. Following on from 7, I added a similar kludge to check the length of the
395 eint vector in pcreposix.c.
396
397 9. Error texts for pcre_compile() are held as one long string to avoid too
398 much relocation at load time. To find a text, the string is searched,
399 counting zeros. There was no check for running off the end of the string,
400 which could happen if a new error number was added without updating the
401 string.
402
403 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
404
405 11. \K was not working if it appeared in an atomic group or in a group that
406 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
407 \K is "not well defined" if used in an assertion. PCRE now accepts it if
408 the assertion is positive, but not if it is negative.
409
410 12. Change 11 fortuitously reduced the size of the stack frame used in the
411 "match()" function of pcre_exec.c by one pointer. Forthcoming
412 implementation of support for (*MARK) will need an extra pointer on the
413 stack; I have reserved it now, so that the stack frame size does not
414 decrease.
415
416 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
417 item in branch that calls a recursion is a subroutine call - as in the
418 second branch in the above example - was incorrectly given the compile-
419 time error "recursive call could loop indefinitely" because pcre_compile()
420 was not correctly checking the subroutine for matching a non-empty string.
421
422 14. The checks for overrunning compiling workspace could trigger after an
423 overrun had occurred. This is a "should never occur" error, but it can be
424 triggered by pathological patterns such as hundreds of nested parentheses.
425 The checks now trigger 100 bytes before the end of the workspace.
426
427 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
428
429
430 Version 8.01 19-Jan-2010
431 ------------------------
432
433 1. If a pattern contained a conditional subpattern with only one branch (in
434 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
435 computed the wrong minimum data length (which is of course zero for such
436 subpatterns). This could cause incorrect "no match" results.
437
438 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
439 the pattern is reset in the first branch, pcre_compile() failed with
440 "internal error: code overflow at offset...". This happened only when
441 the reset was to the original external option setting. (An optimization
442 abstracts leading options settings into an external setting, which was the
443 cause of this.)
444
445 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
446 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
447 assertion pattern did not match (meaning that the assertion was true), it
448 was incorrectly treated as false if the SKIP had been reached during the
449 matching. This also applied to assertions used as conditions.
450
451 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
452 assertion subpattern, including such a pattern used as a condition,
453 unpredictable results occurred, instead of the error return
454 PCRE_ERROR_DFA_UITEM.
455
456 5. The C++ GlobalReplace function was not working like Perl for the special
457 situation when an empty string is matched. It now does the fancy magic
458 stuff that is necessary.
459
460 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
461 removed. (These were left over from very, very early versions of PCRE.)
462
463 7. Some cosmetic changes to the code to make life easier when compiling it
464 as part of something else:
465
466 (a) Change DEBUG to PCRE_DEBUG.
467
468 (b) In pcre_compile(), rename the member of the "branch_chain" structure
469 called "current" as "current_branch", to prevent a collision with the
470 Linux macro when compiled as a kernel module.
471
472 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
473 prevent a collision with the Linux macro when compiled as a kernel
474 module.
475
476 8. In pcre_compile() there are some checks for integer overflows that used to
477 cast potentially large values to (double). This has been changed to that
478 when building, a check for int64_t is made, and if it is found, it is used
479 instead, thus avoiding the use of floating point arithmetic. (There is no
480 other use of FP in PCRE.) If int64_t is not found, the fallback is to
481 double.
482
483 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
484 2005 (difference between two addresses compared to an unsigned value).
485
486 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
487 custom one, because of the following reported problem in Windows:
488
489 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
490 under Win32.
491 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
492 therefore missing the function definition.
493 - The compiler thus generates a "C" signature for the test function.
494 - The linker fails to find the "C" function.
495 - PCRE fails to configure if asked to do so against libbz2.
496
497 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
498 messages were output:
499
500 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
501 rerunning libtoolize, to keep the correct libtool macros in-tree.
502 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
503
504 I have done both of these things.
505
506 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
507 most of the time, it *can* run out if it is given a pattern that contains a
508 runaway infinite recursion. I updated the discussion in the pcrestack man
509 page.
510
511 13. Now that we have gone to the x.xx style of version numbers, the minor
512 version may start with zero. Using 08 or 09 is a bad idea because users
513 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
514 interpreted as invalid octal numbers. I've updated the previous comment in
515 configure.ac, and also added a check that gives an error if 08 or 09 are
516 used.
517
518 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
519 causing partial matching to fail when the end of the subject matched \W
520 in a UTF-8 pattern where \W was quantified with a minimum of 3.
521
522 15. There were some discrepancies between the declarations in pcre_internal.h
523 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
524 their definitions. The declarations used "const uschar *" and the
525 definitions used USPTR. Even though USPTR is normally defined as "const
526 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
527 reported that: "This difference in casting confuses some C++ compilers, for
528 example, SunCC recognizes above declarations as different functions and
529 generates broken code for hbpcre." I have changed the declarations to use
530 USPTR.
531
532 16. GNU libtool is named differently on some systems. The autogen.sh script now
533 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
534 (FreeBSD).
535
536 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
537 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
538 comment: "Figure out how to create a longlong from a string: strtoll and
539 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
540 instance, but it only takes 2 args instead of 3!"
541
542 18. A subtle bug concerned with back references has been fixed by a change of
543 specification, with a corresponding code fix. A pattern such as
544 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
545 refers, was giving matches when it shouldn't. For example, xa=xaaa would
546 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
547 same bug. Such groups have to be quantified to be useful, or contained
548 inside another quantified group. (If there's no repetition, the reference
549 can never match.) The problem arises because, having left the group and
550 moved on to the rest of the pattern, a later failure that backtracks into
551 the group uses the captured value from the final iteration of the group
552 rather than the correct earlier one. I have fixed this in PCRE by forcing
553 any group that contains a reference to itself to be an atomic group; that
554 is, there cannot be any backtracking into it once it has completed. This is
555 similar to recursive and subroutine calls.
556
557
558 Version 8.00 19-Oct-09
559 ----------------------
560
561 1. The table for translating pcre_compile() error codes into POSIX error codes
562 was out-of-date, and there was no check on the pcre_compile() error code
563 being within the table. This could lead to an OK return being given in
564 error.
565
566 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
567 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
568 in a Windows environment.
569
570 3. The pcregrep --count option prints the count for each file even when it is
571 zero, as does GNU grep. However, pcregrep was also printing all files when
572 --files-with-matches was added. Now, when both options are given, it prints
573 counts only for those files that have at least one match. (GNU grep just
574 prints the file name in this circumstance, but including the count seems
575 more useful - otherwise, why use --count?) Also ensured that the
576 combination -clh just lists non-zero counts, with no names.
577
578 4. The long form of the pcregrep -F option was incorrectly implemented as
579 --fixed_strings instead of --fixed-strings. This is an incompatible change,
580 but it seems right to fix it, and I didn't think it was worth preserving
581 the old behaviour.
582
583 5. The command line items --regex=pattern and --regexp=pattern were not
584 recognized by pcregrep, which required --regex pattern or --regexp pattern
585 (with a space rather than an '='). The man page documented the '=' forms,
586 which are compatible with GNU grep; these now work.
587
588 6. No libpcreposix.pc file was created for pkg-config; there was just
589 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
590
591 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
592 when UCP support is not needed, by modifying the Python script that
593 generates it from Unicode data files. This should not matter if the module
594 is correctly used as a library, but I received one complaint about 50K of
595 unwanted data. My guess is that the person linked everything into his
596 program rather than using a library. Anyway, it does no harm.
597
598 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
599 was a minimum greater than 1 for a wide character in a possessive
600 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
601 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
602 character. Chaos in the form of incorrect output or a compiling loop could
603 result.
604
605 9. The restrictions on what a pattern can contain when partial matching is
606 requested for pcre_exec() have been removed. All patterns can now be
607 partially matched by this function. In addition, if there are at least two
608 slots in the offset vector, the offset of the earliest inspected character
609 for the match and the offset of the end of the subject are set in them when
610 PCRE_ERROR_PARTIAL is returned.
611
612 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
613 synonymous with PCRE_PARTIAL, for backwards compatibility, and
614 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
615 and may be more useful for multi-segment matching.
616
617 11. Partial matching with pcre_exec() is now more intuitive. A partial match
618 used to be given if ever the end of the subject was reached; now it is
619 given only if matching could not proceed because another character was
620 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
621 string "Z", which now yields "no match" instead of "partial match". In the
622 case of pcre_dfa_exec(), "no match" is given if every matching path for the
623 final character ended with (*FAIL).
624
625 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
626 if the pattern had a "must contain" character that was already found in the
627 earlier partial match, unless partial matching was again requested. For
628 example, with the pattern /dog.(body)?/, the "must contain" character is
629 "g". If the first part-match was for the string "dog", restarting with
630 "sbody" failed. This bug has been fixed.
631
632 13. The string returned by pcre_dfa_exec() after a partial match has been
633 changed so that it starts at the first inspected character rather than the
634 first character of the match. This makes a difference only if the pattern
635 starts with a lookbehind assertion or \b or \B (\K is not supported by
636 pcre_dfa_exec()). It's an incompatible change, but it makes the two
637 matching functions compatible, and I think it's the right thing to do.
638
639 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
640 so that the demonstration program is easily available in environments where
641 PCRE has not been installed from source.
642
643 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
644 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
645 library.
646
647 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
648 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
649 is not the first non-POSIX option to be added. Clearly some people find
650 these options useful.
651
652 17. If a caller to the POSIX matching function regexec() passes a non-zero
653 value for nmatch with a NULL value for pmatch, the value of
654 nmatch is forced to zero.
655
656 18. RunGrepTest did not have a test for the availability of the -u option of
657 the diff command, as RunTest does. It now checks in the same way as
658 RunTest, and also checks for the -b option.
659
660 19. If an odd number of negated classes containing just a single character
661 interposed, within parentheses, between a forward reference to a named
662 subpattern and the definition of the subpattern, compilation crashed with
663 an internal error, complaining that it could not find the referenced
664 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
665 [The bug was that it was starting one character too far in when skipping
666 over the character class, thus treating the ] as data rather than
667 terminating the class. This meant it could skip too much.]
668
669 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
670 /g option in pcretest when the pattern contains \K, which makes it possible
671 to have an empty string match not at the start, even when the pattern is
672 anchored. Updated pcretest and pcredemo to use this option.
673
674 21. If the maximum number of capturing subpatterns in a recursion was greater
675 than the maximum at the outer level, the higher number was returned, but
676 with unset values at the outer level. The correct (outer level) value is
677 now given.
678
679 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
680 PCRE did not set those parentheses (unlike Perl). I have now found a way to
681 make it do so. The string so far is captured, making this feature
682 compatible with Perl.
683
684 23. The tests have been re-organized, adding tests 11 and 12, to make it
685 possible to check the Perl 5.10 features against Perl 5.10.
686
687 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
688 pattern matches a fixed length string. PCRE did not allow this; now it
689 does. Neither allows recursion.
690
691 25. I finally figured out how to implement a request to provide the minimum
692 length of subject string that was needed in order to match a given pattern.
693 (It was back references and recursion that I had previously got hung up
694 on.) This code has now been added to pcre_study(); it finds a lower bound
695 to the length of subject needed. It is not necessarily the greatest lower
696 bound, but using it to avoid searching strings that are too short does give
697 some useful speed-ups. The value is available to calling programs via
698 pcre_fullinfo().
699
700 26. While implementing 25, I discovered to my embarrassment that pcretest had
701 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
702 study optimizations had never been tested with that matching function.
703 Oops. What is worse, even when it was passed study data, there was a bug in
704 pcre_dfa_exec() that meant it never actually used it. Double oops. There
705 were also very few tests of studied patterns with pcre_dfa_exec().
706
707 27. If (?| is used to create subpatterns with duplicate numbers, they are now
708 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
709 on the other side of the coin, they are no longer allowed to have different
710 names, because these cannot be distinguished in PCRE, and this has caused
711 confusion. (This is a difference from Perl.)
712
713 28. When duplicate subpattern names are present (necessarily with different
714 numbers, as required by 27 above), and a test is made by name in a
715 conditional pattern, either for a subpattern having been matched, or for
716 recursion in such a pattern, all the associated numbered subpatterns are
717 tested, and the overall condition is true if the condition is true for any
718 one of them. This is the way Perl works, and is also more like the way
719 testing by number works.
720
721
722 Version 7.9 11-Apr-09
723 ---------------------
724
725 1. When building with support for bzlib/zlib (pcregrep) and/or readline
726 (pcretest), all targets were linked against these libraries. This included
727 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
728 libraries. This caused unwanted dependencies to be created. This problem
729 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
730 pcretest is linked with readline.
731
732 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
733 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
734 moved outside it again, because FALSE and TRUE are already defined in AIX,
735 but BOOL is not.
736
737 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
738 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
739
740 4. The pcregrep documentation said spaces were inserted as well as colons (or
741 hyphens) following file names and line numbers when outputting matching
742 lines. This is not true; no spaces are inserted. I have also clarified the
743 wording for the --colour (or --color) option.
744
745 5. In pcregrep, when --colour was used with -o, the list of matching strings
746 was not coloured; this is different to GNU grep, so I have changed it to be
747 the same.
748
749 6. When --colo(u)r was used in pcregrep, only the first matching substring in
750 each matching line was coloured. Now it goes on to look for further matches
751 of any of the test patterns, which is the same behaviour as GNU grep.
752
753 7. A pattern that could match an empty string could cause pcregrep to loop; it
754 doesn't make sense to accept an empty string match in pcregrep, so I have
755 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
756 seems to be how GNU grep behaves.
757
758 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
759 start or after a newline", because the conditional assertion was not being
760 correctly handled. The rule now is that both the assertion and what follows
761 in the first alternative must satisfy the test.
762
763 9. If auto-callout was enabled in a pattern with a conditional group whose
764 condition was an assertion, PCRE could crash during matching, both with
765 pcre_exec() and pcre_dfa_exec().
766
767 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
768 used for matching.
769
770 11. Unicode property support in character classes was not working for
771 characters (bytes) greater than 127 when not in UTF-8 mode.
772
773 12. Added the -M command line option to pcretest.
774
775 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
776
777 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
778
779 16. Added comments and documentation about mis-use of no_arg in the C++
780 wrapper.
781
782 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
783 from Martin Jerabek that uses macro names for all relevant character and
784 string constants.
785
786 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
787 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
788 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
789 these, but not everybody uses configure.
790
791 19. A conditional group that had only one branch was not being correctly
792 recognized as an item that could match an empty string. This meant that an
793 enclosing group might also not be so recognized, causing infinite looping
794 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
795 with the subject "ab", where knowledge that the repeated group can match
796 nothing is needed in order to break the loop.
797
798 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
799 exec(), but without supplying a callout function, matching went wrong.
800
801 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
802 leak if the size of the offset vector was greater than 30. When the vector
803 is smaller, the saved offsets during recursion go onto a local stack
804 vector, but for larger vectors malloc() is used. It was failing to free
805 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
806 error, in fact).
807
808 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
809 heapframe that is used only when UTF-8 support is enabled. This caused no
810 problem, but was untidy.
811
812 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
813 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
814 included within another project.
815
816 24. Steven Van Ingelgem's patches to add more options to the CMake support,
817 slightly modified by me:
818
819 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
820 not building pcregrep.
821
822 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
823 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
824
825 25. Forward references, both numeric and by name, in patterns that made use of
826 duplicate group numbers, could behave incorrectly or give incorrect errors,
827 because when scanning forward to find the reference group, PCRE was not
828 taking into account the duplicate group numbers. A pattern such as
829 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
830
831 26. Changed a few more instances of "const unsigned char *" to USPTR, making
832 the feature of a custom pointer more persuasive (as requested by a user).
833
834 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
835 pcretest.c, inside #ifndefs, because it seems they are sometimes already
836 pre-defined.
837
838 28. Added support for (*UTF8) at the start of a pattern.
839
840 29. Arrange for flags added by the "release type" setting in CMake to be shown
841 in the configuration summary.
842
843
844 Version 7.8 05-Sep-08
845 ---------------------
846
847 1. Replaced UCP searching code with optimized version as implemented for Ad
848 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
849 stage table and inline lookup instead of a function, giving speed ups of 2
850 to 5 times on some simple patterns that I tested. Permission was given to
851 distribute the MultiStage2.py script that generates the tables (it's not in
852 the tarball, but is in the Subversion repository).
853
854 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
855 scripts.
856
857 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
858 a group with a zero qualifier. The result of the study could be incorrect,
859 or the function might crash, depending on the pattern.
860
861 4. Caseless matching was not working for non-ASCII characters in back
862 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
863 It now works when Unicode Property Support is available.
864
865 5. In pcretest, an escape such as \x{de} in the data was always generating
866 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
867 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
868 truncation.
869
870 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
871
872 7. Added two (int) casts to pcregrep when printing the difference of two
873 pointers, in case they are 64-bit values.
874
875 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
876 test 2 if it fails.
877
878 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
879 and a #define of that name to empty if it is not externally set. This is to
880 allow users of MSVC to set it if necessary.
881
882 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
883 the convenience functions in the pcre_get.c source file.
884
885 11. An option change at the start of a pattern that had top-level alternatives
886 could cause overwriting and/or a crash. This command provoked a crash in
887 some environments:
888
889 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
890
891 This potential security problem was recorded as CVE-2008-2371.
892
893 12. For a pattern where the match had to start at the beginning or immediately
894 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
895 pcre_dfa_exec() could read past the end of the passed subject if there was
896 no match. To help with detecting such bugs (e.g. with valgrind), I modified
897 pcretest so that it places the subject at the end of its malloc-ed buffer.
898
899 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
900 exec() might read past the end of the data buffer in UTF-8 mode.
901
902 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
903 the data contained the byte 0x85 as part of a UTF-8 character within its
904 first line. This applied both to normal and DFA matching.
905
906 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
907 /^[^d]*?$/8 failed to match "abc".
908
909 16. Added a missing copyright notice to pcrecpp_internal.h.
910
911 17. Make it more clear in the documentation that values returned from
912 pcre_exec() in ovector are byte offsets, not character counts.
913
914 18. Tidied a few places to stop certain compilers from issuing warnings.
915
916 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
917 supplied by Stefan Weber. I made a further small update for 7.8 because
918 there is a change of source arrangements: the pcre_searchfuncs.c module is
919 replaced by pcre_ucd.c.
920
921
922 Version 7.7 07-May-08
923 ---------------------
924
925 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
926 a string to a long long, pretend we don't even have a long long." This is
927 done by checking for the strtoq, strtoll, and _strtoi64 functions.
928
929 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
930 pre-7.6 versions, which defined a global no_arg variable instead of putting
931 it in the RE class. (See also #8 below.)
932
933 3. Remove a line of dead code, identified by coverity and reported by Nuno
934 Lopes.
935
936 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
937
938 (1) The include/exclude patterns were being applied to the whole pathnames
939 of files, instead of just to the final components.
940
941 (2) If there was more than one level of directory, the subdirectories were
942 skipped unless they satisfied the include/exclude conditions. This is
943 inconsistent with GNU grep (and could even be seen as contrary to the
944 pcregrep specification - which I improved to make it absolutely clear).
945 The action now is always to scan all levels of directory, and just
946 apply the include/exclude patterns to regular files.
947
948 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
949 --exclude_dir in the tests to avoid scanning .svn directories.
950
951 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
952 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
953 doesn't support NULs in patterns.
954
955 7. Added some missing "const"s to declarations of static tables in
956 pcre_compile.c and pcre_dfa_exec.c.
957
958 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
959 caused by fix #2 above. (Subsequently also a second patch to fix the
960 first patch. And a third patch - this was a messy problem.)
961
962 9. Applied Craig's patch to remove the use of push_back().
963
964 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
965 matching function regexec().
966
967 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
968 which, however, unlike Perl's \g{...}, are subroutine calls, not back
969 references. PCRE supports relative numbers with this syntax (I don't think
970 Oniguruma does).
971
972 12. Previously, a group with a zero repeat such as (...){0} was completely
973 omitted from the compiled regex. However, this means that if the group
974 was called as a subroutine from elsewhere in the pattern, things went wrong
975 (an internal error was given). Such groups are now left in the compiled
976 pattern, with a new opcode that causes them to be skipped at execution
977 time.
978
979 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
980 to the way PCRE behaves:
981
982 (a) A lone ] character is dis-allowed (Perl treats it as data).
983
984 (b) A back reference to an unmatched subpattern matches an empty string
985 (Perl fails the current match path).
986
987 (c) A data ] in a character class must be notated as \] because if the
988 first data character in a class is ], it defines an empty class. (In
989 Perl it is not possible to have an empty class.) The empty class []
990 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
991 The negative empty class [^] matches any one character, independently
992 of the DOTALL setting.
993
994 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
995 non-existent subpattern following a character class starting with ']' and
996 containing () gave an internal compiling error instead of "reference to
997 non-existent subpattern". Fortunately, when the pattern did exist, the
998 compiled code was correct. (When scanning forwards to check for the
999 existencd of the subpattern, it was treating the data ']' as terminating
1000 the class, so got the count wrong. When actually compiling, the reference
1001 was subsequently set up correctly.)
1002
1003 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1004 it was being rejected as not supported by pcre_dfa_exec(), even though
1005 other assertions are supported. I have made pcre_dfa_exec() support
1006 (*FAIL).
1007
1008 16. The implementation of 13c above involved the invention of a new opcode,
1009 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1010 cannot be changed at match time, I realized I could make a small
1011 improvement to matching performance by compiling OP_ALLANY instead of
1012 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1013 on the OP_ANY path.
1014
1015 17. Compiling pcretest on Windows with readline support failed without the
1016 following two fixes: (1) Make the unistd.h include conditional on
1017 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1018
1019 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1020 ncurses library to be included for pcretest when ReadLine support is
1021 requested, but also to allow for it to be overridden. This patch came from
1022 Daniel Bergström.
1023
1024 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1025 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1026 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1027 spotting this.
1028
1029
1030 Version 7.6 28-Jan-08
1031 ---------------------
1032
1033 1. A character class containing a very large number of characters with
1034 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1035 overflow.
1036
1037 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1038 HAVE_LONG_LONG is not defined.
1039
1040 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1041 bring them up to date and include new features. This patch includes:
1042
1043 - Fixed PH's badly added libz and libbz2 support.
1044 - Fixed a problem with static linking.
1045 - Added pcredemo. [But later removed - see 7 below.]
1046 - Fixed dftables problem and added an option.
1047 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1048 HAVE_LONG_LONG.
1049 - Added readline support for pcretest.
1050 - Added an listing of the option settings after cmake has run.
1051
1052 4. A user submitted a patch to Makefile that makes it easy to create
1053 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1054 Makefile.am that cause it to include this special target, without
1055 affecting anything else. Note that the same mingw target plus all
1056 the other distribution libraries and programs are now supported
1057 when configuring with CMake (see 6 below) instead of with
1058 Configure/Make.
1059
1060 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1061 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1062 exported in the Windows port". It has not yet been confirmed that the patch
1063 solves the problem, but it does no harm.
1064
1065 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1066 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1067 with CMake, and also correct the comment about stack recursion.
1068
1069 7. Remove the automatic building of pcredemo from the ./configure system and
1070 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1071 of a program that users should build themselves after PCRE is installed, so
1072 building it automatically is not really right. What is more, it gave
1073 trouble in some build environments.
1074
1075 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1076
1077
1078 Version 7.5 10-Jan-08
1079 ---------------------
1080
1081 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1082 values in parens when parsing an RE using the C++ wrapper."
1083
1084 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1085 Characters greater than 255 were excluded from the class instead of being
1086 included.
1087
1088 3. The same bug as (2) above applied to negated POSIX classes such as
1089 [:^space:].
1090
1091 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1092 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1093 I have changed it.
1094
1095 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1096 first named subpattern) and a construct such as (?&a) would reference the
1097 first named subpattern whose name started with "a" (in other words, the
1098 length check was missing). Both these problems are fixed. "Subpattern name
1099 expected" is now given for (?&) (a zero-length name), and this patch also
1100 makes it give the same error for \k'' (previously it complained that that
1101 was a reference to a non-existent subpattern).
1102
1103 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1104 this is right because (?- can be followed by option settings as well as by
1105 digits. I have, however, made the messages clearer.
1106
1107 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1108 than the number used in the conditional) now cause a compile-time error.
1109 This is actually not compatible with Perl, which accepts such patterns, but
1110 treats the conditional as always being FALSE (as PCRE used to), but it
1111 seems to me that giving a diagnostic is better.
1112
1113 8. Change "alphameric" to the more common word "alphanumeric" in comments
1114 and messages.
1115
1116 9. Fix two occurrences of "backslash" in comments that should have been
1117 "backspace".
1118
1119 10. Remove two redundant lines of code that can never be obeyed (their function
1120 was moved elsewhere).
1121
1122 11. The program that makes PCRE's Unicode character property table had a bug
1123 which caused it to generate incorrect table entries for sequences of
1124 characters that have the same character type, but are in different scripts.
1125 It amalgamated them into a single range, with the script of the first of
1126 them. In other words, some characters were in the wrong script. There were
1127 thirteen such cases, affecting characters in the following ranges:
1128
1129 U+002b0 - U+002c1
1130 U+0060c - U+0060d
1131 U+0061e - U+00612
1132 U+0064b - U+0065e
1133 U+0074d - U+0076d
1134 U+01800 - U+01805
1135 U+01d00 - U+01d77
1136 U+01d9b - U+01dbf
1137 U+0200b - U+0200f
1138 U+030fc - U+030fe
1139 U+03260 - U+0327f
1140 U+0fb46 - U+0fbb1
1141 U+10450 - U+1049d
1142
1143 12. The -o option (show only the matching part of a line) for pcregrep was not
1144 compatible with GNU grep in that, if there was more than one match in a
1145 line, it showed only the first of them. It now behaves in the same way as
1146 GNU grep.
1147
1148 13. If the -o and -v options were combined for pcregrep, it printed a blank
1149 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1150 does the same. The return code can be used to tell if there were any
1151 non-matching lines.
1152
1153 14. Added --file-offsets and --line-offsets to pcregrep.
1154
1155 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1156 infinitely looping recursion. The bug was that positive lookaheads were not
1157 being skipped when checking for a possible empty match (negative lookaheads
1158 and both kinds of lookbehind were skipped).
1159
1160 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1161 inclusion of <windows.h> to before rather than after the definition of
1162 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1163
1164 17. Specifying a possessive quantifier with a specific limit for a Unicode
1165 character property caused pcre_compile() to compile bad code, which led at
1166 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1167 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1168 caused the error; without that there was no problem.
1169
1170 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1171
1172 19. Added --enable-pcretest-libreadline.
1173
1174 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1175 RE::GlobalReplace(). As a result, the number of replacements returned was
1176 double what it should be. I removed one of the increments, but Craig sent a
1177 later patch that removed the other one (the right fix) and added unit tests
1178 that check the return values (which was not done before).
1179
1180 21. Several CMake things:
1181
1182 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1183 the names libpcre and libpcreposix, not just pcre and pcreposix.
1184
1185 (2) The above change means that pcretest and pcregrep are now correctly
1186 linked with the newly-built libraries, not previously installed ones.
1187
1188 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1189
1190 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1191 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1192 UTF-8 newline character). The key issue is that the pattern starts .*;
1193 this means that the match must be either at the beginning, or after a
1194 newline. The bug was in the code for advancing after a failed match and
1195 checking that the new position followed a newline. It was not taking
1196 account of UTF-8 characters correctly.
1197
1198 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1199 character classes. PCRE was not treating the sequence [:...:] as a
1200 character class unless the ... were all letters. Perl, however, seems to
1201 allow any characters between [: and :], though of course it rejects as
1202 unknown any "names" that contain non-letters, because all the known class
1203 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1204 for example, whereas PCRE did not - it did not recognize a POSIX character
1205 class. This seemed a bit dangerous, so the code has been changed to be
1206 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1207 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1208 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1209 Perl does, and where it didn't before.
1210
1211 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1212 Windows environments %n is disabled by default.
1213
1214
1215 Version 7.4 21-Sep-07
1216 ---------------------
1217
1218 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1219 means that a class such as [\s] counted as "explicit reference to CR or
1220 LF". That isn't really right - the whole point of the change was to try to
1221 help when there was an actual mention of one of the two characters. So now
1222 the change happens only if \r or \n (or a literal CR or LF) character is
1223 encountered.
1224
1225 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1226 of both had grown to the point where there were only 3 bits left.
1227 Fortunately, there was spare space in the data structure, and so I have
1228 moved the internal flags into a new 16-bit field to free up more option
1229 bits.
1230
1231 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1232 but did not set the internal JCHANGED flag - either of these is enough to
1233 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1234 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1235 start sets both bits.
1236
1237 4. Added options (at build time, compile time, exec time) to change \R from
1238 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1239
1240 5. doc/pcresyntax.html was missing from the distribution.
1241
1242 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1243 compatibility, even though it is no longer used.
1244
1245 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1246 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1247 windows.h file is present (where different names are used). [This was
1248 reversed later after testing - see 16 below.]
1249
1250 8. Changed all #include <config.h> to #include "config.h". There were also
1251 some further <pcre.h> cases that I changed to "pcre.h".
1252
1253 9. When pcregrep was used with the --colour option, it missed the line ending
1254 sequence off the lines that it output.
1255
1256 10. It was pointed out to me that arrays of string pointers cause lots of
1257 relocations when a shared library is dynamically loaded. A technique of
1258 using a single long string with a table of offsets can drastically reduce
1259 these. I have refactored PCRE in four places to do this. The result is
1260 dramatic:
1261
1262 Originally: 290
1263 After changing UCP table: 187
1264 After changing error message table: 43
1265 After changing table of "verbs" 36
1266 After changing table of Posix names 22
1267
1268 Thanks to the folks working on Gregex for glib for this insight.
1269
1270 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1271 unicode-properties was also set.
1272
1273 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1274
1275 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1276 checked only for CRLF.
1277
1278 14. Added casts to pcretest.c to avoid compiler warnings.
1279
1280 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1281
1282 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1283 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1284 entirely. This removes changes made in 7 above.
1285
1286 17. The CMake files have been updated, and there is now more information about
1287 building with CMake in the NON-UNIX-USE document.
1288
1289
1290 Version 7.3 28-Aug-07
1291 ---------------------
1292
1293 1. In the rejigging of the build system that eventually resulted in 7.1, the
1294 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1295 brackets there is not right, since it causes compilers to look for an
1296 installed pcre.h, not the version that is in the source that is being
1297 compiled (which of course may be different). I have changed it back to:
1298
1299 #include "pcre.h"
1300
1301 I have a vague recollection that the change was concerned with compiling in
1302 different directories, but in the new build system, that is taken care of
1303 by the VPATH setting the Makefile.
1304
1305 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1306 when the subject happened to end in the byte 0x85 (e.g. if the last
1307 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1308 characters but of course it shouldn't be taken as a newline when it is part
1309 of another character. The bug was that, for an unlimited repeat of . in
1310 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1311 characters when looking for a newline.
1312
1313 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1314
1315 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1316 in debug output.
1317
1318 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1319 long printing in the pcrecpp unittest when running under MinGW.
1320
1321 6. ESC_K was left out of the EBCDIC table.
1322
1323 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1324 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1325 limit also applies to "virtual nesting" when a pattern is recursive, and in
1326 this case 1000 isn't so big. I have been able to remove this limit at the
1327 expense of backing off one optimization in certain circumstances. Normally,
1328 when pcre_exec() would call its internal match() function recursively and
1329 immediately return the result unconditionally, it uses a "tail recursion"
1330 feature to save stack. However, when a subpattern that can match an empty
1331 string has an unlimited repetition quantifier, it no longer makes this
1332 optimization. That gives it a stack frame in which to save the data for
1333 checking that an empty string has been matched. Previously this was taken
1334 from the 1000-entry workspace that had been reserved. So now there is no
1335 explicit limit, but more stack is used.
1336
1337 8. Applied Daniel's patches to solve problems with the import/export magic
1338 syntax that is required for Windows, and which was going wrong for the
1339 pcreposix and pcrecpp parts of the library. These were overlooked when this
1340 problem was solved for the main library.
1341
1342 9. There were some crude static tests to avoid integer overflow when computing
1343 the size of patterns that contain repeated groups with explicit upper
1344 limits. As the maximum quantifier is 65535, the maximum group length was
1345 set at 30,000 so that the product of these two numbers did not overflow a
1346 32-bit integer. However, it turns out that people want to use groups that
1347 are longer than 30,000 bytes (though not repeat them that many times).
1348 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1349 made it possible to implement the integer overflow checks in a much more
1350 dynamic way, which I have now done. The artificial limitation on group
1351 length has been removed - we now have only the limit on the total length of
1352 the compiled pattern, which depends on the LINK_SIZE setting.
1353
1354 10. Fixed a bug in the documentation for get/copy named substring when
1355 duplicate names are permitted. If none of the named substrings are set, the
1356 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1357 empty string.
1358
1359 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1360 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1361 because the ] is interpreted as the first data character and the
1362 terminating ] is not found. PCRE has been made compatible with Perl in this
1363 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1364 cause memory overwriting.
1365
1366 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1367 string has been matched (to stop an infinite loop). It was not recognizing
1368 a conditional subpattern that could match an empty string if that
1369 subpattern was within another subpattern. For example, it looped when
1370 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1371 condition was not nested. This bug has been fixed.
1372
1373 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1374 past the start of the subject in the presence of bytes with the top bit
1375 set, for example "\x8aBCD".
1376
1377 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1378 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1379
1380 14. Optimized (?!) to (*FAIL).
1381
1382 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1383 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1384 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1385 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1386 does: it's just the validity check that is more restrictive.
1387
1388 16. Inserted checks for integer overflows during escape sequence (backslash)
1389 processing, and also fixed erroneous offset values for syntax errors during
1390 backslash processing.
1391
1392 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1393 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1394
1395 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1396 caused an overrun.
1397
1398 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1399 something other than just ASCII characters) inside a group that had an
1400 unlimited repeat caused a loop at compile time (while checking to see
1401 whether the group could match an empty string).
1402
1403 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1404 [\P{Any}] did so. (Error in the code for printing property names.)
1405
1406 21. An orphan \E inside a character class could cause a crash.
1407
1408 22. A repeated capturing bracket such as (A)? could cause a wild memory
1409 reference during compilation.
1410
1411 23. There are several functions in pcre_compile() that scan along a compiled
1412 expression for various reasons (e.g. to see if it's fixed length for look
1413 behind). There were bugs in these functions when a repeated \p or \P was
1414 present in the pattern. These operators have additional parameters compared
1415 with \d, etc, and these were not being taken into account when moving along
1416 the compiled data. Specifically:
1417
1418 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1419 length.
1420
1421 (b) An item such as \pL+ within a repeated group could cause crashes or
1422 loops.
1423
1424 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1425 "reference to non-existent subpattern" error.
1426
1427 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1428
1429 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1430 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1431
1432 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1433
1434 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1435 character were causing crashes (broken optimization).
1436
1437 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1438 \p or \P) caused a compile-time loop.
1439
1440 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1441 break. For example, the unstudied pattern [\r\n]A does not match the string
1442 "\r\nA" because change 7.0/46 below moves the current point on by two
1443 characters after failing to match at the start. However, the pattern \nA
1444 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1445 the same is true. There doesn't seem any very clean way out of this, but
1446 what I have chosen to do makes the common cases work: PCRE now takes note
1447 of whether there can be an explicit match for \r or \n anywhere in the
1448 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1449 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1450 pattern has explicit CR or LF references.
1451
1452 29. Added (*CR) etc for changing newline setting at start of pattern.
1453
1454
1455 Version 7.2 19-Jun-07
1456 ---------------------
1457
1458 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1459 which is apparently normally available under Windows.
1460
1461 2. Re-jig the pcregrep tests with different newline settings in an attempt
1462 to make them independent of the local environment's newline setting.
1463
1464 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1465
1466 4. Some of the "internals" tests were previously cut out when the link size
1467 was not 2, because the output contained actual offsets. The recent new
1468 "Z" feature of pcretest means that these can be cut out, making the tests
1469 usable with all link sizes.
1470
1471 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1472 stack recursion. This gives a massive performance boost under BSD, but just
1473 a small improvement under Linux. However, it saves one field in the frame
1474 in all cases.
1475
1476 6. Added more features from the forthcoming Perl 5.10:
1477
1478 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1479 recursion call. It refers to the nth most recently opened parentheses.
1480
1481 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1482 to be opened parentheses.
1483
1484 (c) Conditions that refer to capturing parentheses can be specified
1485 relatively, for example, (?(-2)... or (?(+3)...
1486
1487 (d) \K resets the start of the current match so that everything before
1488 is not part of it.
1489
1490 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1491
1492 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1493 reference syntax.
1494
1495 (g) (?| introduces a group in which the numbering of parentheses in each
1496 alternative starts with the same number.
1497
1498 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1499
1500 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1501 PCRE_INFO_JCHANGED.
1502
1503 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1504 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1505 for detecting groups that can match an empty string.
1506
1507 9. A pattern with a very large number of alternatives (more than several
1508 hundred) was running out of internal workspace during the pre-compile
1509 phase, where pcre_compile() figures out how much memory will be needed. A
1510 bit of new cunning has reduced the workspace needed for groups with
1511 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1512 workspace instead of running out of the 4096 that are available.
1513
1514 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1515
1516 11. Applied patch from Google to remove an optimization that didn't quite work.
1517 The report of the bug said:
1518
1519 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1520 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1521 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1522
1523 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1524 it matched the wrong number of bytes.
1525
1526
1527 Version 7.1 24-Apr-07
1528 ---------------------
1529
1530 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1531 that is more "standard", making use of automake and other Autotools. There
1532 is some re-arrangement of the files and adjustment of comments consequent
1533 on this.
1534
1535 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1536 for recursive directory scanning broke on some systems because the files
1537 are not scanned in any specific order and on different systems the order
1538 was different. A call to "sort" has been inserted into RunGrepTest for the
1539 approprate test as a short-term fix. In the longer term there may be an
1540 alternative.
1541
1542 3. I had an email from Eric Raymond about problems translating some of PCRE's
1543 man pages to HTML (despite the fact that I distribute HTML pages, some
1544 people do their own conversions for various reasons). The problems
1545 concerned the use of low-level troff macros .br and .in. I have therefore
1546 removed all such uses from the man pages (some were redundant, some could
1547 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1548 HTML has been updated to handle .nf/.fi and to complain if it encounters
1549 .br or .in.
1550
1551 4. Updated comments in configure.ac that get placed in config.h.in and also
1552 arranged for config.h to be included in the distribution, with the name
1553 config.h.generic, for the benefit of those who have to compile without
1554 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1555
1556 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1557 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1558 makevp.bat for the current PCRE, using the additional files
1559 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1560
1561 6. A Windows user reported a minor discrepancy with test 2, which turned out
1562 to be caused by a trailing space on an input line that had got lost in his
1563 copy. The trailing space was an accident, so I've just removed it.
1564
1565 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1566 that is needed.
1567
1568 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1569 as "const" (a) because they are and (b) because it helps the PHP
1570 maintainers who have recently made a script to detect big data structures
1571 in the php code that should be moved to the .rodata section. I remembered
1572 to update Builducptable as well, so it won't revert if ucptable.h is ever
1573 re-created.
1574
1575 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1576 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1577 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1578 support is not required. This saves 1.5-2K of code, which is important in
1579 some applications.
1580
1581 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1582 so as not to refer to the tables, even though these functions will never be
1583 called when UTF-8 support is disabled. Otherwise there are problems with a
1584 shared library.
1585
1586 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1587
1588 (a) It was defining its arguments as char * instead of void *.
1589
1590 (b) It was assuming that all moves were upwards in memory; this was true
1591 a long time ago when I wrote it, but is no longer the case.
1592
1593 The emulated memove() is provided for those environments that have neither
1594 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1595 is clearly not the case, as these two bugs were recently reported.
1596
1597 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1598 and Detrail to create the HTML documentation, the .txt form of the man
1599 pages, and it removes trailing spaces from listed files. It also creates
1600 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1601 case, it wraps all the #defines with #ifndefs. This script should be run
1602 before "make dist".
1603
1604 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1605 with Unicode property support.
1606
1607 (a) For a maximizing quantifier, if the two different cases of the
1608 character were of different lengths in their UTF-8 codings (there are
1609 some cases like this - I found 11), and the matching function had to
1610 back up over a mixture of the two cases, it incorrectly assumed they
1611 were both the same length.
1612
1613 (b) When PCRE was configured to use the heap rather than the stack for
1614 recursion during matching, it was not correctly preserving the data for
1615 the other case of a UTF-8 character when checking ahead for a match
1616 while processing a minimizing repeat. If the check also involved
1617 matching a wide character, but failed, corruption could cause an
1618 erroneous result when trying to check for a repeat of the original
1619 character.
1620
1621 13. Some tidying changes to the testing mechanism:
1622
1623 (a) The RunTest script now detects the internal link size and whether there
1624 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1625 values substituted by "configure". (The RunGrepTest script already did
1626 this for UTF-8.) The configure.ac script no longer substitutes the
1627 relevant variables.
1628
1629 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1630 with length and offset values. This means that the output is different
1631 for different internal link sizes. Test 2 is skipped for link sizes
1632 other than 2 because of this, bypassing the problem. Unfortunately,
1633 there was also a test in test 3 (the locale tests) that used /B and
1634 failed for link sizes other than 2. Rather than cut the whole test out,
1635 I have added a new /Z option to pcretest that replaces the length and
1636 offset values with spaces. This is now used to make test 3 independent
1637 of link size. (Test 2 will be tidied up later.)
1638
1639 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1640 segmentation fault instead of returning the appropriate error message.
1641
1642 15. In multiline mode when the newline sequence was set to "any", the pattern
1643 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1644 This doesn't seem right; it now treats the CRLF combination as the line
1645 ending, and so does not match in that case. It's only a pattern such as ^$
1646 that would hit this one: something like ^ABC$ would have failed after \r
1647 and then tried again after \r\n.
1648
1649 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1650 in an attempt to make files that differ only in their line terminators
1651 compare equal. This works on Linux.
1652
1653 17. Under certain error circumstances pcregrep might try to free random memory
1654 as it exited. This is now fixed, thanks to valgrind.
1655
1656 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1657 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1658 was because its rules for how to advance for /g after matching an empty
1659 string at the end of a line did not allow for this case. They now check for
1660 it specially.
1661
1662 20. pcretest is supposed to handle patterns and data of any length, by
1663 extending its buffers when necessary. It was getting this wrong when the
1664 buffer for a data line had to be extended.
1665
1666 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1667 CRLF as a newline sequence.
1668
1669 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1670 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1671 I have nevertheless tidied it up.
1672
1673 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1674
1675 24. Added a man page for pcre-config.
1676
1677
1678 Version 7.0 19-Dec-06
1679 ---------------------
1680
1681 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1682 moving to gcc 4.1.1.
1683
1684 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1685 sys/time.h, which is documented as needed for this function. It doesn't
1686 seem to matter on Linux, but it showed up on some releases of OS X.
1687
1688 3. It seems that there are systems where bytes whose values are greater than
1689 127 match isprint() in the "C" locale. The "C" locale should be the
1690 default when a C program starts up. In most systems, only ASCII printing
1691 characters match isprint(). This difference caused the output from pcretest
1692 to vary, making some of the tests fail. I have changed pcretest so that:
1693
1694 (a) When it is outputting text in the compiled version of a pattern, bytes
1695 other than 32-126 are always shown as hex escapes.
1696
1697 (b) When it is outputting text that is a matched part of a subject string,
1698 it does the same, unless a different locale has been set for the match
1699 (using the /L modifier). In this case, it uses isprint() to decide.
1700
1701 4. Fixed a major bug that caused incorrect computation of the amount of memory
1702 required for a compiled pattern when options that changed within the
1703 pattern affected the logic of the preliminary scan that determines the
1704 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1705 that the computed length was too small. The symptoms of this bug were
1706 either the PCRE error "internal error: code overflow" from pcre_compile(),
1707 or a glibc crash with a message such as "pcretest: free(): invalid next
1708 size (fast)". Examples of patterns that provoked this bug (shown in
1709 pcretest format) are:
1710
1711 /(?-x: )/x
1712 /(?x)(?-x: \s*#\s*)/
1713 /((?i)[\x{c0}])/8
1714 /(?i:[\x{c0}])/8
1715
1716 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1717 is now done differently.
1718
1719 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1720 wrapper classes; (b) implement a new function in the C++ scanner that is
1721 more efficient than the old way of doing things because it avoids levels of
1722 recursion in the regex matching; (c) add a paragraph to the documentation
1723 for the FullMatch() function.
1724
1725 6. The escape sequence \n was being treated as whatever was defined as
1726 "newline". Not only was this contrary to the documentation, which states
1727 that \n is character 10 (hex 0A), but it also went horribly wrong when
1728 "newline" was defined as CRLF. This has been fixed.
1729
1730 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1731 was being set to -1 for the "end of line" case (supposedly a value that no
1732 character can have). Though this value is never used (the check for end of
1733 line is "zero bytes in current character"), it caused compiler complaints.
1734 I've changed it to 0xffffffff.
1735
1736 8. In pcre_version.c, the version string was being built by a sequence of
1737 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1738 string (as it is for production releases) called a macro with an empty
1739 argument. The C standard says the result of this is undefined. The gcc
1740 compiler treats it as an empty string (which was what was wanted) but it is
1741 reported that Visual C gives an error. The source has been hacked around to
1742 avoid this problem.
1743
1744 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1745 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1746 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1747 of them did).
1748
1749 10. Originally, pcretest opened its input and output without "b"; then I was
1750 told that "b" was needed in some environments, so it was added for release
1751 5.0 to both the input and output. (It makes no difference on Unix-like
1752 systems.) Later I was told that it is wrong for the input on Windows. I've
1753 now abstracted the modes into two macros, to make it easier to fiddle with
1754 them, and removed "b" from the input mode under Windows.
1755
1756 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1757
1758 12. Added -help and --help to pcretest as an official way of being reminded
1759 of the options.
1760
1761 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1762 and pcrecpp.cc because they annoy compilers at high warning levels.
1763
1764 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1765
1766 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1767 scripts are not C programs :-) and which was not noticed because it works
1768 on Linux.
1769
1770 16. pcretest is supposed to handle any length of pattern and data line (as one
1771 line or as a continued sequence of lines) by extending its input buffer if
1772 necessary. This feature was broken for very long pattern lines, leading to
1773 a string of junk being passed to pcre_compile() if the pattern was longer
1774 than about 50K.
1775
1776 17. I have done a major re-factoring of the way pcre_compile() computes the
1777 amount of memory needed for a compiled pattern. Previously, there was code
1778 that made a preliminary scan of the pattern in order to do this. That was
1779 OK when PCRE was new, but as the facilities have expanded, it has become
1780 harder and harder to keep it in step with the real compile phase, and there
1781 have been a number of bugs (see for example, 4 above). I have now found a
1782 cunning way of running the real compile function in a "fake" mode that
1783 enables it to compute how much memory it would need, while actually only
1784 ever using a few hundred bytes of working memory and without too many
1785 tests of the mode. This should make future maintenance and development
1786 easier. A side effect of this work is that the limit of 200 on the nesting
1787 depth of parentheses has been removed (though this was never a serious
1788 limitation, I suspect). However, there is a downside: pcre_compile() now
1789 runs more slowly than before (30% or more, depending on the pattern). I
1790 hope this isn't a big issue. There is no effect on runtime performance.
1791
1792 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1793 newline (only possible for the last line of a file) and it was a
1794 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1795
1796 19. Added additional timing features to pcretest. (1) The -tm option now times
1797 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1798 separate command line item, by a number that specifies the number of
1799 repeats to use when timing. The default is 50000; this gives better
1800 precision, but takes uncomfortably long for very large patterns.
1801
1802 20. Extended pcre_study() to be more clever in cases where a branch of a
1803 subpattern has no definite first character. For example, (a*|b*)[cd] would
1804 previously give no result from pcre_study(). Now it recognizes that the
1805 first character must be a, b, c, or d.
1806
1807 21. There was an incorrect error "recursive call could loop indefinitely" if
1808 a subpattern (or the entire pattern) that was being tested for matching an
1809 empty string contained only one non-empty item after a nested subpattern.
1810 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1811 incorrectly, because the \d was being skipped in the check.
1812
1813 22. The pcretest program now has a new pattern option /B and a command line
1814 option -b, which is equivalent to adding /B to every pattern. This causes
1815 it to show the compiled bytecode, without the additional information that
1816 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1817 is the same as /B/I).
1818
1819 23. A new optimization is now able automatically to treat some sequences such
1820 as a*b as a*+b. More specifically, if something simple (such as a character
1821 or a simple class like \d) has an unlimited quantifier, and is followed by
1822 something that cannot possibly match the quantified thing, the quantifier
1823 is automatically "possessified".
1824
1825 24. A recursive reference to a subpattern whose number was greater than 39
1826 went wrong under certain circumstances in UTF-8 mode. This bug could also
1827 have affected the operation of pcre_study().
1828
1829 25. Realized that a little bit of performance could be had by replacing
1830 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1831
1832 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1833
1834 27. Possessive quantifiers such as a++ were previously implemented by turning
1835 them into atomic groups such as ($>a+). Now they have their own opcodes,
1836 which improves performance. This includes the automatically created ones
1837 from 23 above.
1838
1839 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1840 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1841 the first matched character to be a colon. This applied both to named and
1842 numbered groups.
1843
1844 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1845
1846 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1847 building PCRE on Windows easier, so I have included it in the distribution.
1848
1849 31. There is now a check in pcretest against a ridiculously large number being
1850 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1851 loop, the loop is abandoned.
1852
1853 32. Forward references to subpatterns in conditions such as (?(2)...) where
1854 subpattern 2 is defined later cause pcre_compile() to search forwards in
1855 the pattern for the relevant set of parentheses. This search went wrong
1856 when there were unescaped parentheses in a character class, parentheses
1857 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1858
1859 33. "Subroutine" calls and backreferences were previously restricted to
1860 referencing subpatterns earlier in the regex. This restriction has now
1861 been removed.
1862
1863 34. Added a number of extra features that are going to be in Perl 5.10. On the
1864 whole, these are just syntactic alternatives for features that PCRE had
1865 previously implemented using the Python syntax or my own invention. The
1866 other formats are all retained for compatibility.
1867
1868 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1869 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1870 also .NET compatible.
1871
1872 (b) A recursion or subroutine call to a named group can now be defined as
1873 (?&name) as well as (?P>name).
1874
1875 (c) A backreference to a named group can now be defined as \k<name> or
1876 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1877 5.10, are also .NET compatible.
1878
1879 (d) A conditional reference to a named group can now use the syntax
1880 (?(<name>) or (?('name') as well as (?(name).
1881
1882 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1883 groups (named and numbered) that are never evaluated inline, but can be
1884 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1885 is always false. There may be only one alternative in such a group.
1886
1887 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1888 as the simple (?(R). The condition is true only if the most recent
1889 recursion is that of the given number or name. It does not search out
1890 through the entire recursion stack.
1891
1892 (g) The escape \gN or \g{N} has been added, where N is a positive or
1893 negative number, specifying an absolute or relative reference.
1894
1895 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1896 some "unreachable code" warnings.
1897
1898 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1899 things, this adds five new scripts.
1900
1901 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1902 There were also incompatibilities regarding the handling of \Q..\E inside
1903 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1904 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1905
1906 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1907 matches an empty string, and forcibly breaks the loop. There were bugs in
1908 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1909 against aaaa the result was just "a" rather than "aaaa", for example. Two
1910 separate and independent bugs (that affected different cases) have been
1911 fixed.
1912
1913 39. Refactored the code to abolish the use of different opcodes for small
1914 capturing bracket numbers. This is a tidy that I avoided doing when I
1915 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1916 The new approach is not only tidier, it makes it possible to reduce the
1917 memory needed to fix the previous bug (38).
1918
1919 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1920 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1921 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1922 mode.
1923
1924 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1925 report.
1926
1927 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1928 copy construction and assignment in the C++ wrapper.
1929
1930 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1931 couple of bugs that could have given wrong results in the "--newline=crlf"
1932 case.
1933
1934 44. Added a number of casts and did some reorganization of signed/unsigned int
1935 variables following suggestions from Dair Grant. Also renamed the variable
1936 "this" as "item" because it is a C++ keyword.
1937
1938 45. Arranged for dftables to add
1939
1940 #include "pcre_internal.h"
1941
1942 to pcre_chartables.c because without it, gcc 4.x may remove the array
1943 definition from the final binary if PCRE is built into a static library and
1944 dead code stripping is activated.
1945
1946 46. For an unanchored pattern, if a match attempt fails at the start of a
1947 newline sequence, and the newline setting is CRLF or ANY, and the next two
1948 characters are CRLF, advance by two characters instead of one.
1949
1950
1951 Version 6.7 04-Jul-06
1952 ---------------------
1953
1954 1. In order to handle tests when input lines are enormously long, pcretest has
1955 been re-factored so that it automatically extends its buffers when
1956 necessary. The code is crude, but this _is_ just a test program. The
1957 default size has been increased from 32K to 50K.
1958
1959 2. The code in pcre_study() was using the value of the re argument before
1960 testing it for NULL. (Of course, in any sensible call of the function, it
1961 won't be NULL.)
1962
1963 3. The memmove() emulation function in pcre_internal.h, which is used on
1964 systems that lack both memmove() and bcopy() - that is, hardly ever -
1965 was missing a "static" storage class specifier.
1966
1967 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1968 containing an extended class (one that cannot be represented by a bitmap
1969 because it contains high-valued characters or Unicode property items, e.g.
1970 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1971 pattern, but PCRE should not loop if you do not (it no longer does).
1972 [Detail: two cases were found: (a) a repeated subpattern containing an
1973 extended class; (b) a recursive reference to a subpattern that followed a
1974 previous extended class. It wasn't skipping over the extended class
1975 correctly when UTF-8 mode was not set.]
1976
1977 5. A negated single-character class was not being recognized as fixed-length
1978 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1979 compile error "lookbehind assertion is not fixed length".
1980
1981 6. The RunPerlTest auxiliary script was showing an unexpected difference
1982 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1983 write a Perl script that can interpret lines of an input file either as
1984 byte characters or as UTF-8, which is what "perltest" was being required to
1985 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1986 can't do is switch easily at run time between having the "use utf8;" pragma
1987 or not. In the end, I fudged it by using the RunPerlTest script to insert
1988 "use utf8;" explicitly for the UTF-8 tests.
1989
1990 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1991 the end of the subject string, contrary to the documentation and to what
1992 Perl does. This was true of both matching functions. Now it matches only at
1993 the start of the subject and immediately after *internal* newlines.
1994
1995 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1996 a pointer to an int instead of a pointer to an unsigned long int. This
1997 caused problems on 64-bit systems.
1998
1999 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2000 instance of the 'standard' template library not being so standard".
2001
2002 10. There was no check on the number of named subpatterns nor the maximum
2003 length of a subpattern name. The product of these values is used to compute
2004 the size of the memory block for a compiled pattern. By supplying a very
2005 long subpattern name and a large number of named subpatterns, the size
2006 computation could be caused to overflow. This is now prevented by limiting
2007 the length of names to 32 characters, and the number of named subpatterns
2008 to 10,000.
2009
2010 11. Subpatterns that are repeated with specific counts have to be replicated in
2011 the compiled pattern. The size of memory for this was computed from the
2012 length of the subpattern and the repeat count. The latter is limited to
2013 65535, but there was no limit on the former, meaning that integer overflow
2014 could in principle occur. The compiled length of a repeated subpattern is
2015 now limited to 30,000 bytes in order to prevent this.
2016
2017 12. Added the optional facility to have named substrings with the same name.
2018
2019 13. Added the ability to use a named substring as a condition, using the
2020 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2021 are numbers (not recommended). Forward references are permitted.
2022
2023 14. Added forward references in named backreferences (if you see what I mean).
2024
2025 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2026 pattern could run off the end of the subject. For example, the pattern
2027 "(?s)(.{1,5})"8 did this with the subject "ab".
2028
2029 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2030 PCRE_CASELESS was set when matching characters that were quantified with ?
2031 or *.
2032
2033 17. A character class other than a single negated character that had a minimum
2034 but no maximum quantifier - for example [ab]{6,} - was not handled
2035 correctly by pce_dfa_exec(). It would match only one character.
2036
2037 18. A valid (though odd) pattern that looked like a POSIX character
2038 class but used an invalid character after [ (for example [[,abc,]]) caused
2039 pcre_compile() to give the error "Failed: internal error: code overflow" or
2040 in some cases to crash with a glibc free() error. This could even happen if
2041 the pattern terminated after [[ but there just happened to be a sequence of
2042 letters, a binary zero, and a closing ] in the memory that followed.
2043
2044 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2045 over the years. Originally (before any Unicode support), just the bottom 8
2046 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2047 output from "man perlunicode" includes this:
2048
2049 The regular expression compiler produces polymorphic opcodes. That
2050 is, the pattern adapts to the data and automatically switches to
2051 the Unicode character scheme when presented with Unicode data--or
2052 instead uses a traditional byte scheme when presented with byte
2053 data.
2054
2055 Sadly, a wide octal escape does not cause a switch, and in a string with
2056 no other multibyte characters, these octal escapes are treated as before.
2057 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2058 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2059 Unicode string.
2060
2061 I have not perpetrated such confusion in PCRE. Up till now, it took just
2062 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2063 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2064 translate to the appropriate multibyte character.
2065
2066 29. Applied some refactoring to reduce the number of warnings from Microsoft
2067 and Borland compilers. This has included removing the fudge introduced
2068 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2069 a warning about an unused variable.
2070
2071 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2072 characters since release 4.0, because Perl (from release 5.004) does not.
2073 [Or at least, is documented not to: some releases seem to be in conflict
2074 with the documentation.] However, when a pattern was studied with
2075 pcre_study() and all its branches started with \s, PCRE still included VT
2076 as a possible starting character. Of course, this did no harm; it just
2077 caused an unnecessary match attempt.
2078
2079 22. Removed a now-redundant internal flag bit that recorded the fact that case
2080 dependency changed within the pattern. This was once needed for "required
2081 byte" processing, but is no longer used. This recovers a now-scarce options
2082 bit. Also moved the least significant internal flag bit to the most-
2083 significant bit of the word, which was not previously used (hangover from
2084 the days when it was an int rather than a uint) to free up another bit for
2085 the future.
2086
2087 23. Added support for CRLF line endings as well as CR and LF. As well as the
2088 default being selectable at build time, it can now be changed at runtime
2089 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2090 specify that it is scanning data with non-default line endings.
2091
2092 24. Changed the definition of CXXLINK to make it agree with the definition of
2093 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2094
2095 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2096 recursions. This makes a big different to stack usage for some patterns.
2097
2098 26. If a subpattern containing a named recursion or subroutine reference such
2099 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2100 the space required for the compiled pattern went wrong and gave too small a
2101 value. Depending on the environment, this could lead to "Failed: internal
2102 error: code overflow at offset 49" or "glibc detected double free or
2103 corruption" errors.
2104
2105 27. Applied patches from Google (a) to support the new newline modes and (b) to
2106 advance over multibyte UTF-8 characters in GlobalReplace.
2107
2108 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2109 difference for some implementation of PCRE in some Windows version.
2110
2111 29. Added some extra testing facilities to pcretest:
2112
2113 \q<number> in a data line sets the "match limit" value
2114 \Q<number> in a data line sets the "match recursion limt" value
2115 -S <number> sets the stack size, where <number> is in megabytes
2116
2117 The -S option isn't available for Windows.
2118
2119
2120 Version 6.6 06-Feb-06
2121 ---------------------
2122
2123 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2124 in pcreposix.h. I have copied the definition from pcre.h.
2125
2126 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2127 because pcre.h is no longer a built file.
2128
2129 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2130 not normally included in the compiled code.
2131
2132
2133 Version 6.5 01-Feb-06
2134 ---------------------
2135
2136 1. When using the partial match feature with pcre_dfa_exec(), it was not
2137 anchoring the second and subsequent partial matches at the new starting
2138 point. This could lead to incorrect results. For example, with the pattern
2139 /1234/, partially matching against "123" and then "a4" gave a match.
2140
2141 2. Changes to pcregrep:
2142
2143 (a) All non-match returns from pcre_exec() were being treated as failures
2144 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2145 error message is output. Some extra information is given for the
2146 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2147 probably the only errors that are likely to be caused by users (by
2148 specifying a regex that has nested indefinite repeats, for instance).
2149 If there are more than 20 of these errors, pcregrep is abandoned.
2150
2151 (b) A binary zero was treated as data while matching, but terminated the
2152 output line if it was written out. This has been fixed: binary zeroes
2153 are now no different to any other data bytes.
2154
2155 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2156 used to set a locale for matching. The --locale=xxxx long option has
2157 been added (no short equivalent) to specify a locale explicitly on the
2158 pcregrep command, overriding the environment variables.
2159
2160 (d) When -B was used with -n, some line numbers in the output were one less
2161 than they should have been.
2162
2163 (e) Added the -o (--only-matching) option.
2164
2165 (f) If -A or -C was used with -c (count only), some lines of context were
2166 accidentally printed for the final match.
2167
2168 (g) Added the -H (--with-filename) option.
2169
2170 (h) The combination of options -rh failed to suppress file names for files
2171 that were found from directory arguments.
2172
2173 (i) Added the -D (--devices) and -d (--directories) options.
2174
2175 (j) Added the -F (--fixed-strings) option.
2176
2177 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2178
2179 (l) Added the --colo(u)r option.
2180
2181 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2182 is not present by default.
2183
2184 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2185 items such as (?R) or (?1), when the recursion could match a number of
2186 alternatives. If it matched one of the alternatives, but subsequently,
2187 outside the recursion, there was a failure, the code tried to back up into
2188 the recursion. However, because of the way PCRE is implemented, this is not
2189 possible, and the result was an incorrect result from the match.
2190
2191 In order to prevent this happening, the specification of recursion has
2192 been changed so that all such subpatterns are automatically treated as
2193 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2194
2195 4. I had overlooked the fact that, in some locales, there are characters for
2196 which isalpha() is true but neither isupper() nor islower() are true. In
2197 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2198 and ordfeminine) are like this. This affected the treatment of \w and \W
2199 when they appeared in character classes, but not when they appeared outside
2200 a character class. The bit map for "word" characters is now created
2201 separately from the results of isalnum() instead of just taking it from the
2202 upper, lower, and digit maps. (Plus the underscore character, of course.)
2203
2204 5. The above bug also affected the handling of POSIX character classes such as
2205 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2206 permanent tables. Instead, the bit maps for such a class were previously
2207 created as the appropriate unions of the upper, lower, and digit bitmaps.
2208 Now they are created by subtraction from the [[:word:]] class, which has
2209 its own bitmap.
2210
2211 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2212 It is created by subtracting the vertical space characters (\x09, \x0a,
2213 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2214 subtraction was done in the overall bitmap for a character class, meaning
2215 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2216 be recognized. This bug has been fixed.
2217
2218 7. Patches from the folks at Google:
2219
2220 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2221 real life, but is still worth protecting against".
2222
2223 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2224 regular expressions".
2225
2226 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2227 have it.
2228
2229 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2230 "configure" and the latter not, in order to fix a problem somebody had
2231 with compiling the Arg class on HP-UX.
2232
2233 (e) Improve the error-handling of the C++ wrapper a little bit.
2234
2235 (f) New tests for checking recursion limiting.
2236
2237 8. The pcre_memmove() function, which is used only if the environment does not
2238 have a standard memmove() function (and is therefore rarely compiled),
2239 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2240 returning a result (though PCRE never actually uses the result).
2241
2242 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2243 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2244 returned instead of calling malloc() with an overflowing number that would
2245 most likely cause subsequent chaos.
2246
2247 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2248
2249 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2250 with this option is matched, the nmatch and pmatch options of regexec() are
2251 ignored.
2252
2253 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2254 provided in case anyone wants to the the POSIX interface with UTF-8
2255 strings.
2256
2257 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2258 C++ linking (needed for some HP-UX environments).
2259
2260 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2261 (unused parameter) and in the pcre_printint() function (omitted "default"
2262 switch label when the default is to do nothing).
2263
2264 15. Added some code to make it possible, when PCRE is compiled as a C++
2265 library, to replace subject pointers for pcre_exec() with a smart pointer
2266 class, thus making it possible to process discontinuous strings.
2267
2268 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2269 much the same function. They were added by different people who were trying
2270 to make PCRE easy to compile on non-Unix systems. It has been suggested
2271 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2272 for compiling on Windows systems. I have therefore replaced it with
2273 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2274 defaults to "extern" for C or "extern C" for C++, which works fine on
2275 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2276 SCOPE with something explicit in config.h. In addition:
2277
2278 (a) pcreposix.h still had just "extern" instead of either of these macros;
2279 I have replaced it with PCRE_DATA_SCOPE.
2280
2281 (b) Functions such as _pcre_xclass(), which are internal to the library,
2282 but external in the C sense, all had PCRE_EXPORT in their definitions.
2283 This is apparently wrong for the Windows case, so I have removed it.
2284 (It makes no difference on Unix-like systems.)
2285
2286 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2287 of recursive calls to match(). This is different to MATCH_LIMIT because
2288 that limits the total number of calls to match(), not all of which increase
2289 the depth of recursion. Limiting the recursion depth limits the amount of
2290 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2291 when PCRE is compiled, and changed at run time. A patch from Google adds
2292 this functionality to the C++ interface.
2293
2294 18. Changes to the handling of Unicode character properties:
2295
2296 (a) Updated the table to Unicode 4.1.0.
2297
2298 (b) Recognize characters that are not in the table as "Cn" (undefined).
2299
2300 (c) I revised the way the table is implemented to a much improved format
2301 which includes recognition of ranges. It now supports the ranges that
2302 are defined in UnicodeData.txt, and it also amalgamates other
2303 characters into ranges. This has reduced the number of entries in the
2304 table from around 16,000 to around 3,000, thus reducing its size
2305 considerably. I realized I did not need to use a tree structure after
2306 all - a binary chop search is just as efficient. Having reduced the
2307 number of entries, I extended their size from 6 bytes to 8 bytes to
2308 allow for more data.
2309
2310 (d) Added support for Unicode script names via properties such as \p{Han}.
2311
2312 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2313 matching that character.
2314
2315 20. When matching a repeated Unicode property with a minimum greater than zero,
2316 (for example \pL{2,}), PCRE could look past the end of the subject if it
2317 reached it while seeking the minimum number of characters. This could
2318 happen only if some of the characters were more than one byte long, because
2319 there is a check for at least the minimum number of bytes.
2320
2321 21. Refactored the implementation of \p and \P so as to be more general, to
2322 allow for more different types of property in future. This has changed the
2323 compiled form incompatibly. Anybody with saved compiled patterns that use
2324 \p or \P will have to recompile them.
2325
2326 22. Added "Any" and "L&" to the supported property types.
2327
2328 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2329 but give a compile time error if the value is greater than 0xff.
2330
2331 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2332 accidentally not being installed or uninstalled.
2333
2334 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2335 made were to insert the current release number. This seemed silly, because
2336 it made things harder for people building PCRE on systems that don't run
2337 "configure". I have turned pcre.h into a distributed file, no longer built
2338 by "configure", with the version identification directly included. There is
2339 no longer a pcre.h.in file.
2340
2341 However, this change necessitated a change to the pcre-config script as
2342 well. It is built from pcre-config.in, and one of the substitutions was the
2343 release number. I have updated configure.ac so that ./configure now finds
2344 the release number by grepping pcre.h.
2345
2346 26. Added the ability to run the tests under valgrind.
2347
2348
2349 Version 6.4 05-Sep-05
2350 ---------------------
2351
2352 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2353 "--" to be printed when multiple files were scanned, even when none of the
2354 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2355 consider it to be a bug, and have restored the previous behaviour.
2356
2357 2. A couple of code tidies to get rid of compiler warnings.
2358
2359 3. The pcretest program used to cheat by referring to symbols in the library
2360 whose names begin with _pcre_. These are internal symbols that are not
2361 really supposed to be visible externally, and in some environments it is
2362 possible to suppress them. The cheating is now confined to including
2363 certain files from the library's source, which is a bit cleaner.
2364
2365 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2366 file's purpose clearer.
2367
2368 5. Reorganized pcre_ucp_findchar().
2369
2370
2371 Version 6.3 15-Aug-05
2372 ---------------------
2373
2374 1. The file libpcre.pc.in did not have general read permission in the tarball.
2375
2376 2. There were some problems when building without C++ support:
2377
2378 (a) If C++ support was not built, "make install" and "make test" still
2379 tried to test it.
2380
2381 (b) There were problems when the value of CXX was explicitly set. Some
2382 changes have been made to try to fix these, and ...
2383
2384 (c) --disable-cpp can now be used to explicitly disable C++ support.
2385
2386 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2387 backslash in a target when C++ was disabled. This confuses some
2388 versions of "make", apparently. Using an intermediate variable solves
2389 this. (Same for CPP_LOBJ.)
2390
2391 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2392 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2393 necessary on certain architectures.
2394
2395 4. Added a setting of -export-symbols-regex to the link command to remove
2396 those symbols that are exported in the C sense, but actually are local
2397 within the library, and not documented. Their names all begin with
2398 "_pcre_". This is not a perfect job, because (a) we have to except some
2399 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2400 available (and never for static libraries). I have made a note to try to
2401 find a way round (a) in the future.
2402
2403
2404 Version 6.2 01-Aug-05
2405 ---------------------
2406
2407 1. There was no test for integer overflow of quantifier values. A construction
2408 such as {1111111111111111} would give undefined results. What is worse, if
2409 a minimum quantifier for a parenthesized subpattern overflowed and became
2410 negative, the calculation of the memory size went wrong. This could have
2411 led to memory overwriting.
2412
2413 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2414
2415 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2416 operating environments where this matters.
2417
2418 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2419 PCRE options from within the C++ wrapper.
2420
2421 5. Named capturing subpatterns were not being correctly counted when a pattern
2422 was compiled. This caused two problems: (a) If there were more than 100
2423 such subpatterns, the calculation of the memory needed for the whole
2424 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2425 back references of the form \12, where the number was greater than 9, were
2426 not recognized as back references, even though there were sufficient
2427 previous subpatterns.
2428
2429 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2430 versions of gcc, e.g. 2.95.4.
2431
2432
2433 Version 6.1 21-Jun-05
2434 ---------------------
2435
2436 1. There was one reference to the variable "posix" in pcretest.c that was not
2437 surrounded by "#if !defined NOPOSIX".
2438
2439 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2440 the cross-check on the old pcre_info() function, for the benefit of the
2441 cut-down version of PCRE that is currently imported into Exim.
2442
2443 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2444 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2445 patterns that start (?i) but I don't think that matters. The use of (?i) is
2446 just an example; this all applies to the other options as well.
2447
2448 4. Since libtool seems to echo the compile commands it is issuing, the output
2449 from "make" can be reduced a bit by putting "@" in front of each libtool
2450 compile command.
2451
2452 5. Patch from the folks at Google for configure.in to be a bit more thorough
2453 in checking for a suitable C++ installation before trying to compile the
2454 C++ stuff. This should fix a reported problem when a compiler was present,
2455 but no suitable headers.
2456
2457 6. The man pages all had just "PCRE" as their title. I have changed them to
2458 be the relevant file name. I have also arranged that these names are
2459 retained in the file doc/pcre.txt, which is a concatenation in text format
2460 of all the man pages except the little individual ones for each function.
2461
2462 7. The NON-UNIX-USE file had not been updated for the different set of source
2463 files that come with release 6. I also added a few comments about the C++
2464 wrapper.
2465
2466
2467 Version 6.0 07-Jun-05
2468 ---------------------
2469
2470 1. Some minor internal re-organization to help with my DFA experiments.
2471
2472 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2473 didn't matter for the library itself when fully configured, but did matter
2474 when compiling without UCP support, or within Exim, where the ucp files are
2475 not imported.
2476
2477 3. Refactoring of the library code to split up the various functions into
2478 different source modules. The addition of the new DFA matching code (see
2479 below) to a single monolithic source would have made it really too
2480 unwieldy, quite apart from causing all the code to be include in a
2481 statically linked application, when only some functions are used. This is
2482 relevant even without the DFA addition now that patterns can be compiled in
2483 one application and matched in another.
2484
2485 The downside of splitting up is that there have to be some external
2486 functions and data tables that are used internally in different modules of
2487 the library but which are not part of the API. These have all had their
2488 names changed to start with "_pcre_" so that they are unlikely to clash
2489 with other external names.
2490
2491 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2492 a different (DFA) algorithm. Although it is slower than the original
2493 function, it does have some advantages for certain types of matching
2494 problem.
2495
2496 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2497 including restarting after a partial match.
2498
2499 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2500 defined when compiling for Windows was sent to me. I have put it into the
2501 code, though I have no means of testing or verifying it.
2502
2503 7. Added the pcre_refcount() auxiliary function.
2504
2505 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2506 match before or at the first newline in the subject string. In pcretest,
2507 the /f option on a pattern can be used to set this.
2508
2509 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2510 would behave wrongly. This has been present in PCRE since release 4.0.
2511
2512 10. A number of changes to the pcregrep command:
2513
2514 (a) Refactored how -x works; insert ^(...)$ instead of setting
2515 PCRE_ANCHORED and checking the length, in preparation for adding
2516 something similar for -w.
2517
2518 (b) Added the -w (match as a word) option.
2519
2520 (c) Refactored the way lines are read and buffered so as to have more
2521 than one at a time available.
2522
2523 (d) Implemented a pcregrep test script.
2524
2525 (e) Added the -M (multiline match) option. This allows patterns to match
2526 over several lines of the subject. The buffering ensures that at least
2527 8K, or the rest of the document (whichever is the shorter) is available
2528 for matching (and similarly the previous 8K for lookbehind assertions).
2529
2530 (f) Changed the --help output so that it now says
2531
2532 -w, --word-regex(p)
2533
2534 instead of two lines, one with "regex" and the other with "regexp"
2535 because that confused at least one person since the short forms are the
2536 same. (This required a bit of code, as the output is generated
2537 automatically from a table. It wasn't just a text change.)
2538
2539 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2540 option but starts with a hyphen. Could be a pattern or a path name
2541 starting with a hyphen, for instance.
2542
2543 (h) "-" can be given as a file name to represent stdin.
2544
2545 (i) When file names are being printed, "(standard input)" is used for
2546 the standard input, for compatibility with GNU grep. Previously
2547 "<stdin>" was used.
2548
2549 (j) The option --label=xxx can be used to supply a name to be used for
2550 stdin when file names are being printed. There is no short form.
2551
2552 (k) Re-factored the options decoding logic because we are going to add
2553 two more options that take data. Such options can now be given in four
2554 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2555
2556 (l) Added the -A, -B, and -C options for requesting that lines of context
2557 around matches be printed.
2558
2559 (m) Added the -L option to print the names of files that do not contain
2560 any matching lines, that is, the complement of -l.
2561
2562 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2563 continue to scan other files.
2564
2565 (o) The -s option was incorrectly implemented. For compatibility with other
2566 greps, it now suppresses the error message for a non-existent or non-
2567 accessible file (but not the return code). There is a new option called
2568 -q that suppresses the output of matching lines, which was what -s was
2569 previously doing.
2570
2571 (p) Added --include and --exclude options to specify files for inclusion
2572 and exclusion when recursing.
2573
2574 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2575 Hopefully, it now does.
2576
2577 12. Missing cast in pcre_study().
2578
2579 13. Added an "uninstall" target to the makefile.
2580
2581 14. Replaced "extern" in the function prototypes in Makefile.in with
2582 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2583 world, but is set differently for Windows.
2584
2585 15. Added a second compiling function called pcre_compile2(). The only
2586 difference is that it has an extra argument, which is a pointer to an
2587 integer error code. When there is a compile-time failure, this is set
2588 non-zero, in addition to the error test pointer being set to point to an
2589 error message. The new argument may be NULL if no error number is required
2590 (but then you may as well call pcre_compile(), which is now just a
2591 wrapper). This facility is provided because some applications need a
2592 numeric error indication, but it has also enabled me to tidy up the way
2593 compile-time errors are handled in the POSIX wrapper.
2594
2595 16. Added VPATH=.libs to the makefile; this should help when building with one
2596 prefix path and installing with another. (Or so I'm told by someone who
2597 knows more about this stuff than I do.)
2598
2599 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2600 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2601 match everything, including newlines. This is not POSIX-compatible, but
2602 somebody wanted the feature. From pcretest it can be activated by using
2603 both the P and the s flags.
2604
2605 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2606
2607 19. libpcre.pc was being incorrectly installed as executable.
2608
2609 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2610 it now also looks for '\r' so that it will work unmodified on Windows.
2611
2612 21. Added Google's contributed C++ wrapper to the distribution.
2613
2614 22. Added some untidy missing memory free() calls in pcretest, to keep
2615 Electric Fence happy when testing.
2616
2617
2618
2619 Version 5.0 13-Sep-04
2620 ---------------------
2621
2622 1. Internal change: literal characters are no longer packed up into items
2623 containing multiple characters in a single byte-string. Each character
2624 is now matched using a separate opcode. However, there may be more than one
2625 byte in the character in UTF-8 mode.
2626
2627 2. The pcre_callout_block structure has two new fields: pattern_position and
2628 next_item_length. These contain the offset in the pattern to the next match
2629 item, and its length, respectively.
2630
2631 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2632 insertion of callouts before each pattern item. Added the /C option to
2633 pcretest to make use of this.
2634
2635 4. On the advice of a Windows user, the lines
2636
2637 #if defined(_WIN32) || defined(WIN32)
2638 _setmode( _fileno( stdout ), 0x8000 );
2639 #endif /* defined(_WIN32) || defined(WIN32) */
2640
2641 have been added to the source of pcretest. This apparently does useful
2642 magic in relation to line terminators.
2643
2644 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2645 for the benefit of those environments where the "b" makes a difference.
2646
2647 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2648 to know about it. I have put a hack into configure.in that adds in code
2649 to set GCC=yes if CC=icc. This seems to end up at a point in the
2650 generated configure script that is early enough to affect the setting of
2651 compiler options, which is what is needed, but I have no means of testing
2652 whether it really works. (The user who reported this had patched the
2653 generated configure script, which of course I cannot do.)
2654
2655 LATER: After change 22 below (new libtool files), the configure script
2656 seems to know about icc (and also ecc). Therefore, I have commented out
2657 this hack in configure.in.
2658
2659 7. Added support for pkg-config (2 patches were sent in).
2660
2661 8. Negated POSIX character classes that used a combination of internal tables
2662 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2663 [[:^ascii]]. Typically, they would match almost any characters. The other
2664 POSIX classes were not broken in this way.
2665
2666 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2667 to find the match, as PCRE was deluded into thinking that the match had to
2668 start at the start point or following a newline. The same bug applied to
2669 patterns with negative forward assertions or any backward assertions
2670 preceding ".*" at the start, unless the pattern required a fixed first
2671 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2672
2673 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2674 starting at the last subject character, bytes beyond the end of the subject
2675 string were read.
2676
2677 11. Renamed the variable "class" as "classbits" to make life easier for C++
2678 users. (Previously there was a macro definition, but it apparently wasn't
2679 enough.)
2680
2681 12. Added the new field "tables" to the extra data so that tables can be passed
2682 in at exec time, or the internal tables can be re-selected. This allows
2683 a compiled regex to be saved and re-used at a later time by a different
2684 program that might have everything at different addresses.
2685
2686 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2687 -R library as well as a -L library.
2688
2689 14. The debugging options of pcretest (-d on the command line or D on a
2690 pattern) showed incorrect output for anything following an extended class
2691 that contained multibyte characters and which was followed by a quantifier.
2692
2693 15. Added optional support for general category Unicode character properties
2694 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2695 support. It adds about 90K to the size of the library. The meanings of the
2696 inbuilt class escapes such as \d and \s have NOT been changed.
2697
2698 16. Updated pcredemo.c to include calls to free() to release the memory for the
2699 compiled pattern.
2700
2701 17. The generated file chartables.c was being created in the source directory
2702 instead of in the building directory. This caused the build to fail if the
2703 source directory was different from the building directory, and was
2704 read-only.
2705
2706 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2707 file. No doubt somebody will tell me if they don't make sense... Also added
2708 Dan Mooney's comments about building on OpenVMS.
2709
2710 19. Added support for partial matching via the PCRE_PARTIAL option for
2711 pcre_exec() and the \P data escape in pcretest.
2712
2713 20. Extended pcretest with 3 new pattern features:
2714
2715 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2716 write the compiled pattern to the file whose name is "rest-of-line".
2717 This is a straight binary dump of the data, with the saved pointer to
2718 the character tables forced to be NULL. The study data, if any, is
2719 written too. After writing, pcretest reads a new pattern.
2720
2721 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2722 compiled pattern from the given file. There must not be any
2723 occurrences of "<" in the file name (pretty unlikely); if there are,
2724 pcretest will instead treat the initial "<" as a pattern delimiter.
2725 After reading in the pattern, pcretest goes on to read data lines as
2726 usual.
2727
2728 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2729 and 16-bit fields in a compiled pattern, to simulate a pattern that
2730 was compiled on a host of opposite endianness.
2731
2732 21. The pcre-exec() function can now cope with patterns that were compiled on
2733 hosts of opposite endianness, with this restriction:
2734
2735 As for any compiled expression that is saved and used later, the tables
2736 pointer field cannot be preserved; the extra_data field in the arguments
2737 to pcre_exec() should be used to pass in a tables address if a value
2738 other than the default internal tables were used at compile time.
2739
2740 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2741 now diagnosed as an error. Previously, most of the time, a negative number
2742 would have been treated as zero, but if in addition "ovector" was passed as
2743 NULL, a crash could occur.
2744
2745 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2746 new versions from the libtool 1.5 distribution (the last one is a copy of
2747 a file called libtool.m4). This seems to have fixed the need to patch
2748 "configure" to support Darwin 1.3 (which I used to do). However, I still
2749 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2750 workstation).
2751
2752 24. Changed the PCRE licence to be the more standard "BSD" licence.
2753
2754
2755 Version 4.5 01-Dec-03
2756 ---------------------
2757
2758 1. There has been some re-arrangement of the code for the match() function so
2759 that it can be compiled in a version that does not call itself recursively.
2760 Instead, it keeps those local variables that need separate instances for
2761 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2762 needs to "recurse". Keeping track of where control must go is done by means
2763 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2764 hide most of the details from the main code, and operates only if
2765 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2766 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2767 operating.
2768
2769 To make it easier for callers to provide specially tailored get/free
2770 functions for this usage, two new functions, pcre_stack_malloc, and
2771 pcre_stack_free, are used. They are always called in strict stacking order,
2772 and the size of block requested is always the same.
2773
2774 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2775 PCRE has been compiled to use the stack or the heap for recursion. The
2776 -C option of pcretest uses this to show which version is compiled.
2777
2778 A new data escape \S, is added to pcretest; it causes the amounts of store
2779 obtained and freed by both kinds of malloc/free at match time to be added
2780 to the output.
2781
2782 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2783 what's available on my current Linux desktop machine.
2784
2785 3. When matching a UTF-8 string, the test for a valid string at the start has
2786 been extended. If start_offset is not zero, PCRE now checks that it points
2787 to a byte that is the start of a UTF-8 character. If not, it returns
2788 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2789 this is necessary because there may be backward assertions in the pattern.
2790 When matching the same subject several times, it may save resources to use
2791 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2792
2793 4. The code for checking the validity of UTF-8 strings has been tightened so
2794 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2795 containing "overlong sequences".
2796
2797 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2798 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2799 should have been "&", but it just so happened that all the cases this let
2800 through by mistake were picked up later in the function.
2801
2802 6. I had used a variable called "isblank" - this is a C99 function, causing
2803 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2804
2805 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2806 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2807 so the version is shown; (c) stop "make test" echoing "./RunTest".
2808
2809 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2810
2811 9. The prototype for memmove() for systems that don't have it was using
2812 size_t, but the inclusion of the header that defines size_t was later. I've
2813 moved the #includes for the C headers earlier to avoid this.
2814
2815 10. Added some adjustments to the code to make it easier to compiler on certain
2816 special systems:
2817
2818 (a) Some "const" qualifiers were missing.
2819 (b) Added the macro EXPORT before all exported functions; by default this
2820 is defined to be empty.
2821 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2822 that it reads its output file name as an argument instead of writing
2823 to the standard output and assuming this can be redirected.
2824
2825 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2826 class containing characters with values greater than 255, PCRE compilation
2827 went into a loop.
2828
2829 12. A recursive reference to a subpattern that was within another subpattern
2830 that had a minimum quantifier of zero caused PCRE to crash. For example,
2831 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2832 recursion. If the recursively-called subpattern itself had a zero repeat,
2833 that was OK.
2834
2835 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2836 buffer into which it was copied (for escape processing) was still set at
2837 1024, so long lines caused crashes.
2838
2839 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2840 "internal error: code overflow...". This applied to any character class
2841 that was followed by a possessive quantifier.
2842
2843 15. Modified the Makefile to add libpcre.la as a prerequisite for
2844 libpcreposix.la because I was told this is needed for a parallel build to
2845 work.
2846
2847 16. If a pattern that contained .* following optional items at the start was
2848 studied, the wrong optimizing data was generated, leading to matching
2849 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2850 matching string must start with a or b or c. The correct conclusion for
2851 this pattern is that a match can start with any character.
2852
2853
2854 Version 4.4 13-Aug-03
2855 ---------------------
2856
2857 1. In UTF-8 mode, a character class containing characters with values between
2858 127 and 255 was not handled correctly if the compiled pattern was studied.
2859 In fixing this, I have also improved the studying algorithm for such
2860 classes (slightly).
2861
2862 2. Three internal functions had redundant arguments passed to them. Removal
2863 might give a very teeny performance improvement.
2864
2865 3. Documentation bug: the value of the capture_top field in a callout is *one
2866 more than* the number of the hightest numbered captured substring.
2867
2868 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2869 in incorrectly linking with a previously installed version. They now link
2870 explicitly with libpcre.la.
2871
2872 5. configure.in no longer needs to recognize Cygwin specially.
2873
2874 6. A problem in pcre.in for Windows platforms is fixed.
2875
2876 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2877 pcretest, it used to include the size of the study block as part of its
2878 output. Unfortunately, the structure contains a field that has a different
2879 size on different hardware architectures. This meant that the tests that
2880 showed this size failed. As the block is currently always of a fixed size,
2881 this information isn't actually particularly useful in pcretest output, so
2882 I have just removed it.
2883
2884 8. Three pre-processor statements accidentally did not start in column 1.
2885 Sadly, there are *still* compilers around that complain, even though
2886 standard C has not required this for well over a decade. Sigh.
2887
2888 9. In pcretest, the code for checking callouts passed small integers in the
2889 callout_data field, which is a void * field. However, some picky compilers
2890 complained about the casts involved for this on 64-bit systems. Now
2891 pcretest passes the address of the small integer instead, which should get
2892 rid of the warnings.
2893
2894 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2895 both compile and run time, and gives an error if an invalid UTF-8 sequence
2896 is found. There is a option for disabling this check in cases where the
2897 string is known to be correct and/or the maximum performance is wanted.
2898
2899 11. In response to a bug report, I changed one line in Makefile.in from
2900
2901 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2902 to
2903 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2904
2905 to look similar to other lines, but I have no way of telling whether this
2906 is the right thing to do, as I do not use Windows. No doubt I'll get told
2907 if it's wrong...
2908
2909
2910 Version 4.3 21-May-03
2911 ---------------------
2912
2913 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2914 Makefile.
2915
2916 2. Some refactoring to improve the quality of the code:
2917
2918 (i) The utf8_table... variables are now declared "const".
2919
2920 (ii) The code for \cx, which used the "case flipping" table to upper case
2921 lower case letters, now just substracts 32. This is ASCII-specific,
2922 but the whole concept of \cx is ASCII-specific, so it seems
2923 reasonable.
2924
2925 (iii) PCRE was using its character types table to recognize decimal and
2926 hexadecimal digits in the pattern. This is silly, because it handles
2927 only 0-9, a-f, and A-F, but the character types table is locale-
2928 specific, which means strange things might happen. A private
2929 table is now used for this - though it costs 256 bytes, a table is
2930 much faster than multiple explicit tests. Of course, the standard
2931 character types table is still used for matching digits in subject
2932 strings against \d.
2933
2934 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2935 ending in _t are). So I've renamed it as ESC_tee.
2936
2937 3. The first argument for regexec() in the POSIX wrapper should have been
2938 defined as "const".
2939
2940 4. Changed pcretest to use malloc() for its buffers so that they can be
2941 Electric Fenced for debugging.
2942
2943 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2944 to read one or more bytes before the start of the subject string. Often this
2945 had no effect on PCRE's behaviour, but in some circumstances it could
2946 provoke a segmentation fault.
2947
2948 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2949 to try to read one or more bytes before the start of the subject string.
2950
2951 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2952 UTF-8 support could misbehave in various ways if the subject string
2953 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2954 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2955 back over UTF-8 characters.)
2956
2957
2958 Version 4.2 14-Apr-03
2959 ---------------------
2960
2961 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2962
2963 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2964 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2965 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2966 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2967 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2968 and BUILD_EXEEXT
2969 Note: automatic setting of the BUILD variables is not yet working
2970 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2971 compile-time but not at link-time
2972 [LINK]: use for linking executables only
2973 make different versions for Windows and non-Windows
2974 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2975 libraries
2976 [LINK_FOR_BUILD]: new variable
2977 [OBJEXT]: use throughout
2978 [EXEEXT]: use throughout
2979 <winshared>: new target
2980 <wininstall>: new target
2981 <dftables.o>: use native compiler
2982 <dftables>: use native linker
2983 <install>: handle Windows platform correctly
2984 <clean>: ditto
2985 <check>: ditto
2986 copy DLL to top builddir before testing
2987
2988 As part of these changes, -no-undefined was removed again. This was reported
2989 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2990 in any case.
2991
2992 3. Some tidies to get rid of compiler warnings:
2993
2994 . In the match_data structure, match_limit was an unsigned long int, whereas
2995 match_call_count was an int. I've made them both unsigned long ints.
2996
2997 . In pcretest the fact that a const uschar * doesn't automatically cast to
2998 a void * provoked a warning.
2999
3000 . Turning on some more compiler warnings threw up some "shadow" variables
3001 and a few more missing casts.
3002
3003 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3004 option, a class that contained a single character with a value between 128
3005 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3006
3007 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3008 option, a class that contained several characters, but with at least one
3009 whose value was between 128 and 255 caused PCRE to crash.
3010
3011
3012 Version 4.1 12-Mar-03
3013 ---------------------
3014
3015 1. Compiling with gcc -pedantic found a couple of places where casts were
3016 needed, and a string in dftables.c that was longer than standard compilers are
3017 required to support.
3018
3019 2. Compiling with Sun's compiler found a few more places where the code could
3020 be tidied up in order to avoid warnings.
3021
3022 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3023 first of these names is deprecated in the latest Autoconf in favour of the name
3024 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3025 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3026 analogy I have changed it to CFLAGS_FOR_BUILD.
3027
3028 4. Added -no-undefined to the linking command in the Makefile, because this is
3029 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3030 linking step for the pcreposix library.
3031
3032 5. PCRE was failing to diagnose the case of two named groups with the same
3033 name.
3034
3035 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3036 literal character that is needed in the subject for a match, and scans along to
3037 ensure that it is present before embarking on the full matching process. This
3038 saves time in cases of nested unlimited repeats that are never going to match.
3039 Problem: the scan can take a lot of time if the subject is very long (e.g.
3040 megabytes), thus penalizing straightforward matches. It is now done only if the
3041 amount of subject to be scanned is less than 1000 bytes.
3042
3043 7. A lesser problem with the same optimization is that it was recording the
3044 first character of an anchored pattern as "needed", thus provoking a search
3045 right along the subject, even when the first match of the pattern was going to
3046 fail. The "needed" character is now not set for anchored patterns, unless it
3047 follows something in the pattern that is of non-fixed length. Thus, it still
3048 fulfils its original purpose of finding quick non-matches in cases of nested
3049 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3050
3051
3052 Version 4.0 17-Feb-03
3053 ---------------------
3054
3055 1. If a comment in an extended regex that started immediately after a meta-item
3056 extended to the end of string, PCRE compiled incorrect data. This could lead to
3057 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3058
3059 2. Moved to autoconf 2.53 and libtool 1.4.2.
3060
3061 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3062 the special perltest8 script is no longer needed - all the tests can be run
3063 from a single perltest script.
3064
3065 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3066 by \s. It has now been removed in PCRE. This means it isn't recognized as
3067 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3068 class [:space:] *does* include VT, thereby creating a mess.
3069
3070 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3071 space and tab.
3072
3073 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3074 its new features into the main test script, reducing the number of scripts.
3075
3076 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3077 were backward compatible, and made the (?i) apply to the whole pattern, as if
3078 /i were given. Now it behaves more logically, and applies the option setting
3079 only to what follows. PCRE has been changed to follow suit. However, if it
3080 finds options settings right at the start of the pattern, it extracts them into
3081 the global options, as before. Thus, they show up in the info data.
3082
3083 8. Added support for the \Q...\E escape sequence. Characters in between are
3084 treated as literals. This is slightly different from Perl in that $ and @ are
3085 also handled as literals inside the quotes. In Perl, they will cause variable
3086 interpolation. Note the following examples:
3087
3088 Pattern PCRE matches Perl matches
3089
3090 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3091 \Qabc\$xyz\E abc\$xyz abc\$xyz
3092 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3093
3094 For compatibility with Perl, \Q...\E sequences are recognized inside character
3095 classes as well as outside them.
3096
3097 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3098 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3099 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3100 signed/unsigned warnings.
3101
3102 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3103 option for pcretest, so I've replaced it by a simple function that does just
3104 that job.
3105
3106 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3107 "pcregrep -".
3108
3109 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3110 Java package. This provides some syntactic sugar for simple cases of what my
3111 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3112 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3113 item, you can use this simplified notation. Note that only makes sense with
3114 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3115 greediness, whatever the setting of the PCRE_UNGREEDY option.
3116
3117 13. A change of greediness default within a pattern was not taking effect at
3118 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3119 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3120 was abstracted outside.
3121
3122 14. PCRE now supports the \G assertion. It is true when the current matching
3123 position is at the start point of the match. This differs from \A when the
3124 starting offset is non-zero. Used with the /g option of pcretest (or similar
3125 code), it works in the same way as it does for Perl's /g option. If all
3126 alternatives of a regex begin with \G, the expression is anchored to the start
3127 match position, and the "anchored" flag is set in the compiled expression.
3128
3129 15. Some bugs concerning the handling of certain option changes within patterns
3130 have been fixed. These applied to options other than (?ims). For example,
3131 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3132 been the other way round. Some of this was related to change 7 above.
3133
3134 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3135 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3136 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3137 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3138
3139 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3140 mode. Unlike ".", it always matches newline, whatever the setting of
3141 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3142 assertions. Perl allows it, but it doesn't (in general) work because it can't
3143 calculate the length of the lookbehind. At least, that's the case for Perl
3144 5.8.0 - I've been told they are going to document that it doesn't work in
3145 future.
3146
3147 18. Added an error diagnosis for escapes that PCRE does not support: these are
3148 \L, \l, \N, \P, \p, \U, \u, and \X.
3149
3150 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3151 reading past the end of the pattern in cases such as /[abcd/.
3152
3153 20. PCRE was getting more memory than necessary for patterns with classes that
3154 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3155
3156 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3157 compiling PCRE for use with Virtual Pascal.
3158
3159 22. Small fix to the Makefile to make it work properly if the build is done
3160 outside the source tree.
3161
3162 23. Added a new extension: a condition to go with recursion. If a conditional
3163 subpattern starts with (?(R) the "true" branch is used if recursion has
3164 happened, whereas the "false" branch is used only at the top level.
3165
3166 24. When there was a very long string of literal characters (over 255 bytes
3167 without UTF support, over 250 bytes with UTF support), the computation of how
3168 much memory was required could be incorrect, leading to segfaults or other
3169 strange effects.
3170
3171 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3172 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3173 there was a subsequent back reference to those brackets. This meant that, for
3174 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3175 possible to check for precisely this case. All we can do is abandon the
3176 optimization if .* occurs inside capturing brackets when there are any back
3177 references whatsoever. (See below for a better fix that came later.)
3178
3179 26. The handling of the optimization for finding the first character of a
3180 non-anchored pattern, and for finding a character that is required later in the
3181 match were failing in some cases. This didn't break the matching; it just
3182 failed to optimize when it could. The way this is done has been re-implemented.
3183
3184 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3185
3186 28. Added a new feature that provides some of the functionality that Perl
3187 provides with (?{...}). The facility is termed a "callout". The way it is done
3188 in PCRE is for the caller to provide an optional function, by setting
3189 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3190 global variable. By default it is unset, which disables all calling out. To get
3191 the function called, the regex must include (?C) at appropriate points. This
3192 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3193 This provides a means of identifying different callout points. When PCRE
3194 reaches such a point in the regex, if pcre_callout has been set, the external
3195 function is called. It is provided with data in a structure called
3196 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3197 matching continues; if it returns a non-zero value, the match at the current
3198 point fails. However, backtracking will occur if possible. [This was changed
3199 later and other features added - see item 49 below.]
3200
3201 29. pcretest is upgraded to test the callout functionality. It provides a
3202 callout function that displays information. By default, it shows the start of
3203 the match and the current position in the text. There are some new data escapes
3204 to vary what happens:
3205
3206 \C+ in addition, show current contents of captured substrings
3207 \C- do not supply a callout function
3208 \C!n return 1 when callout number n is reached
3209 \C!n!m return 1 when callout number n is reached for the mth time
3210
3211 30. If pcregrep was called with the -l option and just a single file name, it
3212 output "<stdin>" if a match was found, instead of the file name.
3213
3214 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3215 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3216 pcre_exec(). This saves a malloc/free per call. The default value of
3217 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3218 when configuring.
3219
3220 32. The default maximum size of a compiled pattern is 64K. There have been a
3221 few cases of people hitting this limit. The code now uses macros to handle the
3222 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3223 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3224 configuring. Tests 2 and 5 work only with 2-byte links because they output
3225 debugging information about compiled patterns.
3226
3227 33. Internal code re-arrangements:
3228
3229 (a) Moved the debugging function for printing out a compiled regex into
3230 its own source file (printint.c) and used #include to pull it into
3231 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3232 separate copies.
3233
3234 (b) Defined the list of op-code names for debugging as a macro in
3235 internal.h so that it is next to the definition of the opcodes.
3236
3237 (c) Defined a table of op-code lengths for simpler skipping along compiled
3238 code. This is again a macro in internal.h so that it is next to the
3239 definition of the opcodes.
3240
3241 34. Added support for recursive calls to individual subpatterns, along the
3242 lines of Robin Houston's patch (but implemented somewhat differently).
3243
3244 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3245 allow it to read and process whole directories in Win32. This code was
3246 contributed by Lionel Fourquaux; it has not been tested by me.
3247
3248 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3249 used to name a group. Names consist of alphanumerics and underscores, and must
3250 be unique. Back references use the syntax (?P=name) and recursive calls use
3251 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3252 numbers. The function pcre_fullinfo() can be used after compilation to extract
3253 a name/number map. There are three relevant calls:
3254
3255 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3256 PCRE_INFO_NAMECOUNT yields the number of entries
3257 PCRE_INFO_NAMETABLE yields a pointer to the map.
3258
3259 The map is a vector of fixed-size entries. The size of each entry depends on
3260 the length of the longest name used. The first two bytes of each entry are the
3261 group number, most significant byte first. There follows the corresponding
3262 name, zero terminated. The names are in alphabetical order.
3263
3264 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3265 case instead of 255. Making it the same both with and without UTF-8 support
3266 means that the same test output works with both.
3267
3268 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3269 calling malloc() with a zero argument.
3270
3271 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3272 optimization. I've improved things by keeping a bitmap of backreferences with
3273 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3274 fact referenced, the optimization can be applied. It is unlikely that a
3275 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3276 the match to follow \n) will appear inside brackets with a number greater than
3277 31, but if it does, any back reference > 31 suppresses the optimization.
3278
3279 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3280 of disabling numbered capturing parentheses. Any opening parenthesis that is
3281 not followed by ? behaves as if it were followed by ?: but named parentheses
3282 can still be used for capturing (and they will acquire numbers in the usual
3283 way).
3284
3285 41. Redesigned the return codes from the match() function into yes/no/error so
3286 that errors can be passed back from deep inside the nested calls. A malloc
3287 failure while inside a recursive subpattern call now causes the
3288 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3289
3290 42. It is now possible to set a limit on the number of times the match()
3291 function is called in a call to pcre_exec(). This facility makes it possible to
3292 limit the amount of recursion and backtracking, though not in a directly
3293 obvious way, because the match() function is used in a number of different
3294 circumstances. The count starts from zero for each position in the subject
3295 string (for non-anchored patterns). The default limit is, for compatibility, a
3296 large number, namely 10 000 000. You can change this in two ways:
3297
3298 (a) When configuring PCRE before making, you can use --with-match-limit=n
3299 to set a default value for the compiled library.
3300
3301 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3302 a different value is set. See 45 below.
3303
3304 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3305
3306 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3307 of things that can be changed at compile time. The first argument specifies
3308 what is wanted and the second points to where the information is to be placed.
3309 The current list of available information is:
3310
3311 PCRE_CONFIG_UTF8
3312
3313 The output is an integer that is set to one if UTF-8 support is available;
3314 otherwise it is set to zero.
3315
3316 PCRE_CONFIG_NEWLINE
3317
3318 The output is an integer that it set to the value of the code that is used for
3319 newline. It is either LF (10) or CR (13).
3320
3321 PCRE_CONFIG_LINK_SIZE
3322
3323 The output is an integer that contains the number of bytes used for internal
3324 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3325
3326 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3327
3328 The output is an integer that contains the threshold above which the POSIX
3329 interface uses malloc() for output vectors. See item 31 above.
3330
3331 PCRE_CONFIG_MATCH_LIMIT
3332
3333 The output is an unsigned integer that contains the default limit of the number
3334 of match() calls in a pcre_exec() execution. See 42 above.
3335
3336 44. pcretest has been upgraded by the addition of the -C option. This causes it
3337 to extract all the available output from the new pcre_config() function, and to
3338 output it. The program then exits immediately.
3339
3340 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3341 order to support additional features. One way would have been to define
3342 pcre_exec2() (for example) with extra arguments, but this would not have been
3343 extensible, and would also have required all calls to the original function to
3344 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3345 is used for passing in "extra" data from pcre_study().
3346
3347 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3348 contains the following fields:
3349
3350 flags a bitmap indicating which of the following fields are set
3351 study_data opaque data from pcre_study()
3352 match_limit a way of specifying a limit on match() calls for a specific
3353 call to pcre_exec()
3354 callout_data data for callouts (see 49 below)
3355
3356 The flag bits are also defined in pcre.h, and are
3357
3358 PCRE_EXTRA_STUDY_DATA
3359 PCRE_EXTRA_MATCH_LIMIT
3360 PCRE_EXTRA_CALLOUT_DATA
3361
3362 The pcre_study() function now returns one of these new pcre_extra blocks, with
3363 the actual study data pointed to by the study_data field, and the
3364 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3365 before. That is, this change is entirely upwards-compatible and requires no
3366 change to existing code.
3367
3368 If you want to pass in additional data to pcre_exec(), you can either place it
3369 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3370 block.
3371
3372 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3373 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3374 times with different match limits, until it finds the minimum value needed for
3375 pcre_exec() to complete. The value is then output. This can be instructive; for
3376 most simple matches the number is quite small, but for pathological cases it
3377 gets very large very quickly.
3378
3379 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3380 returns the size of the data block pointed to by the study_data field in a
3381 pcre_extra block, that is, the value that was passed as the argument to
3382 pcre_malloc() when PCRE was getting memory in which to place the information
3383 created by pcre_study(). The fourth argument should point to a size_t variable.
3384 pcretest has been extended so that this information is shown after a successful
3385 pcre_study() call when information about the compiled regex is being displayed.
3386
3387 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3388 because what follows is always an absolute path. (Later: it turns out that this
3389 is more than cosmetic for MinGW, because it doesn't like empty path
3390 components.)
3391
3392 49. Some changes have been made to the callout feature (see 28 above):
3393
3394 (i) A callout function now has three choices for what it returns:
3395
3396 0 => success, carry on matching
3397 > 0 => failure at this point, but backtrack if possible
3398 < 0 => serious error, return this value from pcre_exec()
3399
3400 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3401 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3402 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3403 use by callout functions. It will never be used by PCRE itself.
3404
3405 (ii) The pcre_extra structure (see 45 above) has a void * field called
3406 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3407 pcre_callout_block structure has a field of the same name. The contents of
3408 the field passed in the pcre_extra structure are passed to the callout
3409 function in the corresponding field in the callout block. This makes it
3410 easier to use the same callout-containing regex from multiple threads. For
3411 testing, the pcretest program has a new data escape
3412
3413 \C*n pass the number n (may be negative) as callout_data
3414
3415 If the callout function in pcretest receives a non-zero value as
3416 callout_data, it returns that value.
3417
3418 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3419 there were some redundant $(CFLAGS) in commands that are now specified as
3420 $(LINK), which already includes $(CFLAGS).
3421
3422 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3423 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3424 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3425 one-byte characters throughout. Note that case-insensitive matching applies
3426 only to characters whose values are less than 256. PCRE doesn't support the
3427 notion of cases for higher-valued characters.
3428
3429 (i) A character class whose characters are all within 0-255 is handled as
3430 a bit map, and the map is inverted for negative classes. Previously, a
3431 character > 255 always failed to match such a class; however it should
3432 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3433
3434 (ii) A negated character class with a single character < 255 is coded as
3435 "not this character" (OP_NOT). This wasn't working properly when the test
3436 character was multibyte, either singly or repeated.
3437
3438 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3439 mode, for example: \x{100}{2,3}.
3440
3441 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3442 singly or repeated) now correctly test multibyte characters. However,
3443 PCRE doesn't recognize any characters with values greater than 255 as
3444 digits, spaces, or word characters. Such characters always match \D, \S,
3445 and \W, and never match \d, \s, or \w.
3446
3447 (v) Classes may now contain characters and character ranges with values
3448 greater than 255. For example: [ab\x{100}-\x{400}].
3449
3450 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3451 PCRE in UTF-8 mode.
3452
3453 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3454 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3455 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3456 value.)
3457
3458 53. The single man page has become too large. I have therefore split it up into
3459 a number of separate man pages. These also give rise to individual HTML pages;
3460 these are now put in a separate directory, and there is an index.html page that
3461 lists them all. Some hyperlinking between the pages has been installed.
3462
3463 54. Added convenience functions for handling named capturing parentheses.
3464
3465 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3466 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3467 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3468 are faulted.
3469
3470 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3471 calling configure. These values are used when compiling the dftables.c program
3472 which is run to generate the source of the default character tables. They
3473 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3474 you will need to set these values.
3475
3476 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3477
3478
3479 Version 3.9 02-Jan-02
3480 ---------------------
3481
3482 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3483
3484 2. If --disable-static was given, the building process failed when trying to
3485 build pcretest and pcregrep. (For some reason it was using libtool to compile
3486 them, which is not right, as they aren't part of the library.)
3487
3488
3489 Version 3.8 18-Dec-01
3490 ---------------------
3491
3492 1. The experimental UTF-8 code was completely screwed up. It was packing the
3493 bytes in the wrong order. How dumb can you get?
3494
3495
3496 Version 3.7 29-Oct-01
3497 ---------------------
3498
3499 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3500 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3501 this didn't happen under Solaris 8, where I normally test things.
3502
3503 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3504 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3505 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3506 this fix an hour or so after the initial 3.7 release.)
3507
3508
3509 Version 3.6 23-Oct-01
3510 ---------------------
3511
3512 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3513 offsets passed as NULL with zero offset count.
3514
3515 2. The config.guess and config.sub files had not been updated when I moved to
3516 the latest autoconf.
3517
3518
3519 Version 3.5 15-Aug-01
3520 ---------------------
3521
3522 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3523 had been forgotten.
3524
3525 2. By using declared but undefined structures, we can avoid using "void"
3526 definitions in pcre.h while keeping the internal definitions of the structures
3527 private.
3528
3529 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3530 user point of view, this means that both static and shared libraries are built
3531 by default, but this can be individually controlled. More of the work of
3532 handling this static/shared cases is now inside libtool instead of PCRE's make
3533 file.
3534
3535 4. The pcretest utility is now installed along with pcregrep because it is
3536 useful for users (to test regexs) and by doing this, it automatically gets
3537 relinked by libtool. The documentation has been turned into a man page, so
3538 there are now .1, .txt, and .html versions in /doc.
3539
3540 5. Upgrades to pcregrep:
3541 (i) Added long-form option names like gnu grep.
3542 (ii) Added --help to list all options with an explanatory phrase.
3543 (iii) Added -r, --recursive to recurse into sub-directories.
3544 (iv) Added -f, --file to read patterns from a file.
3545
3546 6. pcre_exec() was referring to its "code" argument before testing that
3547 argument for NULL (and giving an error if it was NULL).
3548
3549 7. Upgraded Makefile.in to allow for compiling in a different directory from
3550 the source directory.
3551
3552 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3553 options bits, the pointer it was passed was to an int instead of to an unsigned
3554 long int. This mattered only on 64-bit systems.
3555
3556 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3557 generated) instead of pcre.in, which it its source. Also made the same change
3558 in several of the .c files.
3559
3560 10. A new release of gcc defines printf() as a macro, which broke pcretest
3561 because it had an ifdef in the middle of a string argument for printf(). Fixed
3562 by using separate calls to printf().
3563
3564 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3565 script, to force use of CR or LF instead of \n in the source. On non-Unix
3566 systems, the value can be set in config.h.
3567
3568 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3569 absolute limit. Changed the text of the error message to make this clear, and
3570 likewise updated the man page.
3571
3572 13. The limit of 99 on the number of capturing subpatterns has been removed.
3573 The new limit is 65535, which I hope will not be a "real" limit.
3574
3575
3576 Version 3.4 22-Aug-00
3577 ---------------------
3578
3579 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3580
3581 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3582
3583
3584 Version 3.3 01-Aug-00
3585 ---------------------
3586
3587 1. If an octal character was given, but the value was greater than \377, it
3588 was not getting masked to the least significant bits, as documented. This could
3589 lead to crashes in some systems.
3590
3591 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3592 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3593
3594 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3595 These just pass their arguments on to (pcre_free)(), but they are provided
3596 because some uses of PCRE bind it to non-C systems that can call its functions,
3597 but cannot call free() or pcre_free() directly.
3598
3599 4. Add "make test" as a synonym for "make check". Corrected some comments in
3600 the Makefile.
3601
3602 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3603 Makefile.
3604
3605 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3606 command called pgrep for grepping around the active processes.
3607
3608 7. Added the beginnings of support for UTF-8 character strings.
3609
3610 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3611 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3612 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3613 out for the ar command.)
3614
3615
3616 Version 3.2 12-May-00
3617 ---------------------
3618
3619 This is purely a bug fixing release.
3620
3621 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3622 of ZA. This was just one example of several cases that could provoke this bug,
3623 which was introduced by change 9 of version 2.00. The code for breaking
3624 infinite loops after an iteration that matches an empty string was't working
3625 correctly.
3626
3627 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3628 when matched against abbab (for example). After matching an empty string, it
3629 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3630 caused it to match further down the string than it should.
3631
3632 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3633 was there because it doesn't seem to be needed, and it causes trouble on some
3634 systems, as it is not a Standard C header. It has been removed.
3635
3636 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3637 were reported on the Macintosh. The changes were from
3638
3639 while ((c = *(++ptr)) != 0 && c != '\n');
3640 to
3641 while ((c = *(++ptr)) != 0 && c != '\n') ;
3642
3643 Totally extraordinary, but if that's what it takes...
3644
3645 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3646 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3647 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3648 assumes the way PCRE uses memmove() (always moving upwards).
3649
3650 6. PCRE is being used in one environment where strchr() is not available. There
3651 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3652 faster code anyway.
3653
3654
3655 Version 3.1 09-Feb-00
3656 ---------------------
3657
3658 The only change in this release is the fixing of some bugs in Makefile.in for
3659 the "install" target:
3660
3661 (1) It was failing to install pcreposix.h.
3662
3663 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3664
3665
3666 Version 3.0 01-Feb-00
3667 ---------------------
3668
3669 1. Add support for the /+ modifier to perltest (to output $` like it does in
3670 pcretest).
3671
3672 2. Add support for the /g modifier to perltest.
3673
3674 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3675 matches null strings.
3676
3677 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3678 pattern. Perl treats empty patterns specially - it reuses the most recent
3679 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3680 effect.
3681
3682 5. The POSIX interface was broken in that it was just handing over the POSIX
3683 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3684 required a bigger vector, with some working space on the end. This means that
3685 the POSIX wrapper now has to get and free some memory, and copy the results.
3686
3687 6. Added some simple autoconf support, placing the test data and the
3688 documentation in separate directories, re-organizing some of the
3689 information files, and making it build pcre-config (a GNU standard). Also added
3690 libtool support for building PCRE as a shared library, which is now the
3691 default.
3692
3693 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3694 09 are not valid octal constants. Single digits will be used for minor values
3695 less than 10.
3696
3697 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3698 existing programs that set these in the POSIX interface can use PCRE without
3699 modification.
3700
3701 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3702 return all that pcre_info() returns, plus additional data. The pcre_info()
3703 function is retained for compatibility, but is considered to be obsolete.
3704
3705 10. Added experimental recursion feature (?R) to handle one common case that
3706 Perl 5.6 will be able to do with (?p{...}).
3707
3708 11. Added support for POSIX character classes like [:alpha:], which Perl is
3709 adopting.
3710
3711
3712 Version 2.08 31-Aug-99
3713 ----------------------
3714
3715 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3716 trying to match at the startoffset position, but instead was moving forward to
3717 the next newline as if a previous match had failed.
3718
3719 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3720 and could get into a loop if a null string was matched other than at the start
3721 of the subject.
3722
3723 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3724 be distinguished at compile time, and for completeness also added PCRE_DATE.
3725
3726 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3727 in GnuWin32 environments.
3728
3729
3730 Version 2.07 29-Jul-99
3731 ----------------------
3732
3733 1. The documentation is now supplied in plain text form and HTML as well as in
3734 the form of man page sources.
3735
3736 2. C++ compilers don't like assigning (void *) values to other pointer types.
3737 In particular this affects malloc(). Although there is no problem in Standard
3738 C, I've put in casts to keep C++ compilers happy.
3739
3740 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3741 should be (const char *).
3742
3743 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3744 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3745 However, I haven't made this a standard facility. The documentation doesn't
3746 mention it, and the Makefile doesn't support it.
3747
3748 5. The Makefile now contains an "install" target, with editable destinations at
3749 the top of the file. The pcretest program is not installed.
3750
3751 6. pgrep -V now gives the PCRE version number and date.
3752
3753 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3754 causing the entire string to be ignored, instead of just the last character.
3755
3756 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3757 non-matching string, it can take a very, very long time, even for strings of
3758 quite modest length, because of the nested recursion. PCRE now does better in
3759 some of these cases. It does this by remembering the last required literal
3760 character in the pattern, and pre-searching the subject to ensure it is present
3761 before running the real match. In other words, it applies a heuristic to detect
3762 some types of certain failure quickly, and in the above example, if presented
3763 with a string that has no trailing " it gives "no match" very quickly.
3764
3765 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3766 other alternatives are tried instead.
3767
3768
3769 Version 2.06 09-Jun-99
3770 ----------------------
3771
3772 1. Change pcretest's output for amount of store used to show just the code
3773 space, because the remainder (the data block) varies in size between 32-bit and
3774 64-bit systems.
3775
3776 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3777 start matching at. This allows lookbehinds to work when searching for multiple
3778 occurrences in a string.
3779
3780 3. Added additional options to pcretest for testing multiple occurrences:
3781
3782 /+ outputs the rest of the string that follows a match
3783 /g loops for multiple occurrences, using the new startoffset argument
3784 /G loops for multiple occurrences by passing an incremented pointer
3785
3786 4. PCRE wasn't doing the "first character" optimization for patterns starting
3787 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3788 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3789 the letter 'x'. On long subject strings, this gives a significant speed-up.
3790
3791
3792 Version 2.05 21-Apr-99
3793 ----------------------
3794
3795 1. Changed the type of magic_number from int to long int so that it works
3796 properly on 16-bit systems.
3797
3798 2. Fixed a bug which caused patterns starting with .* not to work correctly
3799 when the subject string contained newline characters. PCRE was assuming
3800 anchoring for such patterns in all cases, which is not correct because .* will
3801 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3802 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3803 must be retried after every newline in the subject.
3804
3805
3806 Version 2.04 18-Feb-99
3807 ----------------------
3808
3809 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3810 computation of the store needed to hold the pattern was incorrect (too large).
3811 If such patterns were nested a few deep, this could multiply and become a real
3812 problem.
3813
3814 2. Added /M option to pcretest to show the memory requirement of a specific
3815 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3816
3817 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3818 compiled in such a way that the backtracking after subsequent failure was
3819 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3820 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3821
3822
3823 Version 2.03 02-Feb-99
3824 ----------------------
3825
3826 1. Fixed typo and small mistake in man page.
3827
3828 2. Added 4th condition (GPL supersedes if conflict) and created separate
3829 LICENCE file containing the conditions.
3830
3831 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3832 Perl, that is the internal \ allows the delimiter to be included in the
3833 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3834 the final delimiter, add \ to the end of the pattern (to test the error).
3835
3836 4. Added the convenience functions for extracting substrings after a successful
3837 match. Updated pcretest to make it able to test these functions.
3838
3839
3840 Version 2.02 14-Jan-99
3841 ----------------------
3842
3843 1. Initialized the working variables associated with each extraction so that
3844 their saving and restoring doesn't refer to uninitialized store.
3845
3846 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3847 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3848 fix the problem.
3849
3850 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3851 calls, and wasn't printing the correct value for compiling calls. Increased the
3852 default value of LOOPREPEAT, and the number of significant figures in the
3853 times.
3854
3855 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3856
3857 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3858 a building problem on Windows NT with a FAT file system.
3859
3860
3861 Version 2.01 21-Oct-98
3862 ----------------------
3863
3864 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3865 to character tables built by pcre_maketables() in the current locale. If NULL
3866 is passed, the default tables are used.
3867
3868
3869 Version 2.00 24-Sep-98
3870 ----------------------
3871
3872 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3873 it any more.
3874
3875 2. Allow quantification of (?>) groups, and make it work correctly.
3876
3877 3. The first character computation wasn't working for (?>) groups.
3878
3879 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3880 end of the subject) and add 5.005's \z, which really does match only at the
3881 very end of the subject.
3882
3883 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3884
3885 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3886 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3887 localized options. All options to pcre_study() were also removed.
3888
3889 7. Add other new features from 5.005:
3890
3891 $(?<= positive lookbehind
3892 $(?<! negative lookbehind
3893 (?imsx-imsx) added the unsetting capability
3894 such a setting is global if at outer level; local otherwise
3895 (?imsx-imsx:) non-capturing groups with option setting
3896 (?(cond)re|re) conditional pattern matching
3897
3898 A backreference to itself in a repeated group matches the previous
3899 captured string.
3900
3901 8. General tidying up of studying (both automatic and via "study")
3902 consequential on the addition of new assertions.
3903
3904 9. As in 5.005, unlimited repeated groups that could match an empty substring
3905 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3906 runtime if any iteration does actually match an empty substring.
3907
3908 10. Include the RunTest script in the distribution.
3909
3910 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3911 discrepancies, some of which were old and were also with respect to 5.004. They
3912 have now been fixed.
3913
3914
3915 Version 1.09 28-Apr-98
3916 ----------------------
3917
3918 1. A negated single character class followed by a quantifier with a minimum
3919 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3920 program crashes, or just wrong answers. This did not apply to negated classes
3921 containing more than one character, or to minima other than one.
3922
3923
3924 Version 1.08 27-Mar-98
3925 ----------------------
3926
3927 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3928
3929 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3930 latter must appear before anything that relies on it in the pattern.
3931
3932
3933 Version 1.07 16-Feb-98
3934 ----------------------
3935
3936 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3937 repeat of a potentially empty string).
3938
3939
3940 Version 1.06 23-Jan-98
3941 ----------------------
3942
3943 1. Added Markus Oberhumer's little patches for C++.
3944
3945 2. Literal strings longer than 255 characters were broken.
3946
3947
3948 Version 1.05 23-Dec-97
3949 ----------------------
3950
3951 1. Negated character classes containing more than one character were failing if
3952 PCRE_CASELESS was set at run time.
3953
3954
3955 Version 1.04 19-Dec-97
3956 ----------------------
3957
3958 1. Corrected the man page, where some "const" qualifiers had been omitted.
3959
3960 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3961 input syntax.
3962
3963 3. Fixed memory leak which occurred when a regex with back references was
3964 matched with an offsets vector that wasn't big enough. The temporary memory
3965 that is used in this case wasn't being freed if the match failed.
3966
3967 4. Tidied pcretest to ensure it frees memory that it gets.
3968
3969 5. Temporary memory was being obtained in the case where the passed offsets
3970 vector was exactly big enough.
3971
3972 6. Corrected definition of offsetof() from change 5 below.
3973
3974 7. I had screwed up change 6 below and broken the rules for the use of
3975 setjmp(). Now fixed.
3976
3977
3978 Version 1.03 18-Dec-97
3979 ----------------------
3980
3981 1. A erroneous regex with a missing opening parenthesis was correctly
3982 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3983 on some systems.
3984
3985 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3986 it was reported that one broken compiler failed on the former because "code" is
3987 also an independent variable.
3988
3989 3. The erroneous regex a[]b caused an array overrun reference.
3990
3991 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3992 fail on data ending with that character. (It was going on too far, and checking
3993 the next character, typically a binary zero.) This was specific to the
3994 optimized code for single-character negative classes.
3995
3996 5. Added a contributed patch from the TIN world which does the following:
3997
3998 + Add an undef for memmove, in case the the system defines a macro for it.
3999
4000 + Add a definition of offsetof(), in case there isn't one. (I don't know
4001 the reason behind this - offsetof() is part of the ANSI standard - but
4002 it does no harm).
4003
4004 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4005 most of the places where whitespace preceded '#'. I have given up and
4006 allowed the remaining 2 cases to be at the margin.
4007
4008 + Rename some variables in pcre to eliminate shadowing. This seems very
4009 pedantic, but does no harm, of course.
4010
4011 6. Moved the call to setjmp() into its own function, to get rid of warnings
4012 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4013
4014 7. Constructs such as \d{8,} were compiling into the equivalent of
4015 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4016 outcome, but in this particular case used more store than had been allocated,
4017 which caused the bug to be discovered because it threw up an internal error.
4018
4019 8. The debugging code in both pcre and pcretest for outputting the compiled
4020 form of a regex was going wrong in the case of back references followed by
4021 curly-bracketed repeats.
4022
4023
4024 Version 1.02 12-Dec-97
4025 ----------------------
4026
4027 1. Typos in pcre.3 and comments in the source fixed.
4028
4029 2. Applied a contributed patch to get rid of places where it used to remove
4030 'const' from variables, and fixed some signed/unsigned and uninitialized
4031 variable warnings.
4032
4033 3. Added the "runtest" target to Makefile.
4034
4035 4. Set default compiler flag to -O2 rather than just -O.
4036
4037
4038 Version 1.01 19-Nov-97
4039 ----------------------
4040
4041 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4042 like /([ab]*)*/, that is, for classes with more than one character in them.
4043
4044 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4045 as /((?>a*))*/ (a PCRE_EXTRA facility).
4046
4047
4048 Version 1.00 18-Nov-97
4049 ----------------------
4050
4051 1. Added compile-time macros to support systems such as SunOS4 which don't have
4052 memmove() or strerror() but have other things that can be used instead.
4053
4054 2. Arranged that "make clean" removes the executables.
4055
4056
4057 Version 0.99 27-Oct-97
4058 ----------------------
4059
4060 1. Fixed bug in code for optimizing classes with only one character. It was
4061 initializing a 32-byte map regardless, which could cause it to run off the end
4062 of the memory it had got.
4063
4064 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4065
4066
4067 Version 0.98 22-Oct-97
4068 ----------------------
4069
4070 1. Fixed bug in code for handling temporary memory usage when there are more
4071 back references than supplied space in the ovector. This could cause segfaults.
4072
4073
4074 Version 0.97 21-Oct-97
4075 ----------------------
4076
4077 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4078
4079 2. Optimized negated single characters not to use a bit map.
4080
4081 3. Brought error texts together as macro definitions; clarified some of them;
4082 fixed one that was wrong - it said "range out of order" when it meant "invalid
4083 escape sequence".
4084
4085 4. Changed some char * arguments to const char *.
4086
4087 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4088
4089 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4090 pcretest.
4091
4092
4093 Version 0.96 16-Oct-97
4094 ----------------------
4095
4096 1. Added a simple "pgrep" utility to the distribution.
4097
4098 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4099 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4100 where "ddd" means "one or more decimal digits".
4101
4102 3. Fixed serious bug. If a pattern had a back reference, but the call to
4103 pcre_exec() didn't supply a large enough ovector to record the related
4104 identifying subpattern, the match always failed. PCRE now remembers the number
4105 of the largest back reference, and gets some temporary memory in which to save
4106 the offsets during matching if necessary, in order to ensure that
4107 backreferences always work.
4108
4109 4. Increased the compatibility with Perl in a number of ways:
4110
4111 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4112 to request this handling. The option can be set at compile or exec time.
4113
4114 (b) $ matches before a terminating newline by default; an option
4115 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4116 mode). The option can be set at compile or exec time.
4117
4118 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4119 the same as Perl's. If the decimal number it represents is less than 10
4120 or there aren't that many previous left capturing parentheses, an octal
4121 escape is read. Inside a character class, it's always an octal escape,
4122 even if it is a single digit.
4123
4124 (d) An escaped but undefined alphabetic character is taken as a literal,
4125 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4126 escapes.
4127
4128 (e) {0} is now permitted. (The previous item is removed from the compiled
4129 pattern).
4130
4131 5. Changed all the names of code files so that the basic parts are no longer
4132 than 10 characters, and abolished the teeny "globals.c" file.
4133
4134 6. Changed the handling of character classes; they are now done with a 32-byte
4135 bit map always.
4136
4137 7. Added the -d and /D options to pcretest to make it possible to look at the
4138 internals of compilation without having to recompile pcre.
4139
4140
4141 Version 0.95 23-Sep-97
4142 ----------------------
4143
4144 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4145 \x20 at the start of a run of normal characters. These were being treated as
4146 real characters, instead of the source characters being re-checked.
4147
4148
4149 Version 0.94 18-Sep-97
4150 ----------------------
4151
4152 1. The functions are now thread-safe, with the caveat that the global variables
4153 containing pointers to malloc() and free() or alternative functions are the
4154 same for all threads.
4155
4156 2. Get pcre_study() to generate a bitmap of initial characters for non-
4157 anchored patterns when this is possible, and use it if passed to pcre_exec().
4158
4159
4160 Version 0.93 15-Sep-97
4161 ----------------------
4162
4163 1. /(b)|(:+)/ was computing an incorrect first character.
4164
4165 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4166 but not actually doing anything yet.
4167
4168 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4169 as Perl does (e.g. [-az] or [az-]).
4170
4171 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4172 all possible positions.
4173
4174 5. Split up into different modules to avoid including unneeded functions in a
4175 compiled binary. However, compile and exec are still in one module. The "study"
4176 function is split off.
4177
4178 6. The character tables are now in a separate module whose source is generated
4179 by an auxiliary program - but can then be edited by hand if required. There are
4180 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4181 toupper() in the code.
4182
4183 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4184 make them global. Abolish the function for setting them, as the caller can now
4185 set them directly.
4186
4187
4188 Version 0.92 11-Sep-97
4189 ----------------------
4190
4191 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4192 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4193
4194 2. Caseless matching was not working in character classes if the characters in
4195 the pattern were in upper case.
4196
4197 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4198
4199 4. Make PCRE_ANCHORED public and accept as a compile option.
4200
4201 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4202 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4203 pass them.
4204
4205 6. Give an error if bad option bits passed at compile or run time.
4206
4207 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4208 pcretest to cause it to pass that flag.
4209
4210 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4211 options, and the first character, if set.
4212
4213 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4214
4215
4216 Version 0.91 10-Sep-97
4217 ----------------------
4218
4219 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4220 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4221
4222 2. PCRE was looping on encountering an indefinitely repeated back reference to
4223 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4224 Perl does - treats the match as successful.
4225
4226 ****

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