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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 960 - (show annotations)
Thu Apr 19 17:30:38 2012 UTC (3 years, 4 months ago) by ph10
File size: 229381 byte(s)
Error occurred while calculating annotation data.
Fix long-standing DFA testing restart bug in pcretest, and add some 
plausibility checks when restarting in pcre_dfa_exec().
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.31
5 -----------------------------
6
7 1. Fixing a wrong JIT test case and some compiler warnings.
8
9 2. Removed a bashism from the RunTest script.
10
11 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
12 to unsigned type, result still unsigned" that was given by an MS compiler
13 on encountering the code "-sizeof(xxx)".
14
15 4. Partial matching support is added to the JIT compiler.
16
17 5. Fixed several bugs concerned with partial matching of items that consist
18 of more than one character:
19
20 (a) /^(..)\1/ did not partially match "aba" because checking references was
21 done on an "all or nothing" basis. This also applied to repeated
22 references.
23
24 (b) \R did not give a hard partial match if \r was found at the end of the
25 subject.
26
27 (c) \X did not give a hard partial match after matching one or more
28 characters at the end of the subject.
29
30 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
31 a partial match for the string "\r".
32
33 (e) When newline was set to CRLF, the metacharacter "." did not recognize
34 a partial match for a CR character at the end of the subject string.
35
36 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
37 running pcretest, the text "(JIT)" added to the output whenever JIT is
38 actually used to run the match.
39
40 7. Individual JIT compile options can be set in pcretest by following -s+[+]
41 or /S+[+] with a digit between 1 and 7.
42
43 8. OP_NOT now supports any UTF character not just single-byte ones.
44
45 9. (*MARK) control verb is now supported by the JIT compiler.
46
47 10. The command "./RunTest list" lists the available tests without actually
48 running any of them. (Because I keep forgetting what they all are.)
49
50 11. Add PCRE_INFO_MAXLOOKBEHIND.
51
52 12. Applied a (slightly modified) user-supplied patch that improves performance
53 when the heap is used for recursion (compiled with --disable-stack-for-
54 recursion). Instead of malloc and free for each heap frame each time a
55 logical recursion happens, frames are retained on a chain and re-used where
56 possible. This sometimes gives as much as 30% improvement.
57
58 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
59 call.
60
61 14. As documented, (*COMMIT) is now confined to within a positive assertion.
62
63 15. It is now possible to link pcretest with libedit as an alternative to
64 libreadline.
65
66 16. (*COMMIT) control verb is now supported by the JIT compiler.
67
68 17. The Unicode data tables have been updated to Unicode 6.1.0.
69
70 18. Added --file-list option to pcregrep.
71
72 19. Added binary file support to pcregrep, including the -a, --binary-files,
73 -I, and --text options.
74
75 20. The madvise function is renamed for posix_madvise for QNX compatibility
76 reasons. Fixed by Giuseppe D'Angelo.
77
78 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
79
80 22. Changed the option for creating symbolic links for 16-bit man pages from
81 -s to -sf so that re-installing does not cause issues.
82
83 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
84
85 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
86 matches in certain environments (the workspace was not being correctly
87 retained). Also added to pcre_dfa_exec() a simple plausibility check on
88 some of the workspace data at the beginning of a restart.
89
90
91 Version 8.30 04-February-2012
92 -----------------------------
93
94 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
95 name is defined in ctype.h.
96
97 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
98 only in quite long subpatterns.
99
100 3. Removed the function pcre_info(), which has been obsolete and deprecated
101 since it was replaced by pcre_fullinfo() in February 2000.
102
103 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
104 match a (*MARK), and the match failed at the start of the subject, a
105 reference to memory before the start of the subject could occur. This bug
106 was introduced by fix 17 of release 8.21.
107
108 5. A reference to an unset group with zero minimum repetition was giving
109 totally wrong answers (in non-JavaScript-compatibility mode). For example,
110 /(another)?(\1?)test/ matched against "hello world test". This bug was
111 introduced in release 8.13.
112
113 6. Add support for 16-bit character strings (a large amount of work involving
114 many changes and refactorings).
115
116 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
117 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
118 from a file.
119
120 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
121 rounding is not applied in this particular case).
122
123 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
124 if they appear, or are escaped, in patterns.
125
126 10. Get rid of a number of -Wunused-but-set-variable warnings.
127
128 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
129 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
130 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
131 also returns the mark "x". This bug applied to capturing parentheses,
132 non-capturing parentheses, and atomic parentheses. It also applied to some
133 assertions.
134
135 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
136 information out of configure.ac instead of relying on pcre.h.generic, which
137 is not stored in the repository.
138
139 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
140 -lreadline.
141
142 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
143
144 15. Applied Graycode's patch to put the top-level frame on the stack rather
145 than the heap when not using the stack for recursion. This gives a
146 performance improvement in many cases when recursion is not deep.
147
148 16. Experimental code added to "pcretest -C" to output the stack frame size.
149
150
151 Version 8.21 12-Dec-2011
152 ------------------------
153
154 1. Updating the JIT compiler.
155
156 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
157 are added as well.
158
159 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
160 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
161 calling _pcre_jit_exec. Some extra comments are added.
162
163 4. (*MARK) settings inside atomic groups that do not contain any capturing
164 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
165 was introduced by change 18 for 8.20.
166
167 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
168 ECMA-262 standard.
169
170 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
171 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
172 This bug was probably introduced by change 9 of 8.13.
173
174 7. While fixing 6 above, I noticed that a number of other items were being
175 incorrectly rejected as "not fixed length". This arose partly because newer
176 opcodes had not been added to the fixed-length checking code. I have (a)
177 corrected the bug and added tests for these items, and (b) arranged for an
178 error to occur if an unknown opcode is encountered while checking for fixed
179 length instead of just assuming "not fixed length". The items that were
180 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
181 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
182 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
183
184 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
185 being incorrectly compiled and would have given unpredicatble results.
186
187 9. A possessively repeated subpattern with minimum repeat count greater than
188 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
189 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
190 the first (A) could occur when it should not.
191
192 10. Add a cast and remove a redundant test from the code.
193
194 11. JIT should use pcre_malloc/pcre_free for allocation.
195
196 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
197 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
198 is anything other than /usr, -L is still shown).
199
200 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
201
202 14. Perl does not support \N without a following name in a [] class; PCRE now
203 also gives an error.
204
205 15. If a forward reference was repeated with an upper limit of around 2000,
206 it caused the error "internal error: overran compiling workspace". The
207 maximum number of forward references (including repeats) was limited by the
208 internal workspace, and dependent on the LINK_SIZE. The code has been
209 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
210 the default depends on LINK_SIZE. There is a new upper limit (for safety)
211 of around 200,000 forward references. While doing this, I also speeded up
212 the filling in of repeated forward references.
213
214 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
215 incorrectly expecting the subject to contain another "a" after the start.
216
217 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
218 in the match, the SKIP should be ignored. This was not happening; instead
219 the SKIP was being treated as NOMATCH. For patterns such as
220 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
221 tested.
222
223 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
224 now much more compatible with Perl, in particular in cases where the result
225 is a non-match for a non-anchored pattern. For example, if
226 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
227 "m", where previously it did not return a name. A side effect of this
228 change is that for partial matches, the last encountered mark name is
229 returned, as for non matches. A number of tests that were previously not
230 Perl-compatible have been moved into the Perl-compatible test files. The
231 refactoring has had the pleasing side effect of removing one argument from
232 the match() function, thus reducing its stack requirements.
233
234 19. If the /S+ option was used in pcretest to study a pattern using JIT,
235 subsequent uses of /S (without +) incorrectly behaved like /S+.
236
237 21. Retrieve executable code size support for the JIT compiler and fixing
238 some warnings.
239
240 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
241 not work when the shorter character appeared right at the end of the
242 subject string.
243
244 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
245 systems.
246
247 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
248 output it when the /M option is used in pcretest.
249
250 25. The CheckMan script was not being included in the distribution. Also, added
251 an explicit "perl" to run Perl scripts from the PrepareRelease script
252 because this is reportedly needed in Windows.
253
254 26. If study data was being save in a file and studying had not found a set of
255 "starts with" bytes for the pattern, the data written to the file (though
256 never used) was taken from uninitialized memory and so caused valgrind to
257 complain.
258
259 27. Updated RunTest.bat as provided by Sheri Pierce.
260
261 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
262
263 29. Computation of memory usage for the table of capturing group names was
264 giving an unnecessarily large value.
265
266
267 Version 8.20 21-Oct-2011
268 ------------------------
269
270 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
271 a POSIX class. After further experiments with Perl, which convinced me that
272 Perl has bugs and confusions, a closing square bracket is no longer allowed
273 in a POSIX name. This bug also affected patterns with classes that started
274 with full stops.
275
276 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
277 captured substring, but while checking the failing first alternative,
278 substring 1 is temporarily captured. If the output vector supplied to
279 pcre_exec() was not big enough for this capture, the yield of the function
280 was still zero ("insufficient space for captured substrings"). This cannot
281 be totally fixed without adding another stack variable, which seems a lot
282 of expense for a edge case. However, I have improved the situation in cases
283 such as /(a)(b)x|abc/ matched against "abc", where the return code
284 indicates that fewer than the maximum number of slots in the ovector have
285 been set.
286
287 3. Related to (2) above: when there are more back references in a pattern than
288 slots in the output vector, pcre_exec() uses temporary memory during
289 matching, and copies in the captures as far as possible afterwards. It was
290 using the entire output vector, but this conflicts with the specification
291 that only 2/3 is used for passing back captured substrings. Now it uses
292 only the first 2/3, for compatibility. This is, of course, another edge
293 case.
294
295 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
296 main code base, and can be used by building with --enable-jit. When this is
297 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
298 runtime --no-jit option is given.
299
300 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
301 ovector, the return from the function was zero, implying that there were
302 other matches that did not fit. The correct "exactly full" value is now
303 returned.
304
305 6. If a subpattern that was called recursively or as a subroutine contained
306 (*PRUNE) or any other control that caused it to give a non-standard return,
307 invalid errors such as "Error -26 (nested recursion at the same subject
308 position)" or even infinite loops could occur.
309
310 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
311 computing the minimum length on reaching *ACCEPT, and so ended up with the
312 wrong value of 1 rather than 0. Further investigation indicates that
313 computing a minimum subject length in the presence of *ACCEPT is difficult
314 (think back references, subroutine calls), and so I have changed the code
315 so that no minimum is registered for a pattern that contains *ACCEPT.
316
317 8. If (*THEN) was present in the first (true) branch of a conditional group,
318 it was not handled as intended. [But see 16 below.]
319
320 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
321 Sheri Pierce.
322
323 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
324 the first byte in a match must be "a".
325
326 11. Change 17 for 8.13 increased the recursion depth for patterns like
327 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
328 pattern contains any instances of (*THEN). If it does not, the old
329 optimizations are restored. It would be nice to do this on a per-group
330 basis, but at the moment that is not feasible.
331
332 12. In some environments, the output of pcretest -C is CRLF terminated. This
333 broke RunTest's code that checks for the link size. A single white space
334 character after the value is now allowed for.
335
336 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
337 For "fr", it uses the Windows-specific input and output files.
338
339 14. If (*THEN) appeared in a group that was called recursively or as a
340 subroutine, it did not work as intended. [But see next item.]
341
342 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
343 pattern fragments (but not containing any | characters). If A and B are
344 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
345 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
346 D. In other words, Perl considers parentheses that do not contain any |
347 characters to be part of a surrounding alternative, whereas PCRE was
348 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
349 differently. PCRE now behaves in the same way as Perl, except in the case
350 of subroutine/recursion calls such as (?1) which have in any case always
351 been different (but PCRE had them first :-).
352
353 16. Related to 15 above: Perl does not treat the | in a conditional group as
354 creating alternatives. Such a group is treated in the same way as an
355 ordinary group without any | characters when processing (*THEN). PCRE has
356 been changed to match Perl's behaviour.
357
358 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
359 RunGrepTest script failed.
360
361 18. Change 22 for version 13 caused atomic groups to use more stack. This is
362 inevitable for groups that contain captures, but it can lead to a lot of
363 stack use in large patterns. The old behaviour has been restored for atomic
364 groups that do not contain any capturing parentheses.
365
366 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
367 suppress the check for a minimum subject length at run time. (If it was
368 given to pcre_exec() or pcre_dfa_exec() it did work.)
369
370 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
371 fail to work when decoding hex characters in data strings in EBCDIC
372 environments.
373
374 21. It appears that in at least one Mac OS environment, the isxdigit() function
375 is implemented as a macro that evaluates to its argument more than once,
376 contravening the C 90 Standard (I haven't checked a later standard). There
377 was an instance in pcretest which caused it to go wrong when processing
378 \x{...} escapes in subject strings. The has been rewritten to avoid using
379 things like p++ in the argument of isxdigit().
380
381
382 Version 8.13 16-Aug-2011
383 ------------------------
384
385 1. The Unicode data tables have been updated to Unicode 6.0.0.
386
387 2. Two minor typos in pcre_internal.h have been fixed.
388
389 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
390 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
391 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
392
393 4. There were a number of related bugs in the code for matching backrefences
394 caselessly in UTF-8 mode when codes for the characters concerned were
395 different numbers of bytes. For example, U+023A and U+2C65 are an upper
396 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
397 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
398 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
399 2-byte code at the end of the subject (it thought there wasn't enough data
400 left).
401
402 5. Comprehensive information about what went wrong is now returned by
403 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
404 as the output vector has at least 2 elements. The offset of the start of
405 the failing character and a reason code are placed in the vector.
406
407 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
408 now returned is for the first byte of the failing character, instead of the
409 last byte inspected. This is an incompatible change, but I hope it is small
410 enough not to be a problem. It makes the returned offset consistent with
411 pcre_exec() and pcre_dfa_exec().
412
413 7. pcretest now gives a text phrase as well as the error number when
414 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
415 failure, the offset and reason code are output.
416
417 8. When \R was used with a maximizing quantifier it failed to skip backwards
418 over a \r\n pair if the subsequent match failed. Instead, it just skipped
419 back over a single character (\n). This seems wrong (because it treated the
420 two characters as a single entity when going forwards), conflicts with the
421 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
422 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
423 has been changed.
424
425 9. Some internal refactoring has changed the processing so that the handling
426 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
427 time (the PCRE_DOTALL option was changed this way some time ago: version
428 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
429 which was always a bit of a fudge. It also means that there is one less
430 argument for the match() function, which reduces its stack requirements
431 slightly. This change also fixes an incompatibility with Perl: the pattern
432 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
433
434 10. More internal refactoring has drastically reduced the number of recursive
435 calls to match() for possessively repeated groups such as (abc)++ when
436 using pcre_exec().
437
438 11. While implementing 10, a number of bugs in the handling of groups were
439 discovered and fixed:
440
441 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
442 (a|)*(?1) gave a compile-time internal error.
443 ((a|)+)+ did not notice that the outer group could match an empty string.
444 (^a|^)+ was not marked as anchored.
445 (.*a|.*)+ was not marked as matching at start or after a newline.
446
447 12. Yet more internal refactoring has removed another argument from the match()
448 function. Special calls to this function are now indicated by setting a
449 value in a variable in the "match data" data block.
450
451 13. Be more explicit in pcre_study() instead of relying on "default" for
452 opcodes that mean there is no starting character; this means that when new
453 ones are added and accidentally left out of pcre_study(), testing should
454 pick them up.
455
456 14. The -s option of pcretest has been documented for ages as being an old
457 synonym of -m (show memory usage). I have changed it to mean "force study
458 for every regex", that is, assume /S for every regex. This is similar to -i
459 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
460 using it. It makes it easier to run collections of tests with and without
461 study enabled, and thereby test pcre_study() more easily. All the standard
462 tests are now run with and without -s (but some patterns can be marked as
463 "never study" - see 20 below).
464
465 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
466 restoration of the capturing data to the outer values was not happening
467 correctly.
468
469 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
470 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
471 pattern had matched an empty string, and so incorrectly returned a no
472 match.
473
474 17. There was optimizing code for the last branch of non-capturing parentheses,
475 and also for the obeyed branch of a conditional subexpression, which used
476 tail recursion to cut down on stack usage. Unfortunately, now that there is
477 the possibility of (*THEN) occurring in these branches, tail recursion is
478 no longer possible because the return has to be checked for (*THEN). These
479 two optimizations have therefore been removed. [But see 8.20/11 above.]
480
481 18. If a pattern containing \R was studied, it was assumed that \R always
482 matched two bytes, thus causing the minimum subject length to be
483 incorrectly computed because \R can also match just one byte.
484
485 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
486 was incorrectly computed.
487
488 20. If /S is present twice on a test pattern in pcretest input, it now
489 *disables* studying, thereby overriding the use of -s on the command line
490 (see 14 above). This is necessary for one or two tests to keep the output
491 identical in both cases.
492
493 21. When (*ACCEPT) was used in an assertion that matched an empty string and
494 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
495
496 22. When an atomic group that contained a capturing parenthesis was
497 successfully matched, but the branch in which it appeared failed, the
498 capturing was not being forgotten if a higher numbered group was later
499 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
500 group 1 to "a", when in fact it should be unset. This applied to multi-
501 branched capturing and non-capturing groups, repeated or not, and also to
502 positive assertions (capturing in negative assertions does not happen
503 in PCRE) and also to nested atomic groups.
504
505 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
506 subject after a captured substring, to make it easier to tell which of a
507 number of identical substrings has been captured.
508
509 24. The way atomic groups are processed by pcre_exec() has been changed so that
510 if they are repeated, backtracking one repetition now resets captured
511 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
512 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
513 "aaa". Previously, it would have been "a". As part of this code
514 refactoring, the way recursive calls are handled has also been changed.
515
516 25. If an assertion condition captured any substrings, they were not passed
517 back unless some other capturing happened later. For example, if
518 (?(?=(a))a) was matched against "a", no capturing was returned.
519
520 26. When studying a pattern that contained subroutine calls or assertions,
521 the code for finding the minimum length of a possible match was handling
522 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
523 group 1 called group 2 while simultaneously a separate group 2 called group
524 1). A stack overflow occurred in this case. I have fixed this by limiting
525 the recursion depth to 10.
526
527 27. Updated RunTest.bat in the distribution to the version supplied by Tom
528 Fortmann. This supports explicit test numbers on the command line, and has
529 argument validation and error reporting.
530
531 28. An instance of \X with an unlimited repeat could fail if at any point the
532 first character it looked at was a mark character.
533
534 29. Some minor code refactoring concerning Unicode properties and scripts
535 should reduce the stack requirement of match() slightly.
536
537 30. Added the '=' option to pcretest to check the setting of unused capturing
538 slots at the end of the pattern, which are documented as being -1, but are
539 not included in the return count.
540
541 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
542 compiled something random. Now it gives a compile-time error (as does
543 Perl).
544
545 32. A *MARK encountered during the processing of a positive assertion is now
546 recorded and passed back (compatible with Perl).
547
548 33. If --only-matching or --colour was set on a pcregrep call whose pattern
549 had alternative anchored branches, the search for a second match in a line
550 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
551 matched the line "0102" twice. The same bug affected patterns that started
552 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
553 twice.
554
555 34. Previously, PCRE did not allow quantification of assertions. However, Perl
556 does, and because of capturing effects, quantifying parenthesized
557 assertions may at times be useful. Quantifiers are now allowed for
558 parenthesized assertions.
559
560 35. A minor code tidy in pcre_compile() when checking options for \R usage.
561
562 36. \g was being checked for fancy things in a character class, when it should
563 just be a literal "g".
564
565 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
566 appearance of a nested POSIX class supersedes an apparent external class.
567 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
568 unescaped square brackets may also appear as part of class names. For
569 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
570 more like Perl. (But see 8.20/1 above.)
571
572 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
573 was because it thought it was \N{name}, which is not supported).
574
575 39. Add minix to OS list not supporting the -S option in pcretest.
576
577 40. PCRE tries to detect cases of infinite recursion at compile time, but it
578 cannot analyze patterns in sufficient detail to catch mutual recursions
579 such as ((?1))((?2)). There is now a runtime test that gives an error if a
580 subgroup is called recursively as a subpattern for a second time at the
581 same position in the subject string. In previous releases this might have
582 been caught by the recursion limit, or it might have run out of stack.
583
584 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
585 happen only once. PCRE was, however incorrectly giving a compile time error
586 "recursive call could loop indefinitely" because it cannot analyze the
587 pattern in sufficient detail. The compile time test no longer happens when
588 PCRE is compiling a conditional subpattern, but actual runaway loops are
589 now caught at runtime (see 40 above).
590
591 42. It seems that Perl allows any characters other than a closing parenthesis
592 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
593 has been changed to be the same.
594
595 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
596 as not to get warnings when autogen.sh is called. Also changed
597 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
598
599 44. To help people who use pcregrep to scan files containing exceedingly long
600 lines, the following changes have been made:
601
602 (a) The default value of the buffer size parameter has been increased from
603 8K to 20K. (The actual buffer used is three times this size.)
604
605 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
606 PCRE is built.
607
608 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
609 to be set at run time.
610
611 (d) Numerical values in pcregrep options can be followed by K or M, for
612 example --buffer-size=50K.
613
614 (e) If a line being scanned overflows pcregrep's buffer, an error is now
615 given and the return code is set to 2.
616
617 45. Add a pointer to the latest mark to the callout data block.
618
619 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
620 partial match of an empty string instead of no match. This was specific to
621 the use of ".".
622
623 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
624 complete match instead of a partial match. This bug was dependent on both
625 the PCRE_UTF8 and PCRE_DOTALL options being set.
626
627 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
628 starting byte set, because \b was not being ignored.
629
630
631 Version 8.12 15-Jan-2011
632 ------------------------
633
634 1. Fixed some typos in the markup of the man pages, and wrote a script that
635 checks for such things as part of the documentation building process.
636
637 2. On a big-endian 64-bit system, pcregrep did not correctly process the
638 --match-limit and --recursion-limit options (added for 8.11). In
639 particular, this made one of the standard tests fail. (The integer value
640 went into the wrong half of a long int.)
641
642 3. If the --colour option was given to pcregrep with -v (invert match), it
643 did strange things, either producing crazy output, or crashing. It should,
644 of course, ignore a request for colour when reporting lines that do not
645 match.
646
647 4. Another pcregrep bug caused similar problems if --colour was specified with
648 -M (multiline) and the pattern match finished with a line ending.
649
650 5. In pcregrep, when a pattern that ended with a literal newline sequence was
651 matched in multiline mode, the following line was shown as part of the
652 match. This seems wrong, so I have changed it.
653
654 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
655 the check for further matches in the same line (so they could be coloured)
656 to overrun the end of the current line. If another match was found, it was
657 incorrectly shown (and then shown again when found in the next line).
658
659 7. If pcregrep was compiled under Windows, there was a reference to the
660 function pcregrep_exit() before it was defined. I am assuming this was
661 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
662 reported by a user. I've moved the definition above the reference.
663
664
665 Version 8.11 10-Dec-2010
666 ------------------------
667
668 1. (*THEN) was not working properly if there were untried alternatives prior
669 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
670 backtracked to try for "b" instead of moving to the next alternative branch
671 at the same level (in this case, to look for "c"). The Perl documentation
672 is clear that when (*THEN) is backtracked onto, it goes to the "next
673 alternative in the innermost enclosing group".
674
675 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
676 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
677 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
678 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
679 (*THEN).
680
681 3. If \s appeared in a character class, it removed the VT character from
682 the class, even if it had been included by some previous item, for example
683 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
684 of \s, but is part of the POSIX "space" class.)
685
686 4. A partial match never returns an empty string (because you can always
687 match an empty string at the end of the subject); however the checking for
688 an empty string was starting at the "start of match" point. This has been
689 changed to the "earliest inspected character" point, because the returned
690 data for a partial match starts at this character. This means that, for
691 example, /(?<=abc)def/ gives a partial match for the subject "abc"
692 (previously it gave "no match").
693
694 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
695 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
696 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
697 has an implication that the given string is incomplete (because a partial
698 match is preferred over a full match). For this reason, these items now
699 give a partial match in this situation. [Aside: previously, the one case
700 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
701 match rather than a full match, which was wrong by the old rules, but is
702 now correct.]
703
704 6. There was a bug in the handling of #-introduced comments, recognized when
705 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
706 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
707 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
708 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
709 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
710 places in pcre_compile().
711
712 7. Related to (6) above, when pcre_compile() was skipping #-introduced
713 comments when looking ahead for named forward references to subpatterns,
714 the only newline sequence it recognized was NL. It now handles newlines
715 according to the set newline convention.
716
717 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
718 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
719 cater for a lack of strerror(). These oversights have been fixed.
720
721 9. Added --match-limit and --recursion-limit to pcregrep.
722
723 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
724
725 11. When the -o option was used, pcregrep was setting a return code of 1, even
726 when matches were found, and --line-buffered was not being honoured.
727
728 12. Added an optional parentheses number to the -o and --only-matching options
729 of pcregrep.
730
731 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
732 can match an empty string. The code to do it in pcretest and pcredemo
733 needed fixing:
734
735 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
736 only one byte after an empty string match just before CRLF (this case
737 just got forgotten; "any" and "anycrlf" were OK).
738
739 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
740 mode when an empty string match preceded an ASCII character followed by
741 a non-ASCII character. (The code for advancing by one character rather
742 than one byte was nonsense.)
743
744 (c) The pcredemo.c sample program did not have any code at all to handle
745 the cases when CRLF is a valid newline sequence.
746
747 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
748 as a starting offset was within the subject string. There is now a new
749 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
750 negative or greater than the length of the string. In order to test this,
751 pcretest is extended to allow the setting of negative starting offsets.
752
753 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
754 starting offset points to the beginning of a UTF-8 character was
755 unnecessarily clumsy. I tidied it up.
756
757 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
758 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
759
760 17. Nobody had reported that the --include_dir option, which was added in
761 release 7.7 should have been called --include-dir (hyphen, not underscore)
762 for compatibility with GNU grep. I have changed it to --include-dir, but
763 left --include_dir as an undocumented synonym, and the same for
764 --exclude-dir, though that is not available in GNU grep, at least as of
765 release 2.5.4.
766
767 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
768 characters from a string of bytes) have been redefined so as not to use
769 loops, in order to improve performance in some environments. At the same
770 time, I abstracted some of the common code into auxiliary macros to save
771 repetition (this should not affect the compiled code).
772
773 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
774 compile-time error is now given if \c is not followed by an ASCII
775 character, that is, a byte less than 128. (In EBCDIC mode, the code is
776 different, and any byte value is allowed.)
777
778 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
779 START_OPTIMIZE option, which is now allowed at compile time - but just
780 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
781 to pcregrep and other applications that have no direct access to PCRE
782 options. The new /Y option in pcretest sets this option when calling
783 pcre_compile().
784
785 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
786 back references. Groups containing recursive back references were forced to
787 be atomic by that change, but in the case of named groups, the amount of
788 memory required was incorrectly computed, leading to "Failed: internal
789 error: code overflow". This has been fixed.
790
791 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
792 pcretest.c, to avoid build problems in some Borland environments.
793
794
795 Version 8.10 25-Jun-2010
796 ------------------------
797
798 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
799 THEN.
800
801 2. (*ACCEPT) was not working when inside an atomic group.
802
803 3. Inside a character class, \B is treated as a literal by default, but
804 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
805 causes the error). The code is unchanged, but I tidied the documentation.
806
807 4. Inside a character class, PCRE always treated \R and \X as literals,
808 whereas Perl faults them if its -w option is set. I have changed PCRE so
809 that it faults them when PCRE_EXTRA is set.
810
811 5. Added support for \N, which always matches any character other than
812 newline. (It is the same as "." when PCRE_DOTALL is not set.)
813
814 6. When compiling pcregrep with newer versions of gcc which may have
815 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
816 declared with attribute warn_unused_result" were given. Just casting the
817 result to (void) does not stop the warnings; a more elaborate fudge is
818 needed. I've used a macro to implement this.
819
820 7. Minor change to pcretest.c to avoid a compiler warning.
821
822 8. Added four artifical Unicode properties to help with an option to make
823 \s etc use properties (see next item). The new properties are: Xan
824 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
825
826 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
827 use Unicode properties. (*UCP) at the start of a pattern can be used to set
828 this option. Modified pcretest to add /W to test this facility. Added
829 REG_UCP to make it available via the POSIX interface.
830
831 10. Added --line-buffered to pcregrep.
832
833 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
834 studied, and the match started with a letter with a code point greater than
835 127 whose first byte was different to the first byte of the other case of
836 the letter, the other case of this starting letter was not recognized
837 (#976).
838
839 12. If a pattern that was studied started with a repeated Unicode property
840 test, for example, \p{Nd}+, there was the theoretical possibility of
841 setting up an incorrect bitmap of starting bytes, but fortunately it could
842 not have actually happened in practice until change 8 above was made (it
843 added property types that matched character-matching opcodes).
844
845 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
846 possible starting bytes for non-anchored patterns.
847
848 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
849 \R, and also a number of cases that involve Unicode properties, both
850 explicit and implicit when PCRE_UCP is set.
851
852 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
853 input, it could crash or give wrong results if characters with values
854 greater than 0xc0 were present in the subject string. (Detail: it assumed
855 UTF-8 input when processing these items.)
856
857 16. Added a lot of (int) casts to avoid compiler warnings in systems where
858 size_t is 64-bit (#991).
859
860 17. Added a check for running out of memory when PCRE is compiled with
861 --disable-stack-for-recursion (#990).
862
863 18. If the last data line in a file for pcretest does not have a newline on
864 the end, a newline was missing in the output.
865
866 19. The default pcre_chartables.c file recognizes only ASCII characters (values
867 less than 128) in its various bitmaps. However, there is a facility for
868 generating tables according to the current locale when PCRE is compiled. It
869 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
870 characters, are recognized by isspace() and therefore were getting set in
871 these tables, and indeed these tables seem to approximate to ISO 8859. This
872 caused a problem in UTF-8 mode when pcre_study() was used to create a list
873 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
874 which of course cannot start UTF-8 characters. I have changed the code so
875 that only real ASCII characters (less than 128) and the correct starting
876 bytes for UTF-8 encodings are set for characters greater than 127 when in
877 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
878 altogether.)
879
880 20. Added the /T option to pcretest so as to be able to run tests with non-
881 standard character tables, thus making it possible to include the tests
882 used for 19 above in the standard set of tests.
883
884 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
885 reference to a subpattern the other side of a comment that contains an
886 opening parenthesis caused either an internal compiling error, or a
887 reference to the wrong subpattern.
888
889
890 Version 8.02 19-Mar-2010
891 ------------------------
892
893 1. The Unicode data tables have been updated to Unicode 5.2.0.
894
895 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
896 configured.
897
898 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
899 original author of that file, following a query about its status.
900
901 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
902 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
903
904 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
905 quantifier applied to a forward-referencing subroutine call, could compile
906 incorrect code or give the error "internal error: previously-checked
907 referenced subpattern not found".
908
909 6. Both MS Visual Studio and Symbian OS have problems with initializing
910 variables to point to external functions. For these systems, therefore,
911 pcre_malloc etc. are now initialized to local functions that call the
912 relevant global functions.
913
914 7. There were two entries missing in the vectors called coptable and poptable
915 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
916 I've fixed the data, and added a kludgy way of testing at compile time that
917 the lengths are correct (equal to the number of opcodes).
918
919 8. Following on from 7, I added a similar kludge to check the length of the
920 eint vector in pcreposix.c.
921
922 9. Error texts for pcre_compile() are held as one long string to avoid too
923 much relocation at load time. To find a text, the string is searched,
924 counting zeros. There was no check for running off the end of the string,
925 which could happen if a new error number was added without updating the
926 string.
927
928 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
929
930 11. \K was not working if it appeared in an atomic group or in a group that
931 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
932 \K is "not well defined" if used in an assertion. PCRE now accepts it if
933 the assertion is positive, but not if it is negative.
934
935 12. Change 11 fortuitously reduced the size of the stack frame used in the
936 "match()" function of pcre_exec.c by one pointer. Forthcoming
937 implementation of support for (*MARK) will need an extra pointer on the
938 stack; I have reserved it now, so that the stack frame size does not
939 decrease.
940
941 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
942 item in branch that calls a recursion is a subroutine call - as in the
943 second branch in the above example - was incorrectly given the compile-
944 time error "recursive call could loop indefinitely" because pcre_compile()
945 was not correctly checking the subroutine for matching a non-empty string.
946
947 14. The checks for overrunning compiling workspace could trigger after an
948 overrun had occurred. This is a "should never occur" error, but it can be
949 triggered by pathological patterns such as hundreds of nested parentheses.
950 The checks now trigger 100 bytes before the end of the workspace.
951
952 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
953
954
955 Version 8.01 19-Jan-2010
956 ------------------------
957
958 1. If a pattern contained a conditional subpattern with only one branch (in
959 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
960 computed the wrong minimum data length (which is of course zero for such
961 subpatterns). This could cause incorrect "no match" results.
962
963 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
964 the pattern is reset in the first branch, pcre_compile() failed with
965 "internal error: code overflow at offset...". This happened only when
966 the reset was to the original external option setting. (An optimization
967 abstracts leading options settings into an external setting, which was the
968 cause of this.)
969
970 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
971 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
972 assertion pattern did not match (meaning that the assertion was true), it
973 was incorrectly treated as false if the SKIP had been reached during the
974 matching. This also applied to assertions used as conditions.
975
976 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
977 assertion subpattern, including such a pattern used as a condition,
978 unpredictable results occurred, instead of the error return
979 PCRE_ERROR_DFA_UITEM.
980
981 5. The C++ GlobalReplace function was not working like Perl for the special
982 situation when an empty string is matched. It now does the fancy magic
983 stuff that is necessary.
984
985 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
986 removed. (These were left over from very, very early versions of PCRE.)
987
988 7. Some cosmetic changes to the code to make life easier when compiling it
989 as part of something else:
990
991 (a) Change DEBUG to PCRE_DEBUG.
992
993 (b) In pcre_compile(), rename the member of the "branch_chain" structure
994 called "current" as "current_branch", to prevent a collision with the
995 Linux macro when compiled as a kernel module.
996
997 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
998 prevent a collision with the Linux macro when compiled as a kernel
999 module.
1000
1001 8. In pcre_compile() there are some checks for integer overflows that used to
1002 cast potentially large values to (double). This has been changed to that
1003 when building, a check for int64_t is made, and if it is found, it is used
1004 instead, thus avoiding the use of floating point arithmetic. (There is no
1005 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1006 double.
1007
1008 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1009 2005 (difference between two addresses compared to an unsigned value).
1010
1011 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1012 custom one, because of the following reported problem in Windows:
1013
1014 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1015 under Win32.
1016 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1017 therefore missing the function definition.
1018 - The compiler thus generates a "C" signature for the test function.
1019 - The linker fails to find the "C" function.
1020 - PCRE fails to configure if asked to do so against libbz2.
1021
1022 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1023 messages were output:
1024
1025 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1026 rerunning libtoolize, to keep the correct libtool macros in-tree.
1027 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1028
1029 I have done both of these things.
1030
1031 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1032 most of the time, it *can* run out if it is given a pattern that contains a
1033 runaway infinite recursion. I updated the discussion in the pcrestack man
1034 page.
1035
1036 13. Now that we have gone to the x.xx style of version numbers, the minor
1037 version may start with zero. Using 08 or 09 is a bad idea because users
1038 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1039 interpreted as invalid octal numbers. I've updated the previous comment in
1040 configure.ac, and also added a check that gives an error if 08 or 09 are
1041 used.
1042
1043 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1044 causing partial matching to fail when the end of the subject matched \W
1045 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1046
1047 15. There were some discrepancies between the declarations in pcre_internal.h
1048 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1049 their definitions. The declarations used "const uschar *" and the
1050 definitions used USPTR. Even though USPTR is normally defined as "const
1051 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1052 reported that: "This difference in casting confuses some C++ compilers, for
1053 example, SunCC recognizes above declarations as different functions and
1054 generates broken code for hbpcre." I have changed the declarations to use
1055 USPTR.
1056
1057 16. GNU libtool is named differently on some systems. The autogen.sh script now
1058 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1059 (FreeBSD).
1060
1061 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1062 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1063 comment: "Figure out how to create a longlong from a string: strtoll and
1064 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1065 instance, but it only takes 2 args instead of 3!"
1066
1067 18. A subtle bug concerned with back references has been fixed by a change of
1068 specification, with a corresponding code fix. A pattern such as
1069 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1070 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1071 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1072 same bug. Such groups have to be quantified to be useful, or contained
1073 inside another quantified group. (If there's no repetition, the reference
1074 can never match.) The problem arises because, having left the group and
1075 moved on to the rest of the pattern, a later failure that backtracks into
1076 the group uses the captured value from the final iteration of the group
1077 rather than the correct earlier one. I have fixed this in PCRE by forcing
1078 any group that contains a reference to itself to be an atomic group; that
1079 is, there cannot be any backtracking into it once it has completed. This is
1080 similar to recursive and subroutine calls.
1081
1082
1083 Version 8.00 19-Oct-09
1084 ----------------------
1085
1086 1. The table for translating pcre_compile() error codes into POSIX error codes
1087 was out-of-date, and there was no check on the pcre_compile() error code
1088 being within the table. This could lead to an OK return being given in
1089 error.
1090
1091 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1092 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1093 in a Windows environment.
1094
1095 3. The pcregrep --count option prints the count for each file even when it is
1096 zero, as does GNU grep. However, pcregrep was also printing all files when
1097 --files-with-matches was added. Now, when both options are given, it prints
1098 counts only for those files that have at least one match. (GNU grep just
1099 prints the file name in this circumstance, but including the count seems
1100 more useful - otherwise, why use --count?) Also ensured that the
1101 combination -clh just lists non-zero counts, with no names.
1102
1103 4. The long form of the pcregrep -F option was incorrectly implemented as
1104 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1105 but it seems right to fix it, and I didn't think it was worth preserving
1106 the old behaviour.
1107
1108 5. The command line items --regex=pattern and --regexp=pattern were not
1109 recognized by pcregrep, which required --regex pattern or --regexp pattern
1110 (with a space rather than an '='). The man page documented the '=' forms,
1111 which are compatible with GNU grep; these now work.
1112
1113 6. No libpcreposix.pc file was created for pkg-config; there was just
1114 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1115
1116 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1117 when UCP support is not needed, by modifying the Python script that
1118 generates it from Unicode data files. This should not matter if the module
1119 is correctly used as a library, but I received one complaint about 50K of
1120 unwanted data. My guess is that the person linked everything into his
1121 program rather than using a library. Anyway, it does no harm.
1122
1123 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1124 was a minimum greater than 1 for a wide character in a possessive
1125 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1126 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1127 character. Chaos in the form of incorrect output or a compiling loop could
1128 result.
1129
1130 9. The restrictions on what a pattern can contain when partial matching is
1131 requested for pcre_exec() have been removed. All patterns can now be
1132 partially matched by this function. In addition, if there are at least two
1133 slots in the offset vector, the offset of the earliest inspected character
1134 for the match and the offset of the end of the subject are set in them when
1135 PCRE_ERROR_PARTIAL is returned.
1136
1137 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1138 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1139 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1140 and may be more useful for multi-segment matching.
1141
1142 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1143 used to be given if ever the end of the subject was reached; now it is
1144 given only if matching could not proceed because another character was
1145 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1146 string "Z", which now yields "no match" instead of "partial match". In the
1147 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1148 final character ended with (*FAIL).
1149
1150 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1151 if the pattern had a "must contain" character that was already found in the
1152 earlier partial match, unless partial matching was again requested. For
1153 example, with the pattern /dog.(body)?/, the "must contain" character is
1154 "g". If the first part-match was for the string "dog", restarting with
1155 "sbody" failed. This bug has been fixed.
1156
1157 13. The string returned by pcre_dfa_exec() after a partial match has been
1158 changed so that it starts at the first inspected character rather than the
1159 first character of the match. This makes a difference only if the pattern
1160 starts with a lookbehind assertion or \b or \B (\K is not supported by
1161 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1162 matching functions compatible, and I think it's the right thing to do.
1163
1164 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1165 so that the demonstration program is easily available in environments where
1166 PCRE has not been installed from source.
1167
1168 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1169 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1170 library.
1171
1172 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1173 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1174 is not the first non-POSIX option to be added. Clearly some people find
1175 these options useful.
1176
1177 17. If a caller to the POSIX matching function regexec() passes a non-zero
1178 value for nmatch with a NULL value for pmatch, the value of
1179 nmatch is forced to zero.
1180
1181 18. RunGrepTest did not have a test for the availability of the -u option of
1182 the diff command, as RunTest does. It now checks in the same way as
1183 RunTest, and also checks for the -b option.
1184
1185 19. If an odd number of negated classes containing just a single character
1186 interposed, within parentheses, between a forward reference to a named
1187 subpattern and the definition of the subpattern, compilation crashed with
1188 an internal error, complaining that it could not find the referenced
1189 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1190 [The bug was that it was starting one character too far in when skipping
1191 over the character class, thus treating the ] as data rather than
1192 terminating the class. This meant it could skip too much.]
1193
1194 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1195 /g option in pcretest when the pattern contains \K, which makes it possible
1196 to have an empty string match not at the start, even when the pattern is
1197 anchored. Updated pcretest and pcredemo to use this option.
1198
1199 21. If the maximum number of capturing subpatterns in a recursion was greater
1200 than the maximum at the outer level, the higher number was returned, but
1201 with unset values at the outer level. The correct (outer level) value is
1202 now given.
1203
1204 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1205 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1206 make it do so. The string so far is captured, making this feature
1207 compatible with Perl.
1208
1209 23. The tests have been re-organized, adding tests 11 and 12, to make it
1210 possible to check the Perl 5.10 features against Perl 5.10.
1211
1212 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1213 pattern matches a fixed length string. PCRE did not allow this; now it
1214 does. Neither allows recursion.
1215
1216 25. I finally figured out how to implement a request to provide the minimum
1217 length of subject string that was needed in order to match a given pattern.
1218 (It was back references and recursion that I had previously got hung up
1219 on.) This code has now been added to pcre_study(); it finds a lower bound
1220 to the length of subject needed. It is not necessarily the greatest lower
1221 bound, but using it to avoid searching strings that are too short does give
1222 some useful speed-ups. The value is available to calling programs via
1223 pcre_fullinfo().
1224
1225 26. While implementing 25, I discovered to my embarrassment that pcretest had
1226 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1227 study optimizations had never been tested with that matching function.
1228 Oops. What is worse, even when it was passed study data, there was a bug in
1229 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1230 were also very few tests of studied patterns with pcre_dfa_exec().
1231
1232 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1233 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1234 on the other side of the coin, they are no longer allowed to have different
1235 names, because these cannot be distinguished in PCRE, and this has caused
1236 confusion. (This is a difference from Perl.)
1237
1238 28. When duplicate subpattern names are present (necessarily with different
1239 numbers, as required by 27 above), and a test is made by name in a
1240 conditional pattern, either for a subpattern having been matched, or for
1241 recursion in such a pattern, all the associated numbered subpatterns are
1242 tested, and the overall condition is true if the condition is true for any
1243 one of them. This is the way Perl works, and is also more like the way
1244 testing by number works.
1245
1246
1247 Version 7.9 11-Apr-09
1248 ---------------------
1249
1250 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1251 (pcretest), all targets were linked against these libraries. This included
1252 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1253 libraries. This caused unwanted dependencies to be created. This problem
1254 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1255 pcretest is linked with readline.
1256
1257 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1258 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1259 moved outside it again, because FALSE and TRUE are already defined in AIX,
1260 but BOOL is not.
1261
1262 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1263 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1264
1265 4. The pcregrep documentation said spaces were inserted as well as colons (or
1266 hyphens) following file names and line numbers when outputting matching
1267 lines. This is not true; no spaces are inserted. I have also clarified the
1268 wording for the --colour (or --color) option.
1269
1270 5. In pcregrep, when --colour was used with -o, the list of matching strings
1271 was not coloured; this is different to GNU grep, so I have changed it to be
1272 the same.
1273
1274 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1275 each matching line was coloured. Now it goes on to look for further matches
1276 of any of the test patterns, which is the same behaviour as GNU grep.
1277
1278 7. A pattern that could match an empty string could cause pcregrep to loop; it
1279 doesn't make sense to accept an empty string match in pcregrep, so I have
1280 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1281 seems to be how GNU grep behaves.
1282
1283 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1284 start or after a newline", because the conditional assertion was not being
1285 correctly handled. The rule now is that both the assertion and what follows
1286 in the first alternative must satisfy the test.
1287
1288 9. If auto-callout was enabled in a pattern with a conditional group whose
1289 condition was an assertion, PCRE could crash during matching, both with
1290 pcre_exec() and pcre_dfa_exec().
1291
1292 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1293 used for matching.
1294
1295 11. Unicode property support in character classes was not working for
1296 characters (bytes) greater than 127 when not in UTF-8 mode.
1297
1298 12. Added the -M command line option to pcretest.
1299
1300 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1301
1302 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1303
1304 16. Added comments and documentation about mis-use of no_arg in the C++
1305 wrapper.
1306
1307 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1308 from Martin Jerabek that uses macro names for all relevant character and
1309 string constants.
1310
1311 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1312 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1313 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1314 these, but not everybody uses configure.
1315
1316 19. A conditional group that had only one branch was not being correctly
1317 recognized as an item that could match an empty string. This meant that an
1318 enclosing group might also not be so recognized, causing infinite looping
1319 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1320 with the subject "ab", where knowledge that the repeated group can match
1321 nothing is needed in order to break the loop.
1322
1323 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1324 exec(), but without supplying a callout function, matching went wrong.
1325
1326 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1327 leak if the size of the offset vector was greater than 30. When the vector
1328 is smaller, the saved offsets during recursion go onto a local stack
1329 vector, but for larger vectors malloc() is used. It was failing to free
1330 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1331 error, in fact).
1332
1333 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1334 heapframe that is used only when UTF-8 support is enabled. This caused no
1335 problem, but was untidy.
1336
1337 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1338 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1339 included within another project.
1340
1341 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1342 slightly modified by me:
1343
1344 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1345 not building pcregrep.
1346
1347 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1348 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1349
1350 25. Forward references, both numeric and by name, in patterns that made use of
1351 duplicate group numbers, could behave incorrectly or give incorrect errors,
1352 because when scanning forward to find the reference group, PCRE was not
1353 taking into account the duplicate group numbers. A pattern such as
1354 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1355
1356 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1357 the feature of a custom pointer more persuasive (as requested by a user).
1358
1359 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1360 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1361 pre-defined.
1362
1363 28. Added support for (*UTF8) at the start of a pattern.
1364
1365 29. Arrange for flags added by the "release type" setting in CMake to be shown
1366 in the configuration summary.
1367
1368
1369 Version 7.8 05-Sep-08
1370 ---------------------
1371
1372 1. Replaced UCP searching code with optimized version as implemented for Ad
1373 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1374 stage table and inline lookup instead of a function, giving speed ups of 2
1375 to 5 times on some simple patterns that I tested. Permission was given to
1376 distribute the MultiStage2.py script that generates the tables (it's not in
1377 the tarball, but is in the Subversion repository).
1378
1379 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1380 scripts.
1381
1382 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1383 a group with a zero qualifier. The result of the study could be incorrect,
1384 or the function might crash, depending on the pattern.
1385
1386 4. Caseless matching was not working for non-ASCII characters in back
1387 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1388 It now works when Unicode Property Support is available.
1389
1390 5. In pcretest, an escape such as \x{de} in the data was always generating
1391 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1392 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1393 truncation.
1394
1395 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1396
1397 7. Added two (int) casts to pcregrep when printing the difference of two
1398 pointers, in case they are 64-bit values.
1399
1400 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1401 test 2 if it fails.
1402
1403 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1404 and a #define of that name to empty if it is not externally set. This is to
1405 allow users of MSVC to set it if necessary.
1406
1407 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1408 the convenience functions in the pcre_get.c source file.
1409
1410 11. An option change at the start of a pattern that had top-level alternatives
1411 could cause overwriting and/or a crash. This command provoked a crash in
1412 some environments:
1413
1414 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1415
1416 This potential security problem was recorded as CVE-2008-2371.
1417
1418 12. For a pattern where the match had to start at the beginning or immediately
1419 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1420 pcre_dfa_exec() could read past the end of the passed subject if there was
1421 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1422 pcretest so that it places the subject at the end of its malloc-ed buffer.
1423
1424 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1425 exec() might read past the end of the data buffer in UTF-8 mode.
1426
1427 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1428 the data contained the byte 0x85 as part of a UTF-8 character within its
1429 first line. This applied both to normal and DFA matching.
1430
1431 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1432 /^[^d]*?$/8 failed to match "abc".
1433
1434 16. Added a missing copyright notice to pcrecpp_internal.h.
1435
1436 17. Make it more clear in the documentation that values returned from
1437 pcre_exec() in ovector are byte offsets, not character counts.
1438
1439 18. Tidied a few places to stop certain compilers from issuing warnings.
1440
1441 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1442 supplied by Stefan Weber. I made a further small update for 7.8 because
1443 there is a change of source arrangements: the pcre_searchfuncs.c module is
1444 replaced by pcre_ucd.c.
1445
1446
1447 Version 7.7 07-May-08
1448 ---------------------
1449
1450 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1451 a string to a long long, pretend we don't even have a long long." This is
1452 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1453
1454 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1455 pre-7.6 versions, which defined a global no_arg variable instead of putting
1456 it in the RE class. (See also #8 below.)
1457
1458 3. Remove a line of dead code, identified by coverity and reported by Nuno
1459 Lopes.
1460
1461 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1462
1463 (1) The include/exclude patterns were being applied to the whole pathnames
1464 of files, instead of just to the final components.
1465
1466 (2) If there was more than one level of directory, the subdirectories were
1467 skipped unless they satisfied the include/exclude conditions. This is
1468 inconsistent with GNU grep (and could even be seen as contrary to the
1469 pcregrep specification - which I improved to make it absolutely clear).
1470 The action now is always to scan all levels of directory, and just
1471 apply the include/exclude patterns to regular files.
1472
1473 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1474 --exclude_dir in the tests to avoid scanning .svn directories.
1475
1476 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1477 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1478 doesn't support NULs in patterns.
1479
1480 7. Added some missing "const"s to declarations of static tables in
1481 pcre_compile.c and pcre_dfa_exec.c.
1482
1483 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1484 caused by fix #2 above. (Subsequently also a second patch to fix the
1485 first patch. And a third patch - this was a messy problem.)
1486
1487 9. Applied Craig's patch to remove the use of push_back().
1488
1489 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1490 matching function regexec().
1491
1492 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1493 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1494 references. PCRE supports relative numbers with this syntax (I don't think
1495 Oniguruma does).
1496
1497 12. Previously, a group with a zero repeat such as (...){0} was completely
1498 omitted from the compiled regex. However, this means that if the group
1499 was called as a subroutine from elsewhere in the pattern, things went wrong
1500 (an internal error was given). Such groups are now left in the compiled
1501 pattern, with a new opcode that causes them to be skipped at execution
1502 time.
1503
1504 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1505 to the way PCRE behaves:
1506
1507 (a) A lone ] character is dis-allowed (Perl treats it as data).
1508
1509 (b) A back reference to an unmatched subpattern matches an empty string
1510 (Perl fails the current match path).
1511
1512 (c) A data ] in a character class must be notated as \] because if the
1513 first data character in a class is ], it defines an empty class. (In
1514 Perl it is not possible to have an empty class.) The empty class []
1515 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1516 The negative empty class [^] matches any one character, independently
1517 of the DOTALL setting.
1518
1519 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1520 non-existent subpattern following a character class starting with ']' and
1521 containing () gave an internal compiling error instead of "reference to
1522 non-existent subpattern". Fortunately, when the pattern did exist, the
1523 compiled code was correct. (When scanning forwards to check for the
1524 existencd of the subpattern, it was treating the data ']' as terminating
1525 the class, so got the count wrong. When actually compiling, the reference
1526 was subsequently set up correctly.)
1527
1528 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1529 it was being rejected as not supported by pcre_dfa_exec(), even though
1530 other assertions are supported. I have made pcre_dfa_exec() support
1531 (*FAIL).
1532
1533 16. The implementation of 13c above involved the invention of a new opcode,
1534 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1535 cannot be changed at match time, I realized I could make a small
1536 improvement to matching performance by compiling OP_ALLANY instead of
1537 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1538 on the OP_ANY path.
1539
1540 17. Compiling pcretest on Windows with readline support failed without the
1541 following two fixes: (1) Make the unistd.h include conditional on
1542 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1543
1544 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1545 ncurses library to be included for pcretest when ReadLine support is
1546 requested, but also to allow for it to be overridden. This patch came from
1547 Daniel Bergström.
1548
1549 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1550 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1551 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1552 spotting this.
1553
1554
1555 Version 7.6 28-Jan-08
1556 ---------------------
1557
1558 1. A character class containing a very large number of characters with
1559 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1560 overflow.
1561
1562 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1563 HAVE_LONG_LONG is not defined.
1564
1565 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1566 bring them up to date and include new features. This patch includes:
1567
1568 - Fixed PH's badly added libz and libbz2 support.
1569 - Fixed a problem with static linking.
1570 - Added pcredemo. [But later removed - see 7 below.]
1571 - Fixed dftables problem and added an option.
1572 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1573 HAVE_LONG_LONG.
1574 - Added readline support for pcretest.
1575 - Added an listing of the option settings after cmake has run.
1576
1577 4. A user submitted a patch to Makefile that makes it easy to create
1578 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1579 Makefile.am that cause it to include this special target, without
1580 affecting anything else. Note that the same mingw target plus all
1581 the other distribution libraries and programs are now supported
1582 when configuring with CMake (see 6 below) instead of with
1583 Configure/Make.
1584
1585 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1586 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1587 exported in the Windows port". It has not yet been confirmed that the patch
1588 solves the problem, but it does no harm.
1589
1590 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1591 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1592 with CMake, and also correct the comment about stack recursion.
1593
1594 7. Remove the automatic building of pcredemo from the ./configure system and
1595 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1596 of a program that users should build themselves after PCRE is installed, so
1597 building it automatically is not really right. What is more, it gave
1598 trouble in some build environments.
1599
1600 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1601
1602
1603 Version 7.5 10-Jan-08
1604 ---------------------
1605
1606 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1607 values in parens when parsing an RE using the C++ wrapper."
1608
1609 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1610 Characters greater than 255 were excluded from the class instead of being
1611 included.
1612
1613 3. The same bug as (2) above applied to negated POSIX classes such as
1614 [:^space:].
1615
1616 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1617 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1618 I have changed it.
1619
1620 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1621 first named subpattern) and a construct such as (?&a) would reference the
1622 first named subpattern whose name started with "a" (in other words, the
1623 length check was missing). Both these problems are fixed. "Subpattern name
1624 expected" is now given for (?&) (a zero-length name), and this patch also
1625 makes it give the same error for \k'' (previously it complained that that
1626 was a reference to a non-existent subpattern).
1627
1628 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1629 this is right because (?- can be followed by option settings as well as by
1630 digits. I have, however, made the messages clearer.
1631
1632 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1633 than the number used in the conditional) now cause a compile-time error.
1634 This is actually not compatible with Perl, which accepts such patterns, but
1635 treats the conditional as always being FALSE (as PCRE used to), but it
1636 seems to me that giving a diagnostic is better.
1637
1638 8. Change "alphameric" to the more common word "alphanumeric" in comments
1639 and messages.
1640
1641 9. Fix two occurrences of "backslash" in comments that should have been
1642 "backspace".
1643
1644 10. Remove two redundant lines of code that can never be obeyed (their function
1645 was moved elsewhere).
1646
1647 11. The program that makes PCRE's Unicode character property table had a bug
1648 which caused it to generate incorrect table entries for sequences of
1649 characters that have the same character type, but are in different scripts.
1650 It amalgamated them into a single range, with the script of the first of
1651 them. In other words, some characters were in the wrong script. There were
1652 thirteen such cases, affecting characters in the following ranges:
1653
1654 U+002b0 - U+002c1
1655 U+0060c - U+0060d
1656 U+0061e - U+00612
1657 U+0064b - U+0065e
1658 U+0074d - U+0076d
1659 U+01800 - U+01805
1660 U+01d00 - U+01d77
1661 U+01d9b - U+01dbf
1662 U+0200b - U+0200f
1663 U+030fc - U+030fe
1664 U+03260 - U+0327f
1665 U+0fb46 - U+0fbb1
1666 U+10450 - U+1049d
1667
1668 12. The -o option (show only the matching part of a line) for pcregrep was not
1669 compatible with GNU grep in that, if there was more than one match in a
1670 line, it showed only the first of them. It now behaves in the same way as
1671 GNU grep.
1672
1673 13. If the -o and -v options were combined for pcregrep, it printed a blank
1674 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1675 does the same. The return code can be used to tell if there were any
1676 non-matching lines.
1677
1678 14. Added --file-offsets and --line-offsets to pcregrep.
1679
1680 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1681 infinitely looping recursion. The bug was that positive lookaheads were not
1682 being skipped when checking for a possible empty match (negative lookaheads
1683 and both kinds of lookbehind were skipped).
1684
1685 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1686 inclusion of <windows.h> to before rather than after the definition of
1687 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1688
1689 17. Specifying a possessive quantifier with a specific limit for a Unicode
1690 character property caused pcre_compile() to compile bad code, which led at
1691 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1692 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1693 caused the error; without that there was no problem.
1694
1695 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1696
1697 19. Added --enable-pcretest-libreadline.
1698
1699 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1700 RE::GlobalReplace(). As a result, the number of replacements returned was
1701 double what it should be. I removed one of the increments, but Craig sent a
1702 later patch that removed the other one (the right fix) and added unit tests
1703 that check the return values (which was not done before).
1704
1705 21. Several CMake things:
1706
1707 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1708 the names libpcre and libpcreposix, not just pcre and pcreposix.
1709
1710 (2) The above change means that pcretest and pcregrep are now correctly
1711 linked with the newly-built libraries, not previously installed ones.
1712
1713 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1714
1715 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1716 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1717 UTF-8 newline character). The key issue is that the pattern starts .*;
1718 this means that the match must be either at the beginning, or after a
1719 newline. The bug was in the code for advancing after a failed match and
1720 checking that the new position followed a newline. It was not taking
1721 account of UTF-8 characters correctly.
1722
1723 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1724 character classes. PCRE was not treating the sequence [:...:] as a
1725 character class unless the ... were all letters. Perl, however, seems to
1726 allow any characters between [: and :], though of course it rejects as
1727 unknown any "names" that contain non-letters, because all the known class
1728 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1729 for example, whereas PCRE did not - it did not recognize a POSIX character
1730 class. This seemed a bit dangerous, so the code has been changed to be
1731 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1732 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1733 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1734 Perl does, and where it didn't before.
1735
1736 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1737 Windows environments %n is disabled by default.
1738
1739
1740 Version 7.4 21-Sep-07
1741 ---------------------
1742
1743 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1744 means that a class such as [\s] counted as "explicit reference to CR or
1745 LF". That isn't really right - the whole point of the change was to try to
1746 help when there was an actual mention of one of the two characters. So now
1747 the change happens only if \r or \n (or a literal CR or LF) character is
1748 encountered.
1749
1750 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1751 of both had grown to the point where there were only 3 bits left.
1752 Fortunately, there was spare space in the data structure, and so I have
1753 moved the internal flags into a new 16-bit field to free up more option
1754 bits.
1755
1756 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1757 but did not set the internal JCHANGED flag - either of these is enough to
1758 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1759 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1760 start sets both bits.
1761
1762 4. Added options (at build time, compile time, exec time) to change \R from
1763 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1764
1765 5. doc/pcresyntax.html was missing from the distribution.
1766
1767 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1768 compatibility, even though it is no longer used.
1769
1770 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1771 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1772 windows.h file is present (where different names are used). [This was
1773 reversed later after testing - see 16 below.]
1774
1775 8. Changed all #include <config.h> to #include "config.h". There were also
1776 some further <pcre.h> cases that I changed to "pcre.h".
1777
1778 9. When pcregrep was used with the --colour option, it missed the line ending
1779 sequence off the lines that it output.
1780
1781 10. It was pointed out to me that arrays of string pointers cause lots of
1782 relocations when a shared library is dynamically loaded. A technique of
1783 using a single long string with a table of offsets can drastically reduce
1784 these. I have refactored PCRE in four places to do this. The result is
1785 dramatic:
1786
1787 Originally: 290
1788 After changing UCP table: 187
1789 After changing error message table: 43
1790 After changing table of "verbs" 36
1791 After changing table of Posix names 22
1792
1793 Thanks to the folks working on Gregex for glib for this insight.
1794
1795 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1796 unicode-properties was also set.
1797
1798 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1799
1800 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1801 checked only for CRLF.
1802
1803 14. Added casts to pcretest.c to avoid compiler warnings.
1804
1805 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1806
1807 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1808 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1809 entirely. This removes changes made in 7 above.
1810
1811 17. The CMake files have been updated, and there is now more information about
1812 building with CMake in the NON-UNIX-USE document.
1813
1814
1815 Version 7.3 28-Aug-07
1816 ---------------------
1817
1818 1. In the rejigging of the build system that eventually resulted in 7.1, the
1819 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1820 brackets there is not right, since it causes compilers to look for an
1821 installed pcre.h, not the version that is in the source that is being
1822 compiled (which of course may be different). I have changed it back to:
1823
1824 #include "pcre.h"
1825
1826 I have a vague recollection that the change was concerned with compiling in
1827 different directories, but in the new build system, that is taken care of
1828 by the VPATH setting the Makefile.
1829
1830 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1831 when the subject happened to end in the byte 0x85 (e.g. if the last
1832 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1833 characters but of course it shouldn't be taken as a newline when it is part
1834 of another character. The bug was that, for an unlimited repeat of . in
1835 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1836 characters when looking for a newline.
1837
1838 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1839
1840 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1841 in debug output.
1842
1843 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1844 long printing in the pcrecpp unittest when running under MinGW.
1845
1846 6. ESC_K was left out of the EBCDIC table.
1847
1848 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1849 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1850 limit also applies to "virtual nesting" when a pattern is recursive, and in
1851 this case 1000 isn't so big. I have been able to remove this limit at the
1852 expense of backing off one optimization in certain circumstances. Normally,
1853 when pcre_exec() would call its internal match() function recursively and
1854 immediately return the result unconditionally, it uses a "tail recursion"
1855 feature to save stack. However, when a subpattern that can match an empty
1856 string has an unlimited repetition quantifier, it no longer makes this
1857 optimization. That gives it a stack frame in which to save the data for
1858 checking that an empty string has been matched. Previously this was taken
1859 from the 1000-entry workspace that had been reserved. So now there is no
1860 explicit limit, but more stack is used.
1861
1862 8. Applied Daniel's patches to solve problems with the import/export magic
1863 syntax that is required for Windows, and which was going wrong for the
1864 pcreposix and pcrecpp parts of the library. These were overlooked when this
1865 problem was solved for the main library.
1866
1867 9. There were some crude static tests to avoid integer overflow when computing
1868 the size of patterns that contain repeated groups with explicit upper
1869 limits. As the maximum quantifier is 65535, the maximum group length was
1870 set at 30,000 so that the product of these two numbers did not overflow a
1871 32-bit integer. However, it turns out that people want to use groups that
1872 are longer than 30,000 bytes (though not repeat them that many times).
1873 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1874 made it possible to implement the integer overflow checks in a much more
1875 dynamic way, which I have now done. The artificial limitation on group
1876 length has been removed - we now have only the limit on the total length of
1877 the compiled pattern, which depends on the LINK_SIZE setting.
1878
1879 10. Fixed a bug in the documentation for get/copy named substring when
1880 duplicate names are permitted. If none of the named substrings are set, the
1881 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1882 empty string.
1883
1884 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1885 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1886 because the ] is interpreted as the first data character and the
1887 terminating ] is not found. PCRE has been made compatible with Perl in this
1888 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1889 cause memory overwriting.
1890
1891 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1892 string has been matched (to stop an infinite loop). It was not recognizing
1893 a conditional subpattern that could match an empty string if that
1894 subpattern was within another subpattern. For example, it looped when
1895 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1896 condition was not nested. This bug has been fixed.
1897
1898 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1899 past the start of the subject in the presence of bytes with the top bit
1900 set, for example "\x8aBCD".
1901
1902 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1903 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1904
1905 14. Optimized (?!) to (*FAIL).
1906
1907 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1908 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1909 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1910 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1911 does: it's just the validity check that is more restrictive.
1912
1913 16. Inserted checks for integer overflows during escape sequence (backslash)
1914 processing, and also fixed erroneous offset values for syntax errors during
1915 backslash processing.
1916
1917 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1918 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1919
1920 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1921 caused an overrun.
1922
1923 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1924 something other than just ASCII characters) inside a group that had an
1925 unlimited repeat caused a loop at compile time (while checking to see
1926 whether the group could match an empty string).
1927
1928 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1929 [\P{Any}] did so. (Error in the code for printing property names.)
1930
1931 21. An orphan \E inside a character class could cause a crash.
1932
1933 22. A repeated capturing bracket such as (A)? could cause a wild memory
1934 reference during compilation.
1935
1936 23. There are several functions in pcre_compile() that scan along a compiled
1937 expression for various reasons (e.g. to see if it's fixed length for look
1938 behind). There were bugs in these functions when a repeated \p or \P was
1939 present in the pattern. These operators have additional parameters compared
1940 with \d, etc, and these were not being taken into account when moving along
1941 the compiled data. Specifically:
1942
1943 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1944 length.
1945
1946 (b) An item such as \pL+ within a repeated group could cause crashes or
1947 loops.
1948
1949 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1950 "reference to non-existent subpattern" error.
1951
1952 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1953
1954 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1955 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1956
1957 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1958
1959 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1960 character were causing crashes (broken optimization).
1961
1962 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1963 \p or \P) caused a compile-time loop.
1964
1965 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1966 break. For example, the unstudied pattern [\r\n]A does not match the string
1967 "\r\nA" because change 7.0/46 below moves the current point on by two
1968 characters after failing to match at the start. However, the pattern \nA
1969 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1970 the same is true. There doesn't seem any very clean way out of this, but
1971 what I have chosen to do makes the common cases work: PCRE now takes note
1972 of whether there can be an explicit match for \r or \n anywhere in the
1973 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1974 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1975 pattern has explicit CR or LF references.
1976
1977 29. Added (*CR) etc for changing newline setting at start of pattern.
1978
1979
1980 Version 7.2 19-Jun-07
1981 ---------------------
1982
1983 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1984 which is apparently normally available under Windows.
1985
1986 2. Re-jig the pcregrep tests with different newline settings in an attempt
1987 to make them independent of the local environment's newline setting.
1988
1989 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1990
1991 4. Some of the "internals" tests were previously cut out when the link size
1992 was not 2, because the output contained actual offsets. The recent new
1993 "Z" feature of pcretest means that these can be cut out, making the tests
1994 usable with all link sizes.
1995
1996 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1997 stack recursion. This gives a massive performance boost under BSD, but just
1998 a small improvement under Linux. However, it saves one field in the frame
1999 in all cases.
2000
2001 6. Added more features from the forthcoming Perl 5.10:
2002
2003 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2004 recursion call. It refers to the nth most recently opened parentheses.
2005
2006 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2007 to be opened parentheses.
2008
2009 (c) Conditions that refer to capturing parentheses can be specified
2010 relatively, for example, (?(-2)... or (?(+3)...
2011
2012 (d) \K resets the start of the current match so that everything before
2013 is not part of it.
2014
2015 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2016
2017 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2018 reference syntax.
2019
2020 (g) (?| introduces a group in which the numbering of parentheses in each
2021 alternative starts with the same number.
2022
2023 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2024
2025 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2026 PCRE_INFO_JCHANGED.
2027
2028 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2029 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2030 for detecting groups that can match an empty string.
2031
2032 9. A pattern with a very large number of alternatives (more than several
2033 hundred) was running out of internal workspace during the pre-compile
2034 phase, where pcre_compile() figures out how much memory will be needed. A
2035 bit of new cunning has reduced the workspace needed for groups with
2036 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2037 workspace instead of running out of the 4096 that are available.
2038
2039 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2040
2041 11. Applied patch from Google to remove an optimization that didn't quite work.
2042 The report of the bug said:
2043
2044 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2045 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2046 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2047
2048 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2049 it matched the wrong number of bytes.
2050
2051
2052 Version 7.1 24-Apr-07
2053 ---------------------
2054
2055 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2056 that is more "standard", making use of automake and other Autotools. There
2057 is some re-arrangement of the files and adjustment of comments consequent
2058 on this.
2059
2060 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2061 for recursive directory scanning broke on some systems because the files
2062 are not scanned in any specific order and on different systems the order
2063 was different. A call to "sort" has been inserted into RunGrepTest for the
2064 approprate test as a short-term fix. In the longer term there may be an
2065 alternative.
2066
2067 3. I had an email from Eric Raymond about problems translating some of PCRE's
2068 man pages to HTML (despite the fact that I distribute HTML pages, some
2069 people do their own conversions for various reasons). The problems
2070 concerned the use of low-level troff macros .br and .in. I have therefore
2071 removed all such uses from the man pages (some were redundant, some could
2072 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2073 HTML has been updated to handle .nf/.fi and to complain if it encounters
2074 .br or .in.
2075
2076 4. Updated comments in configure.ac that get placed in config.h.in and also
2077 arranged for config.h to be included in the distribution, with the name
2078 config.h.generic, for the benefit of those who have to compile without
2079 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2080
2081 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2082 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2083 makevp.bat for the current PCRE, using the additional files
2084 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2085
2086 6. A Windows user reported a minor discrepancy with test 2, which turned out
2087 to be caused by a trailing space on an input line that had got lost in his
2088 copy. The trailing space was an accident, so I've just removed it.
2089
2090 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2091 that is needed.
2092
2093 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2094 as "const" (a) because they are and (b) because it helps the PHP
2095 maintainers who have recently made a script to detect big data structures
2096 in the php code that should be moved to the .rodata section. I remembered
2097 to update Builducptable as well, so it won't revert if ucptable.h is ever
2098 re-created.
2099
2100 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2101 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2102 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2103 support is not required. This saves 1.5-2K of code, which is important in
2104 some applications.
2105
2106 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2107 so as not to refer to the tables, even though these functions will never be
2108 called when UTF-8 support is disabled. Otherwise there are problems with a
2109 shared library.
2110
2111 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2112
2113 (a) It was defining its arguments as char * instead of void *.
2114
2115 (b) It was assuming that all moves were upwards in memory; this was true
2116 a long time ago when I wrote it, but is no longer the case.
2117
2118 The emulated memove() is provided for those environments that have neither
2119 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2120 is clearly not the case, as these two bugs were recently reported.
2121
2122 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2123 and Detrail to create the HTML documentation, the .txt form of the man
2124 pages, and it removes trailing spaces from listed files. It also creates
2125 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2126 case, it wraps all the #defines with #ifndefs. This script should be run
2127 before "make dist".
2128
2129 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2130 with Unicode property support.
2131
2132 (a) For a maximizing quantifier, if the two different cases of the
2133 character were of different lengths in their UTF-8 codings (there are
2134 some cases like this - I found 11), and the matching function had to
2135 back up over a mixture of the two cases, it incorrectly assumed they
2136 were both the same length.
2137
2138 (b) When PCRE was configured to use the heap rather than the stack for
2139 recursion during matching, it was not correctly preserving the data for
2140 the other case of a UTF-8 character when checking ahead for a match
2141 while processing a minimizing repeat. If the check also involved
2142 matching a wide character, but failed, corruption could cause an
2143 erroneous result when trying to check for a repeat of the original
2144 character.
2145
2146 13. Some tidying changes to the testing mechanism:
2147
2148 (a) The RunTest script now detects the internal link size and whether there
2149 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2150 values substituted by "configure". (The RunGrepTest script already did
2151 this for UTF-8.) The configure.ac script no longer substitutes the
2152 relevant variables.
2153
2154 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2155 with length and offset values. This means that the output is different
2156 for different internal link sizes. Test 2 is skipped for link sizes
2157 other than 2 because of this, bypassing the problem. Unfortunately,
2158 there was also a test in test 3 (the locale tests) that used /B and
2159 failed for link sizes other than 2. Rather than cut the whole test out,
2160 I have added a new /Z option to pcretest that replaces the length and
2161 offset values with spaces. This is now used to make test 3 independent
2162 of link size. (Test 2 will be tidied up later.)
2163
2164 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2165 segmentation fault instead of returning the appropriate error message.
2166
2167 15. In multiline mode when the newline sequence was set to "any", the pattern
2168 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2169 This doesn't seem right; it now treats the CRLF combination as the line
2170 ending, and so does not match in that case. It's only a pattern such as ^$
2171 that would hit this one: something like ^ABC$ would have failed after \r
2172 and then tried again after \r\n.
2173
2174 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2175 in an attempt to make files that differ only in their line terminators
2176 compare equal. This works on Linux.
2177
2178 17. Under certain error circumstances pcregrep might try to free random memory
2179 as it exited. This is now fixed, thanks to valgrind.
2180
2181 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2182 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2183 was because its rules for how to advance for /g after matching an empty
2184 string at the end of a line did not allow for this case. They now check for
2185 it specially.
2186
2187 20. pcretest is supposed to handle patterns and data of any length, by
2188 extending its buffers when necessary. It was getting this wrong when the
2189 buffer for a data line had to be extended.
2190
2191 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2192 CRLF as a newline sequence.
2193
2194 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2195 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2196 I have nevertheless tidied it up.
2197
2198 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2199
2200 24. Added a man page for pcre-config.
2201
2202
2203 Version 7.0 19-Dec-06
2204 ---------------------
2205
2206 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2207 moving to gcc 4.1.1.
2208
2209 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2210 sys/time.h, which is documented as needed for this function. It doesn't
2211 seem to matter on Linux, but it showed up on some releases of OS X.
2212
2213 3. It seems that there are systems where bytes whose values are greater than
2214 127 match isprint() in the "C" locale. The "C" locale should be the
2215 default when a C program starts up. In most systems, only ASCII printing
2216 characters match isprint(). This difference caused the output from pcretest
2217 to vary, making some of the tests fail. I have changed pcretest so that:
2218
2219 (a) When it is outputting text in the compiled version of a pattern, bytes
2220 other than 32-126 are always shown as hex escapes.
2221
2222 (b) When it is outputting text that is a matched part of a subject string,
2223 it does the same, unless a different locale has been set for the match
2224 (using the /L modifier). In this case, it uses isprint() to decide.
2225
2226 4. Fixed a major bug that caused incorrect computation of the amount of memory
2227 required for a compiled pattern when options that changed within the
2228 pattern affected the logic of the preliminary scan that determines the
2229 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2230 that the computed length was too small. The symptoms of this bug were
2231 either the PCRE error "internal error: code overflow" from pcre_compile(),
2232 or a glibc crash with a message such as "pcretest: free(): invalid next
2233 size (fast)". Examples of patterns that provoked this bug (shown in
2234 pcretest format) are:
2235
2236 /(?-x: )/x
2237 /(?x)(?-x: \s*#\s*)/
2238 /((?i)[\x{c0}])/8
2239 /(?i:[\x{c0}])/8
2240
2241 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2242 is now done differently.
2243
2244 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2245 wrapper classes; (b) implement a new function in the C++ scanner that is
2246 more efficient than the old way of doing things because it avoids levels of
2247 recursion in the regex matching; (c) add a paragraph to the documentation
2248 for the FullMatch() function.
2249
2250 6. The escape sequence \n was being treated as whatever was defined as
2251 "newline". Not only was this contrary to the documentation, which states
2252 that \n is character 10 (hex 0A), but it also went horribly wrong when
2253 "newline" was defined as CRLF. This has been fixed.
2254
2255 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2256 was being set to -1 for the "end of line" case (supposedly a value that no
2257 character can have). Though this value is never used (the check for end of
2258 line is "zero bytes in current character"), it caused compiler complaints.
2259 I've changed it to 0xffffffff.
2260
2261 8. In pcre_version.c, the version string was being built by a sequence of
2262 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2263 string (as it is for production releases) called a macro with an empty
2264 argument. The C standard says the result of this is undefined. The gcc
2265 compiler treats it as an empty string (which was what was wanted) but it is
2266 reported that Visual C gives an error. The source has been hacked around to
2267 avoid this problem.
2268
2269 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2270 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2271 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2272 of them did).
2273
2274 10. Originally, pcretest opened its input and output without "b"; then I was
2275 told that "b" was needed in some environments, so it was added for release
2276 5.0 to both the input and output. (It makes no difference on Unix-like
2277 systems.) Later I was told that it is wrong for the input on Windows. I've
2278 now abstracted the modes into two macros, to make it easier to fiddle with
2279 them, and removed "b" from the input mode under Windows.
2280
2281 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2282
2283 12. Added -help and --help to pcretest as an official way of being reminded
2284 of the options.
2285
2286 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2287 and pcrecpp.cc because they annoy compilers at high warning levels.
2288
2289 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2290
2291 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2292 scripts are not C programs :-) and which was not noticed because it works
2293 on Linux.
2294
2295 16. pcretest is supposed to handle any length of pattern and data line (as one
2296 line or as a continued sequence of lines) by extending its input buffer if
2297 necessary. This feature was broken for very long pattern lines, leading to
2298 a string of junk being passed to pcre_compile() if the pattern was longer
2299 than about 50K.
2300
2301 17. I have done a major re-factoring of the way pcre_compile() computes the
2302 amount of memory needed for a compiled pattern. Previously, there was code
2303 that made a preliminary scan of the pattern in order to do this. That was
2304 OK when PCRE was new, but as the facilities have expanded, it has become
2305 harder and harder to keep it in step with the real compile phase, and there
2306 have been a number of bugs (see for example, 4 above). I have now found a
2307 cunning way of running the real compile function in a "fake" mode that
2308 enables it to compute how much memory it would need, while actually only
2309 ever using a few hundred bytes of working memory and without too many
2310 tests of the mode. This should make future maintenance and development
2311 easier. A side effect of this work is that the limit of 200 on the nesting
2312 depth of parentheses has been removed (though this was never a serious
2313 limitation, I suspect). However, there is a downside: pcre_compile() now
2314 runs more slowly than before (30% or more, depending on the pattern). I
2315 hope this isn't a big issue. There is no effect on runtime performance.
2316
2317 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2318 newline (only possible for the last line of a file) and it was a
2319 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2320
2321 19. Added additional timing features to pcretest. (1) The -tm option now times
2322 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2323 separate command line item, by a number that specifies the number of
2324 repeats to use when timing. The default is 50000; this gives better
2325 precision, but takes uncomfortably long for very large patterns.
2326
2327 20. Extended pcre_study() to be more clever in cases where a branch of a
2328 subpattern has no definite first character. For example, (a*|b*)[cd] would
2329 previously give no result from pcre_study(). Now it recognizes that the
2330 first character must be a, b, c, or d.
2331
2332 21. There was an incorrect error "recursive call could loop indefinitely" if
2333 a subpattern (or the entire pattern) that was being tested for matching an
2334 empty string contained only one non-empty item after a nested subpattern.
2335 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2336 incorrectly, because the \d was being skipped in the check.
2337
2338 22. The pcretest program now has a new pattern option /B and a command line
2339 option -b, which is equivalent to adding /B to every pattern. This causes
2340 it to show the compiled bytecode, without the additional information that
2341 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2342 is the same as /B/I).
2343
2344 23. A new optimization is now able automatically to treat some sequences such
2345 as a*b as a*+b. More specifically, if something simple (such as a character
2346 or a simple class like \d) has an unlimited quantifier, and is followed by
2347 something that cannot possibly match the quantified thing, the quantifier
2348 is automatically "possessified".
2349
2350 24. A recursive reference to a subpattern whose number was greater than 39
2351 went wrong under certain circumstances in UTF-8 mode. This bug could also
2352 have affected the operation of pcre_study().
2353
2354 25. Realized that a little bit of performance could be had by replacing
2355 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2356
2357 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2358
2359 27. Possessive quantifiers such as a++ were previously implemented by turning
2360 them into atomic groups such as ($>a+). Now they have their own opcodes,
2361 which improves performance. This includes the automatically created ones
2362 from 23 above.
2363
2364 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2365 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2366 the first matched character to be a colon. This applied both to named and
2367 numbered groups.
2368
2369 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2370
2371 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2372 building PCRE on Windows easier, so I have included it in the distribution.
2373
2374 31. There is now a check in pcretest against a ridiculously large number being
2375 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2376 loop, the loop is abandoned.
2377
2378 32. Forward references to subpatterns in conditions such as (?(2)...) where
2379 subpattern 2 is defined later cause pcre_compile() to search forwards in
2380 the pattern for the relevant set of parentheses. This search went wrong
2381 when there were unescaped parentheses in a character class, parentheses
2382 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2383
2384 33. "Subroutine" calls and backreferences were previously restricted to
2385 referencing subpatterns earlier in the regex. This restriction has now
2386 been removed.
2387
2388 34. Added a number of extra features that are going to be in Perl 5.10. On the
2389 whole, these are just syntactic alternatives for features that PCRE had
2390 previously implemented using the Python syntax or my own invention. The
2391 other formats are all retained for compatibility.
2392
2393 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2394 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2395 also .NET compatible.
2396
2397 (b) A recursion or subroutine call to a named group can now be defined as
2398 (?&name) as well as (?P>name).
2399
2400 (c) A backreference to a named group can now be defined as \k<name> or
2401 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2402 5.10, are also .NET compatible.
2403
2404 (d) A conditional reference to a named group can now use the syntax
2405 (?(<name>) or (?('name') as well as (?(name).
2406
2407 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2408 groups (named and numbered) that are never evaluated inline, but can be
2409 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2410 is always false. There may be only one alternative in such a group.
2411
2412 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2413 as the simple (?(R). The condition is true only if the most recent
2414 recursion is that of the given number or name. It does not search out
2415 through the entire recursion stack.
2416
2417 (g) The escape \gN or \g{N} has been added, where N is a positive or
2418 negative number, specifying an absolute or relative reference.
2419
2420 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2421 some "unreachable code" warnings.
2422
2423 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2424 things, this adds five new scripts.
2425
2426 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2427 There were also incompatibilities regarding the handling of \Q..\E inside
2428 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2429 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2430
2431 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2432 matches an empty string, and forcibly breaks the loop. There were bugs in
2433 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2434 against aaaa the result was just "a" rather than "aaaa", for example. Two
2435 separate and independent bugs (that affected different cases) have been
2436 fixed.
2437
2438 39. Refactored the code to abolish the use of different opcodes for small
2439 capturing bracket numbers. This is a tidy that I avoided doing when I
2440 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2441 The new approach is not only tidier, it makes it possible to reduce the
2442 memory needed to fix the previous bug (38).
2443
2444 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2445 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2446 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2447 mode.
2448
2449 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2450 report.
2451
2452 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2453 copy construction and assignment in the C++ wrapper.
2454
2455 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2456 couple of bugs that could have given wrong results in the "--newline=crlf"
2457 case.
2458
2459 44. Added a number of casts and did some reorganization of signed/unsigned int
2460 variables following suggestions from Dair Grant. Also renamed the variable
2461 "this" as "item" because it is a C++ keyword.
2462
2463 45. Arranged for dftables to add
2464
2465 #include "pcre_internal.h"
2466
2467 to pcre_chartables.c because without it, gcc 4.x may remove the array
2468 definition from the final binary if PCRE is built into a static library and
2469 dead code stripping is activated.
2470
2471 46. For an unanchored pattern, if a match attempt fails at the start of a
2472 newline sequence, and the newline setting is CRLF or ANY, and the next two
2473 characters are CRLF, advance by two characters instead of one.
2474
2475
2476 Version 6.7 04-Jul-06
2477 ---------------------
2478
2479 1. In order to handle tests when input lines are enormously long, pcretest has
2480 been re-factored so that it automatically extends its buffers when
2481 necessary. The code is crude, but this _is_ just a test program. The
2482 default size has been increased from 32K to 50K.
2483
2484 2. The code in pcre_study() was using the value of the re argument before
2485 testing it for NULL. (Of course, in any sensible call of the function, it
2486 won't be NULL.)
2487
2488 3. The memmove() emulation function in pcre_internal.h, which is used on
2489 systems that lack both memmove() and bcopy() - that is, hardly ever -
2490 was missing a "static" storage class specifier.
2491
2492 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2493 containing an extended class (one that cannot be represented by a bitmap
2494 because it contains high-valued characters or Unicode property items, e.g.
2495 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2496 pattern, but PCRE should not loop if you do not (it no longer does).
2497 [Detail: two cases were found: (a) a repeated subpattern containing an
2498 extended class; (b) a recursive reference to a subpattern that followed a
2499 previous extended class. It wasn't skipping over the extended class
2500 correctly when UTF-8 mode was not set.]
2501
2502 5. A negated single-character class was not being recognized as fixed-length
2503 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2504 compile error "lookbehind assertion is not fixed length".
2505
2506 6. The RunPerlTest auxiliary script was showing an unexpected difference
2507 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2508 write a Perl script that can interpret lines of an input file either as
2509 byte characters or as UTF-8, which is what "perltest" was being required to
2510 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2511 can't do is switch easily at run time between having the "use utf8;" pragma
2512 or not. In the end, I fudged it by using the RunPerlTest script to insert
2513 "use utf8;" explicitly for the UTF-8 tests.
2514
2515 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2516 the end of the subject string, contrary to the documentation and to what
2517 Perl does. This was true of both matching functions. Now it matches only at
2518 the start of the subject and immediately after *internal* newlines.
2519
2520 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2521 a pointer to an int instead of a pointer to an unsigned long int. This
2522 caused problems on 64-bit systems.
2523
2524 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2525 instance of the 'standard' template library not being so standard".
2526
2527 10. There was no check on the number of named subpatterns nor the maximum
2528 length of a subpattern name. The product of these values is used to compute
2529 the size of the memory block for a compiled pattern. By supplying a very
2530 long subpattern name and a large number of named subpatterns, the size
2531 computation could be caused to overflow. This is now prevented by limiting
2532 the length of names to 32 characters, and the number of named subpatterns
2533 to 10,000.
2534
2535 11. Subpatterns that are repeated with specific counts have to be replicated in
2536 the compiled pattern. The size of memory for this was computed from the
2537 length of the subpattern and the repeat count. The latter is limited to
2538 65535, but there was no limit on the former, meaning that integer overflow
2539 could in principle occur. The compiled length of a repeated subpattern is
2540 now limited to 30,000 bytes in order to prevent this.
2541
2542 12. Added the optional facility to have named substrings with the same name.
2543
2544 13. Added the ability to use a named substring as a condition, using the
2545 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2546 are numbers (not recommended). Forward references are permitted.
2547
2548 14. Added forward references in named backreferences (if you see what I mean).
2549
2550 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2551 pattern could run off the end of the subject. For example, the pattern
2552 "(?s)(.{1,5})"8 did this with the subject "ab".
2553
2554 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2555 PCRE_CASELESS was set when matching characters that were quantified with ?
2556 or *.
2557
2558 17. A character class other than a single negated character that had a minimum
2559 but no maximum quantifier - for example [ab]{6,} - was not handled
2560 correctly by pce_dfa_exec(). It would match only one character.
2561
2562 18. A valid (though odd) pattern that looked like a POSIX character
2563 class but used an invalid character after [ (for example [[,abc,]]) caused
2564 pcre_compile() to give the error "Failed: internal error: code overflow" or
2565 in some cases to crash with a glibc free() error. This could even happen if
2566 the pattern terminated after [[ but there just happened to be a sequence of
2567 letters, a binary zero, and a closing ] in the memory that followed.
2568
2569 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2570 over the years. Originally (before any Unicode support), just the bottom 8
2571 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2572 output from "man perlunicode" includes this:
2573
2574 The regular expression compiler produces polymorphic opcodes. That
2575 is, the pattern adapts to the data and automatically switches to
2576 the Unicode character scheme when presented with Unicode data--or
2577 instead uses a traditional byte scheme when presented with byte
2578 data.
2579
2580 Sadly, a wide octal escape does not cause a switch, and in a string with
2581 no other multibyte characters, these octal escapes are treated as before.
2582 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2583 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2584 Unicode string.
2585
2586 I have not perpetrated such confusion in PCRE. Up till now, it took just
2587 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2588 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2589 translate to the appropriate multibyte character.
2590
2591 29. Applied some refactoring to reduce the number of warnings from Microsoft
2592 and Borland compilers. This has included removing the fudge introduced
2593 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2594 a warning about an unused variable.
2595
2596 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2597 characters since release 4.0, because Perl (from release 5.004) does not.
2598 [Or at least, is documented not to: some releases seem to be in conflict
2599 with the documentation.] However, when a pattern was studied with
2600 pcre_study() and all its branches started with \s, PCRE still included VT
2601 as a possible starting character. Of course, this did no harm; it just
2602 caused an unnecessary match attempt.
2603
2604 22. Removed a now-redundant internal flag bit that recorded the fact that case
2605 dependency changed within the pattern. This was once needed for "required
2606 byte" processing, but is no longer used. This recovers a now-scarce options
2607 bit. Also moved the least significant internal flag bit to the most-
2608 significant bit of the word, which was not previously used (hangover from
2609 the days when it was an int rather than a uint) to free up another bit for
2610 the future.
2611
2612 23. Added support for CRLF line endings as well as CR and LF. As well as the
2613 default being selectable at build time, it can now be changed at runtime
2614 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2615 specify that it is scanning data with non-default line endings.
2616
2617 24. Changed the definition of CXXLINK to make it agree with the definition of
2618 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2619
2620 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2621 recursions. This makes a big different to stack usage for some patterns.
2622
2623 26. If a subpattern containing a named recursion or subroutine reference such
2624 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2625 the space required for the compiled pattern went wrong and gave too small a
2626 value. Depending on the environment, this could lead to "Failed: internal
2627 error: code overflow at offset 49" or "glibc detected double free or
2628 corruption" errors.
2629
2630 27. Applied patches from Google (a) to support the new newline modes and (b) to
2631 advance over multibyte UTF-8 characters in GlobalReplace.
2632
2633 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2634 difference for some implementation of PCRE in some Windows version.
2635
2636 29. Added some extra testing facilities to pcretest:
2637
2638 \q<number> in a data line sets the "match limit" value
2639 \Q<number> in a data line sets the "match recursion limt" value
2640 -S <number> sets the stack size, where <number> is in megabytes
2641
2642 The -S option isn't available for Windows.
2643
2644
2645 Version 6.6 06-Feb-06
2646 ---------------------
2647
2648 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2649 in pcreposix.h. I have copied the definition from pcre.h.
2650
2651 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2652 because pcre.h is no longer a built file.
2653
2654 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2655 not normally included in the compiled code.
2656
2657
2658 Version 6.5 01-Feb-06
2659 ---------------------
2660
2661 1. When using the partial match feature with pcre_dfa_exec(), it was not
2662 anchoring the second and subsequent partial matches at the new starting
2663 point. This could lead to incorrect results. For example, with the pattern
2664 /1234/, partially matching against "123" and then "a4" gave a match.
2665
2666 2. Changes to pcregrep:
2667
2668 (a) All non-match returns from pcre_exec() were being treated as failures
2669 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2670 error message is output. Some extra information is given for the
2671 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2672 probably the only errors that are likely to be caused by users (by
2673 specifying a regex that has nested indefinite repeats, for instance).
2674 If there are more than 20 of these errors, pcregrep is abandoned.
2675
2676 (b) A binary zero was treated as data while matching, but terminated the
2677 output line if it was written out. This has been fixed: binary zeroes
2678 are now no different to any other data bytes.
2679
2680 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2681 used to set a locale for matching. The --locale=xxxx long option has
2682 been added (no short equivalent) to specify a locale explicitly on the
2683 pcregrep command, overriding the environment variables.
2684
2685 (d) When -B was used with -n, some line numbers in the output were one less
2686 than they should have been.
2687
2688 (e) Added the -o (--only-matching) option.
2689
2690 (f) If -A or -C was used with -c (count only), some lines of context were
2691 accidentally printed for the final match.
2692
2693 (g) Added the -H (--with-filename) option.
2694
2695 (h) The combination of options -rh failed to suppress file names for files
2696 that were found from directory arguments.
2697
2698 (i) Added the -D (--devices) and -d (--directories) options.
2699
2700 (j) Added the -F (--fixed-strings) option.
2701
2702 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2703
2704 (l) Added the --colo(u)r option.
2705
2706 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2707 is not present by default.
2708
2709 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2710 items such as (?R) or (?1), when the recursion could match a number of
2711 alternatives. If it matched one of the alternatives, but subsequently,
2712 outside the recursion, there was a failure, the code tried to back up into
2713 the recursion. However, because of the way PCRE is implemented, this is not
2714 possible, and the result was an incorrect result from the match.
2715
2716 In order to prevent this happening, the specification of recursion has
2717 been changed so that all such subpatterns are automatically treated as
2718 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2719
2720 4. I had overlooked the fact that, in some locales, there are characters for
2721 which isalpha() is true but neither isupper() nor islower() are true. In
2722 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2723 and ordfeminine) are like this. This affected the treatment of \w and \W
2724 when they appeared in character classes, but not when they appeared outside
2725 a character class. The bit map for "word" characters is now created
2726 separately from the results of isalnum() instead of just taking it from the
2727 upper, lower, and digit maps. (Plus the underscore character, of course.)
2728
2729 5. The above bug also affected the handling of POSIX character classes such as
2730 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2731 permanent tables. Instead, the bit maps for such a class were previously
2732 created as the appropriate unions of the upper, lower, and digit bitmaps.
2733 Now they are created by subtraction from the [[:word:]] class, which has
2734 its own bitmap.
2735
2736 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2737 It is created by subtracting the vertical space characters (\x09, \x0a,
2738 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2739 subtraction was done in the overall bitmap for a character class, meaning
2740 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2741 be recognized. This bug has been fixed.
2742
2743 7. Patches from the folks at Google:
2744
2745 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2746 real life, but is still worth protecting against".
2747
2748 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2749 regular expressions".
2750
2751 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2752 have it.
2753
2754 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2755 "configure" and the latter not, in order to fix a problem somebody had
2756 with compiling the Arg class on HP-UX.
2757
2758 (e) Improve the error-handling of the C++ wrapper a little bit.
2759
2760 (f) New tests for checking recursion limiting.
2761
2762 8. The pcre_memmove() function, which is used only if the environment does not
2763 have a standard memmove() function (and is therefore rarely compiled),
2764 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2765 returning a result (though PCRE never actually uses the result).
2766
2767 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2768 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2769 returned instead of calling malloc() with an overflowing number that would
2770 most likely cause subsequent chaos.
2771
2772 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2773
2774 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2775 with this option is matched, the nmatch and pmatch options of regexec() are
2776 ignored.
2777
2778 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2779 provided in case anyone wants to the the POSIX interface with UTF-8
2780 strings.
2781
2782 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2783 C++ linking (needed for some HP-UX environments).
2784
2785 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2786 (unused parameter) and in the pcre_printint() function (omitted "default"
2787 switch label when the default is to do nothing).
2788
2789 15. Added some code to make it possible, when PCRE is compiled as a C++
2790 library, to replace subject pointers for pcre_exec() with a smart pointer
2791 class, thus making it possible to process discontinuous strings.
2792
2793 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2794 much the same function. They were added by different people who were trying
2795 to make PCRE easy to compile on non-Unix systems. It has been suggested
2796 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2797 for compiling on Windows systems. I have therefore replaced it with
2798 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2799 defaults to "extern" for C or "extern C" for C++, which works fine on
2800 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2801 SCOPE with something explicit in config.h. In addition:
2802
2803 (a) pcreposix.h still had just "extern" instead of either of these macros;
2804 I have replaced it with PCRE_DATA_SCOPE.
2805
2806 (b) Functions such as _pcre_xclass(), which are internal to the library,
2807 but external in the C sense, all had PCRE_EXPORT in their definitions.
2808 This is apparently wrong for the Windows case, so I have removed it.
2809 (It makes no difference on Unix-like systems.)
2810
2811 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2812 of recursive calls to match(). This is different to MATCH_LIMIT because
2813 that limits the total number of calls to match(), not all of which increase
2814 the depth of recursion. Limiting the recursion depth limits the amount of
2815 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2816 when PCRE is compiled, and changed at run time. A patch from Google adds
2817 this functionality to the C++ interface.
2818
2819 18. Changes to the handling of Unicode character properties:
2820
2821 (a) Updated the table to Unicode 4.1.0.
2822
2823 (b) Recognize characters that are not in the table as "Cn" (undefined).
2824
2825 (c) I revised the way the table is implemented to a much improved format
2826 which includes recognition of ranges. It now supports the ranges that
2827 are defined in UnicodeData.txt, and it also amalgamates other
2828 characters into ranges. This has reduced the number of entries in the
2829 table from around 16,000 to around 3,000, thus reducing its size
2830 considerably. I realized I did not need to use a tree structure after
2831 all - a binary chop search is just as efficient. Having reduced the
2832 number of entries, I extended their size from 6 bytes to 8 bytes to
2833 allow for more data.
2834
2835 (d) Added support for Unicode script names via properties such as \p{Han}.
2836
2837 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2838 matching that character.
2839
2840 20. When matching a repeated Unicode property with a minimum greater than zero,
2841 (for example \pL{2,}), PCRE could look past the end of the subject if it
2842 reached it while seeking the minimum number of characters. This could
2843 happen only if some of the characters were more than one byte long, because
2844 there is a check for at least the minimum number of bytes.
2845
2846 21. Refactored the implementation of \p and \P so as to be more general, to
2847 allow for more different types of property in future. This has changed the
2848 compiled form incompatibly. Anybody with saved compiled patterns that use
2849 \p or \P will have to recompile them.
2850
2851 22. Added "Any" and "L&" to the supported property types.
2852
2853 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2854 but give a compile time error if the value is greater than 0xff.
2855
2856 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2857 accidentally not being installed or uninstalled.
2858
2859 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2860 made were to insert the current release number. This seemed silly, because
2861 it made things harder for people building PCRE on systems that don't run
2862 "configure". I have turned pcre.h into a distributed file, no longer built
2863 by "configure", with the version identification directly included. There is
2864 no longer a pcre.h.in file.
2865
2866 However, this change necessitated a change to the pcre-config script as
2867 well. It is built from pcre-config.in, and one of the substitutions was the
2868 release number. I have updated configure.ac so that ./configure now finds
2869 the release number by grepping pcre.h.
2870
2871 26. Added the ability to run the tests under valgrind.
2872
2873
2874 Version 6.4 05-Sep-05
2875 ---------------------
2876
2877 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2878 "--" to be printed when multiple files were scanned, even when none of the
2879 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2880 consider it to be a bug, and have restored the previous behaviour.
2881
2882 2. A couple of code tidies to get rid of compiler warnings.
2883
2884 3. The pcretest program used to cheat by referring to symbols in the library
2885 whose names begin with _pcre_. These are internal symbols that are not
2886 really supposed to be visible externally, and in some environments it is
2887 possible to suppress them. The cheating is now confined to including
2888 certain files from the library's source, which is a bit cleaner.
2889
2890 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2891 file's purpose clearer.
2892
2893 5. Reorganized pcre_ucp_findchar().
2894
2895
2896 Version 6.3 15-Aug-05
2897 ---------------------
2898
2899 1. The file libpcre.pc.in did not have general read permission in the tarball.
2900
2901 2. There were some problems when building without C++ support:
2902
2903 (a) If C++ support was not built, "make install" and "make test" still
2904 tried to test it.
2905
2906 (b) There were problems when the value of CXX was explicitly set. Some
2907 changes have been made to try to fix these, and ...
2908
2909 (c) --disable-cpp can now be used to explicitly disable C++ support.
2910
2911 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2912 backslash in a target when C++ was disabled. This confuses some
2913 versions of "make", apparently. Using an intermediate variable solves
2914 this. (Same for CPP_LOBJ.)
2915
2916 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2917 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2918 necessary on certain architectures.
2919
2920 4. Added a setting of -export-symbols-regex to the link command to remove
2921 those symbols that are exported in the C sense, but actually are local
2922 within the library, and not documented. Their names all begin with
2923 "_pcre_". This is not a perfect job, because (a) we have to except some
2924 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2925 available (and never for static libraries). I have made a note to try to
2926 find a way round (a) in the future.
2927
2928
2929 Version 6.2 01-Aug-05
2930 ---------------------
2931
2932 1. There was no test for integer overflow of quantifier values. A construction
2933 such as {1111111111111111} would give undefined results. What is worse, if
2934 a minimum quantifier for a parenthesized subpattern overflowed and became
2935 negative, the calculation of the memory size went wrong. This could have
2936 led to memory overwriting.
2937
2938 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2939
2940 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2941 operating environments where this matters.
2942
2943 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2944 PCRE options from within the C++ wrapper.
2945
2946 5. Named capturing subpatterns were not being correctly counted when a pattern
2947 was compiled. This caused two problems: (a) If there were more than 100
2948 such subpatterns, the calculation of the memory needed for the whole
2949 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2950 back references of the form \12, where the number was greater than 9, were
2951 not recognized as back references, even though there were sufficient
2952 previous subpatterns.
2953
2954 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2955 versions of gcc, e.g. 2.95.4.
2956
2957
2958 Version 6.1 21-Jun-05
2959 ---------------------
2960
2961 1. There was one reference to the variable "posix" in pcretest.c that was not
2962 surrounded by "#if !defined NOPOSIX".
2963
2964 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2965 the cross-check on the old pcre_info() function, for the benefit of the
2966 cut-down version of PCRE that is currently imported into Exim.
2967
2968 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2969 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2970 patterns that start (?i) but I don't think that matters. The use of (?i) is
2971 just an example; this all applies to the other options as well.
2972
2973 4. Since libtool seems to echo the compile commands it is issuing, the output
2974 from "make" can be reduced a bit by putting "@" in front of each libtool
2975 compile command.
2976
2977 5. Patch from the folks at Google for configure.in to be a bit more thorough
2978 in checking for a suitable C++ installation before trying to compile the
2979 C++ stuff. This should fix a reported problem when a compiler was present,
2980 but no suitable headers.
2981
2982 6. The man pages all had just "PCRE" as their title. I have changed them to
2983 be the relevant file name. I have also arranged that these names are
2984 retained in the file doc/pcre.txt, which is a concatenation in text format
2985 of all the man pages except the little individual ones for each function.
2986
2987 7. The NON-UNIX-USE file had not been updated for the different set of source
2988 files that come with release 6. I also added a few comments about the C++
2989 wrapper.
2990
2991
2992 Version 6.0 07-Jun-05
2993 ---------------------
2994
2995 1. Some minor internal re-organization to help with my DFA experiments.
2996
2997 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2998 didn't matter for the library itself when fully configured, but did matter
2999 when compiling without UCP support, or within Exim, where the ucp files are
3000 not imported.
3001
3002 3. Refactoring of the library code to split up the various functions into
3003 different source modules. The addition of the new DFA matching code (see
3004 below) to a single monolithic source would have made it really too
3005 unwieldy, quite apart from causing all the code to be include in a
3006 statically linked application, when only some functions are used. This is
3007 relevant even without the DFA addition now that patterns can be compiled in
3008 one application and matched in another.
3009
3010 The downside of splitting up is that there have to be some external
3011 functions and data tables that are used internally in different modules of
3012 the library but which are not part of the API. These have all had their
3013 names changed to start with "_pcre_" so that they are unlikely to clash
3014 with other external names.
3015
3016 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3017 a different (DFA) algorithm. Although it is slower than the original
3018 function, it does have some advantages for certain types of matching
3019 problem.
3020
3021 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3022 including restarting after a partial match.
3023
3024 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3025 defined when compiling for Windows was sent to me. I have put it into the
3026 code, though I have no means of testing or verifying it.
3027
3028 7. Added the pcre_refcount() auxiliary function.
3029
3030 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3031 match before or at the first newline in the subject string. In pcretest,
3032 the /f option on a pattern can be used to set this.
3033
3034 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3035 would behave wrongly. This has been present in PCRE since release 4.0.
3036
3037 10. A number of changes to the pcregrep command:
3038
3039 (a) Refactored how -x works; insert ^(...)$ instead of setting
3040 PCRE_ANCHORED and checking the length, in preparation for adding
3041 something similar for -w.
3042
3043 (b) Added the -w (match as a word) option.
3044
3045 (c) Refactored the way lines are read and buffered so as to have more
3046 than one at a time available.
3047
3048 (d) Implemented a pcregrep test script.
3049
3050 (e) Added the -M (multiline match) option. This allows patterns to match
3051 over several lines of the subject. The buffering ensures that at least
3052 8K, or the rest of the document (whichever is the shorter) is available
3053 for matching (and similarly the previous 8K for lookbehind assertions).
3054
3055 (f) Changed the --help output so that it now says
3056
3057 -w, --word-regex(p)
3058
3059 instead of two lines, one with "regex" and the other with "regexp"
3060 because that confused at least one person since the short forms are the
3061 same. (This required a bit of code, as the output is generated
3062 automatically from a table. It wasn't just a text change.)
3063
3064 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3065 option but starts with a hyphen. Could be a pattern or a path name
3066 starting with a hyphen, for instance.
3067
3068 (h) "-" can be given as a file name to represent stdin.
3069
3070 (i) When file names are being printed, "(standard input)" is used for
3071 the standard input, for compatibility with GNU grep. Previously
3072 "<stdin>" was used.
3073
3074 (j) The option --label=xxx can be used to supply a name to be used for
3075 stdin when file names are being printed. There is no short form.
3076
3077 (k) Re-factored the options decoding logic because we are going to add
3078 two more options that take data. Such options can now be given in four
3079 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3080
3081 (l) Added the -A, -B, and -C options for requesting that lines of context
3082 around matches be printed.
3083
3084 (m) Added the -L option to print the names of files that do not contain
3085 any matching lines, that is, the complement of -l.
3086
3087 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3088 continue to scan other files.
3089
3090 (o) The -s option was incorrectly implemented. For compatibility with other
3091 greps, it now suppresses the error message for a non-existent or non-
3092 accessible file (but not the return code). There is a new option called
3093 -q that suppresses the output of matching lines, which was what -s was
3094 previously doing.
3095
3096 (p) Added --include and --exclude options to specify files for inclusion
3097 and exclusion when recursing.
3098
3099 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3100 Hopefully, it now does.
3101
3102 12. Missing cast in pcre_study().
3103
3104 13. Added an "uninstall" target to the makefile.
3105
3106 14. Replaced "extern" in the function prototypes in Makefile.in with
3107 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3108 world, but is set differently for Windows.
3109
3110 15. Added a second compiling function called pcre_compile2(). The only
3111 difference is that it has an extra argument, which is a pointer to an
3112 integer error code. When there is a compile-time failure, this is set
3113 non-zero, in addition to the error test pointer being set to point to an
3114 error message. The new argument may be NULL if no error number is required
3115 (but then you may as well call pcre_compile(), which is now just a
3116 wrapper). This facility is provided because some applications need a
3117 numeric error indication, but it has also enabled me to tidy up the way
3118 compile-time errors are handled in the POSIX wrapper.
3119
3120 16. Added VPATH=.libs to the makefile; this should help when building with one
3121 prefix path and installing with another. (Or so I'm told by someone who
3122 knows more about this stuff than I do.)
3123
3124 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3125 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3126 match everything, including newlines. This is not POSIX-compatible, but
3127 somebody wanted the feature. From pcretest it can be activated by using
3128 both the P and the s flags.
3129
3130 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3131
3132 19. libpcre.pc was being incorrectly installed as executable.
3133
3134 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3135 it now also looks for '\r' so that it will work unmodified on Windows.
3136
3137 21. Added Google's contributed C++ wrapper to the distribution.
3138
3139 22. Added some untidy missing memory free() calls in pcretest, to keep
3140 Electric Fence happy when testing.
3141
3142
3143
3144 Version 5.0 13-Sep-04
3145 ---------------------
3146
3147 1. Internal change: literal characters are no longer packed up into items
3148 containing multiple characters in a single byte-string. Each character
3149 is now matched using a separate opcode. However, there may be more than one
3150 byte in the character in UTF-8 mode.
3151
3152 2. The pcre_callout_block structure has two new fields: pattern_position and
3153 next_item_length. These contain the offset in the pattern to the next match
3154 item, and its length, respectively.
3155
3156 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3157 insertion of callouts before each pattern item. Added the /C option to
3158 pcretest to make use of this.
3159
3160 4. On the advice of a Windows user, the lines
3161
3162 #if defined(_WIN32) || defined(WIN32)
3163 _setmode( _fileno( stdout ), 0x8000 );
3164 #endif /* defined(_WIN32) || defined(WIN32) */
3165
3166 have been added to the source of pcretest. This apparently does useful
3167 magic in relation to line terminators.
3168
3169 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3170 for the benefit of those environments where the "b" makes a difference.
3171
3172 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3173 to know about it. I have put a hack into configure.in that adds in code
3174 to set GCC=yes if CC=icc. This seems to end up at a point in the
3175 generated configure script that is early enough to affect the setting of
3176 compiler options, which is what is needed, but I have no means of testing
3177 whether it really works. (The user who reported this had patched the
3178 generated configure script, which of course I cannot do.)
3179
3180 LATER: After change 22 below (new libtool files), the configure script
3181 seems to know about icc (and also ecc). Therefore, I have commented out
3182 this hack in configure.in.
3183
3184 7. Added support for pkg-config (2 patches were sent in).
3185
3186 8. Negated POSIX character classes that used a combination of internal tables
3187 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3188 [[:^ascii]]. Typically, they would match almost any characters. The other
3189 POSIX classes were not broken in this way.
3190
3191 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3192 to find the match, as PCRE was deluded into thinking that the match had to
3193 start at the start point or following a newline. The same bug applied to
3194 patterns with negative forward assertions or any backward assertions
3195 preceding ".*" at the start, unless the pattern required a fixed first
3196 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3197
3198 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3199 starting at the last subject character, bytes beyond the end of the subject
3200 string were read.
3201
3202 11. Renamed the variable "class" as "classbits" to make life easier for C++
3203 users. (Previously there was a macro definition, but it apparently wasn't
3204 enough.)
3205
3206 12. Added the new field "tables" to the extra data so that tables can be passed
3207 in at exec time, or the internal tables can be re-selected. This allows
3208 a compiled regex to be saved and re-used at a later time by a different
3209 program that might have everything at different addresses.
3210
3211 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3212 -R library as well as a -L library.
3213
3214 14. The debugging options of pcretest (-d on the command line or D on a
3215 pattern) showed incorrect output for anything following an extended class
3216 that contained multibyte characters and which was followed by a quantifier.
3217
3218 15. Added optional support for general category Unicode character properties
3219 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3220 support. It adds about 90K to the size of the library. The meanings of the
3221 inbuilt class escapes such as \d and \s have NOT been changed.
3222
3223 16. Updated pcredemo.c to include calls to free() to release the memory for the
3224 compiled pattern.
3225
3226 17. The generated file chartables.c was being created in the source directory
3227 instead of in the building directory. This caused the build to fail if the
3228 source directory was different from the building directory, and was
3229 read-only.
3230
3231 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3232 file. No doubt somebody will tell me if they don't make sense... Also added
3233 Dan Mooney's comments about building on OpenVMS.
3234
3235 19. Added support for partial matching via the PCRE_PARTIAL option for
3236 pcre_exec() and the \P data escape in pcretest.
3237
3238 20. Extended pcretest with 3 new pattern features:
3239
3240 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3241 write the compiled pattern to the file whose name is "rest-of-line".
3242 This is a straight binary dump of the data, with the saved pointer to
3243 the character tables forced to be NULL. The study data, if any, is
3244 written too. After writing, pcretest reads a new pattern.
3245
3246 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3247 compiled pattern from the given file. There must not be any
3248 occurrences of "<" in the file name (pretty unlikely); if there are,
3249 pcretest will instead treat the initial "<" as a pattern delimiter.
3250 After reading in the pattern, pcretest goes on to read data lines as
3251 usual.
3252
3253 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3254 and 16-bit fields in a compiled pattern, to simulate a pattern that
3255 was compiled on a host of opposite endianness.
3256
3257 21. The pcre-exec() function can now cope with patterns that were compiled on
3258 hosts of opposite endianness, with this restriction:
3259
3260 As for any compiled expression that is saved and used later, the tables
3261 pointer field cannot be preserved; the extra_data field in the arguments
3262 to pcre_exec() should be used to pass in a tables address if a value
3263 other than the default internal tables were used at compile time.
3264
3265 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3266 now diagnosed as an error. Previously, most of the time, a negative number
3267 would have been treated as zero, but if in addition "ovector" was passed as
3268 NULL, a crash could occur.
3269
3270 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3271 new versions from the libtool 1.5 distribution (the last one is a copy of
3272 a file called libtool.m4). This seems to have fixed the need to patch
3273 "configure" to support Darwin 1.3 (which I used to do). However, I still
3274 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3275 workstation).
3276
3277 24. Changed the PCRE licence to be the more standard "BSD" licence.
3278
3279
3280 Version 4.5 01-Dec-03
3281 ---------------------
3282
3283 1. There has been some re-arrangement of the code for the match() function so
3284 that it can be compiled in a version that does not call itself recursively.
3285 Instead, it keeps those local variables that need separate instances for
3286 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3287 needs to "recurse". Keeping track of where control must go is done by means
3288 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3289 hide most of the details from the main code, and operates only if
3290 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3291 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3292 operating.
3293
3294 To make it easier for callers to provide specially tailored get/free
3295 functions for this usage, two new functions, pcre_stack_malloc, and
3296 pcre_stack_free, are used. They are always called in strict stacking order,
3297 and the size of block requested is always the same.
3298
3299 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3300 PCRE has been compiled to use the stack or the heap for recursion. The
3301 -C option of pcretest uses this to show which version is compiled.
3302
3303 A new data escape \S, is added to pcretest; it causes the amounts of store
3304 obtained and freed by both kinds of malloc/free at match time to be added
3305 to the output.
3306
3307 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3308 what's available on my current Linux desktop machine.
3309
3310 3. When matching a UTF-8 string, the test for a valid string at the start has
3311 been extended. If start_offset is not zero, PCRE now checks that it points
3312 to a byte that is the start of a UTF-8 character. If not, it returns
3313 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3314 this is necessary because there may be backward assertions in the pattern.
3315 When matching the same subject several times, it may save resources to use
3316 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3317
3318 4. The code for checking the validity of UTF-8 strings has been tightened so
3319 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3320 containing "overlong sequences".
3321
3322 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3323 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3324 should have been "&", but it just so happened that all the cases this let
3325 through by mistake were picked up later in the function.
3326
3327 6. I had used a variable called "isblank" - this is a C99 function, causing
3328 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3329
3330 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3331 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3332 so the version is shown; (c) stop "make test" echoing "./RunTest".
3333
3334 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3335
3336 9. The prototype for memmove() for systems that don't have it was using
3337 size_t, but the inclusion of the header that defines size_t was later. I've
3338 moved the #includes for the C headers earlier to avoid this.
3339
3340 10. Added some adjustments to the code to make it easier to compiler on certain
3341 special systems:
3342
3343 (a) Some "const" qualifiers were missing.
3344 (b) Added the macro EXPORT before all exported functions; by default this
3345 is defined to be empty.
3346 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3347 that it reads its output file name as an argument instead of writing
3348 to the standard output and assuming this can be redirected.
3349
3350 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3351 class containing characters with values greater than 255, PCRE compilation
3352 went into a loop.
3353
3354 12. A recursive reference to a subpattern that was within another subpattern
3355 that had a minimum quantifier of zero caused PCRE to crash. For example,
3356 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3357 recursion. If the recursively-called subpattern itself had a zero repeat,
3358 that was OK.
3359
3360 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3361 buffer into which it was copied (for escape processing) was still set at
3362 1024, so long lines caused crashes.
3363
3364 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3365 "internal error: code overflow...". This applied to any character class
3366 that was followed by a possessive quantifier.
3367
3368 15. Modified the Makefile to add libpcre.la as a prerequisite for
3369 libpcreposix.la because I was told this is needed for a parallel build to
3370 work.
3371
3372 16. If a pattern that contained .* following optional items at the start was
3373 studied, the wrong optimizing data was generated, leading to matching
3374 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3375 matching string must start with a or b or c. The correct conclusion for
3376 this pattern is that a match can start with any character.
3377
3378
3379 Version 4.4 13-Aug-03
3380 ---------------------
3381
3382 1. In UTF-8 mode, a character class containing characters with values between
3383 127 and 255 was not handled correctly if the compiled pattern was studied.
3384 In fixing this, I have also improved the studying algorithm for such
3385 classes (slightly).
3386
3387 2. Three internal functions had redundant arguments passed to them. Removal
3388 might give a very teeny performance improvement.
3389
3390 3. Documentation bug: the value of the capture_top field in a callout is *one
3391 more than* the number of the hightest numbered captured substring.
3392
3393 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3394 in incorrectly linking with a previously installed version. They now link
3395 explicitly with libpcre.la.
3396
3397 5. configure.in no longer needs to recognize Cygwin specially.
3398
3399 6. A problem in pcre.in for Windows platforms is fixed.
3400
3401 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3402 pcretest, it used to include the size of the study block as part of its
3403 output. Unfortunately, the structure contains a field that has a different
3404 size on different hardware architectures. This meant that the tests that
3405 showed this size failed. As the block is currently always of a fixed size,
3406 this information isn't actually particularly useful in pcretest output, so
3407 I have just removed it.
3408
3409 8. Three pre-processor statements accidentally did not start in column 1.
3410 Sadly, there are *still* compilers around that complain, even though
3411 standard C has not required this for well over a decade. Sigh.
3412
3413 9. In pcretest, the code for checking callouts passed small integers in the
3414 callout_data field, which is a void * field. However, some picky compilers
3415 complained about the casts involved for this on 64-bit systems. Now
3416 pcretest passes the address of the small integer instead, which should get
3417 rid of the warnings.
3418
3419 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3420 both compile and run time, and gives an error if an invalid UTF-8 sequence
3421 is found. There is a option for disabling this check in cases where the
3422 string is known to be correct and/or the maximum performance is wanted.
3423
3424 11. In response to a bug report, I changed one line in Makefile.in from
3425
3426 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3427 to
3428 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3429
3430 to look similar to other lines, but I have no way of telling whether this
3431 is the right thing to do, as I do not use Windows. No doubt I'll get told
3432 if it's wrong...
3433
3434
3435 Version 4.3 21-May-03
3436 ---------------------
3437
3438 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3439 Makefile.
3440
3441 2. Some refactoring to improve the quality of the code:
3442
3443 (i) The utf8_table... variables are now declared "const".
3444
3445 (ii) The code for \cx, which used the "case flipping" table to upper case
3446 lower case letters, now just substracts 32. This is ASCII-specific,
3447 but the whole concept of \cx is ASCII-specific, so it seems
3448 reasonable.
3449
3450 (iii) PCRE was using its character types table to recognize decimal and
3451 hexadecimal digits in the pattern. This is silly, because it handles
3452 only 0-9, a-f, and A-F, but the character types table is locale-
3453 specific, which means strange things might happen. A private
3454 table is now used for this - though it costs 256 bytes, a table is
3455 much faster than multiple explicit tests. Of course, the standard
3456 character types table is still used for matching digits in subject
3457 strings against \d.
3458
3459 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3460 ending in _t are). So I've renamed it as ESC_tee.
3461
3462 3. The first argument for regexec() in the POSIX wrapper should have been
3463 defined as "const".
3464
3465 4. Changed pcretest to use malloc() for its buffers so that they can be
3466 Electric Fenced for debugging.
3467
3468 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3469 to read one or more bytes before the start of the subject string. Often this
3470 had no effect on PCRE's behaviour, but in some circumstances it could
3471 provoke a segmentation fault.
3472
3473 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3474 to try to read one or more bytes before the start of the subject string.
3475
3476 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3477 UTF-8 support could misbehave in various ways if the subject string
3478 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3479 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3480 back over UTF-8 characters.)
3481
3482
3483 Version 4.2 14-Apr-03
3484 ---------------------
3485
3486 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3487
3488 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3489 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3490 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3491 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3492 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3493 and BUILD_EXEEXT
3494 Note: automatic setting of the BUILD variables is not yet working
3495 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3496 compile-time but not at link-time
3497 [LINK]: use for linking executables only
3498 make different versions for Windows and non-Windows
3499 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3500 libraries
3501 [LINK_FOR_BUILD]: new variable
3502 [OBJEXT]: use throughout
3503 [EXEEXT]: use throughout
3504 <winshared>: new target
3505 <wininstall>: new target
3506 <dftables.o>: use native compiler
3507 <dftables>: use native linker
3508 <install>: handle Windows platform correctly
3509 <clean>: ditto
3510 <check>: ditto
3511 copy DLL to top builddir before testing
3512
3513 As part of these changes, -no-undefined was removed again. This was reported
3514 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3515 in any case.
3516
3517 3. Some tidies to get rid of compiler warnings:
3518
3519 . In the match_data structure, match_limit was an unsigned long int, whereas
3520 match_call_count was an int. I've made them both unsigned long ints.
3521
3522 . In pcretest the fact that a const uschar * doesn't automatically cast to
3523 a void * provoked a warning.
3524
3525 . Turning on some more compiler warnings threw up some "shadow" variables
3526 and a few more missing casts.
3527
3528 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3529 option, a class that contained a single character with a value between 128
3530 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3531
3532 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3533 option, a class that contained several characters, but with at least one
3534 whose value was between 128 and 255 caused PCRE to crash.
3535
3536
3537 Version 4.1 12-Mar-03
3538 ---------------------
3539
3540 1. Compiling with gcc -pedantic found a couple of places where casts were
3541 needed, and a string in dftables.c that was longer than standard compilers are
3542 required to support.
3543
3544 2. Compiling with Sun's compiler found a few more places where the code could
3545 be tidied up in order to avoid warnings.
3546
3547 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3548 first of these names is deprecated in the latest Autoconf in favour of the name
3549 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3550 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3551 analogy I have changed it to CFLAGS_FOR_BUILD.
3552
3553 4. Added -no-undefined to the linking command in the Makefile, because this is
3554 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3555 linking step for the pcreposix library.
3556
3557 5. PCRE was failing to diagnose the case of two named groups with the same
3558 name.
3559
3560 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3561 literal character that is needed in the subject for a match, and scans along to
3562 ensure that it is present before embarking on the full matching process. This
3563 saves time in cases of nested unlimited repeats that are never going to match.
3564 Problem: the scan can take a lot of time if the subject is very long (e.g.
3565 megabytes), thus penalizing straightforward matches. It is now done only if the
3566 amount of subject to be scanned is less than 1000 bytes.
3567
3568 7. A lesser problem with the same optimization is that it was recording the
3569 first character of an anchored pattern as "needed", thus provoking a search
3570 right along the subject, even when the first match of the pattern was going to
3571 fail. The "needed" character is now not set for anchored patterns, unless it
3572 follows something in the pattern that is of non-fixed length. Thus, it still
3573 fulfils its original purpose of finding quick non-matches in cases of nested
3574 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3575
3576
3577 Version 4.0 17-Feb-03
3578 ---------------------
3579
3580 1. If a comment in an extended regex that started immediately after a meta-item
3581 extended to the end of string, PCRE compiled incorrect data. This could lead to
3582 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3583
3584 2. Moved to autoconf 2.53 and libtool 1.4.2.
3585
3586 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3587 the special perltest8 script is no longer needed - all the tests can be run
3588 from a single perltest script.
3589
3590 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3591 by \s. It has now been removed in PCRE. This means it isn't recognized as
3592 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3593 class [:space:] *does* include VT, thereby creating a mess.
3594
3595 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3596 space and tab.
3597
3598 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3599 its new features into the main test script, reducing the number of scripts.
3600
3601 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3602 were backward compatible, and made the (?i) apply to the whole pattern, as if
3603 /i were given. Now it behaves more logically, and applies the option setting
3604 only to what follows. PCRE has been changed to follow suit. However, if it
3605 finds options settings right at the start of the pattern, it extracts them into
3606 the global options, as before. Thus, they show up in the info data.
3607
3608 8. Added support for the \Q...\E escape sequence. Characters in between are
3609 treated as literals. This is slightly different from Perl in that $ and @ are
3610 also handled as literals inside the quotes. In Perl, they will cause variable
3611 interpolation. Note the following examples:
3612
3613 Pattern PCRE matches Perl matches
3614
3615 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3616 \Qabc\$xyz\E abc\$xyz abc\$xyz
3617 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3618
3619 For compatibility with Perl, \Q...\E sequences are recognized inside character
3620 classes as well as outside them.
3621
3622 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3623 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3624 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3625 signed/unsigned warnings.
3626
3627 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3628 option for pcretest, so I've replaced it by a simple function that does just
3629 that job.
3630
3631 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3632 "pcregrep -".
3633
3634 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3635 Java package. This provides some syntactic sugar for simple cases of what my
3636 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3637 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3638 item, you can use this simplified notation. Note that only makes sense with
3639 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3640 greediness, whatever the setting of the PCRE_UNGREEDY option.
3641
3642 13. A change of greediness default within a pattern was not taking effect at
3643 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3644 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3645 was abstracted outside.
3646
3647 14. PCRE now supports the \G assertion. It is true when the current matching
3648 position is at the start point of the match. This differs from \A when the
3649 starting offset is non-zero. Used with the /g option of pcretest (or similar
3650 code), it works in the same way as it does for Perl's /g option. If all
3651 alternatives of a regex begin with \G, the expression is anchored to the start
3652 match position, and the "anchored" flag is set in the compiled expression.
3653
3654 15. Some bugs concerning the handling of certain option changes within patterns
3655 have been fixed. These applied to options other than (?ims). For example,
3656 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3657 been the other way round. Some of this was related to change 7 above.
3658
3659 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3660 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3661 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3662 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3663
3664 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3665 mode. Unlike ".", it always matches newline, whatever the setting of
3666 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3667 assertions. Perl allows it, but it doesn't (in general) work because it can't
3668 calculate the length of the lookbehind. At least, that's the case for Perl
3669 5.8.0 - I've been told they are going to document that it doesn't work in
3670 future.
3671
3672 18. Added an error diagnosis for escapes that PCRE does not support: these are
3673 \L, \l, \N, \P, \p, \U, \u, and \X.
3674
3675 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3676 reading past the end of the pattern in cases such as /[abcd/.
3677
3678 20. PCRE was getting more memory than necessary for patterns with classes that
3679 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3680
3681 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3682 compiling PCRE for use with Virtual Pascal.
3683
3684 22. Small fix to the Makefile to make it work properly if the build is done
3685 outside the source tree.
3686
3687 23. Added a new extension: a condition to go with recursion. If a conditional
3688 subpattern starts with (?(R) the "true" branch is used if recursion has
3689 happened, whereas the "false" branch is used only at the top level.
3690
3691 24. When there was a very long string of literal characters (over 255 bytes
3692 without UTF support, over 250 bytes with UTF support), the computation of how
3693 much memory was required could be incorrect, leading to segfaults or other
3694 strange effects.
3695
3696 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3697 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3698 there was a subsequent back reference to those brackets. This meant that, for
3699 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3700 possible to check for precisely this case. All we can do is abandon the
3701 optimization if .* occurs inside capturing brackets when there are any back
3702 references whatsoever. (See below for a better fix that came later.)
3703
3704 26. The handling of the optimization for finding the first character of a
3705 non-anchored pattern, and for finding a character that is required later in the
3706 match were failing in some cases. This didn't break the matching; it just
3707 failed to optimize when it could. The way this is done has been re-implemented.
3708
3709 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3710
3711 28. Added a new feature that provides some of the functionality that Perl
3712 provides with (?{...}). The facility is termed a "callout". The way it is done
3713 in PCRE is for the caller to provide an optional function, by setting
3714 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3715 global variable. By default it is unset, which disables all calling out. To get
3716 the function called, the regex must include (?C) at appropriate points. This
3717 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3718 This provides a means of identifying different callout points. When PCRE
3719 reaches such a point in the regex, if pcre_callout has been set, the external
3720 function is called. It is provided with data in a structure called
3721 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3722 matching continues; if it returns a non-zero value, the match at the current
3723 point fails. However, backtracking will occur if possible. [This was changed
3724 later and other features added - see item 49 below.]
3725
3726 29. pcretest is upgraded to test the callout functionality. It provides a
3727 callout function that displays information. By default, it shows the start of
3728 the match and the current position in the text. There are some new data escapes
3729 to vary what happens:
3730
3731 \C+ in addition, show current contents of captured substrings
3732 \C- do not supply a callout function
3733 \C!n return 1 when callout number n is reached
3734 \C!n!m return 1 when callout number n is reached for the mth time
3735
3736 30. If pcregrep was called with the -l option and just a single file name, it
3737 output "<stdin>" if a match was found, instead of the file name.
3738
3739 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3740 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3741 pcre_exec(). This saves a malloc/free per call. The default value of
3742 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3743 when configuring.
3744
3745 32. The default maximum size of a compiled pattern is 64K. There have been a
3746 few cases of people hitting this limit. The code now uses macros to handle the
3747 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3748 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3749 configuring. Tests 2 and 5 work only with 2-byte links because they output
3750 debugging information about compiled patterns.
3751
3752 33. Internal code re-arrangements:
3753
3754 (a) Moved the debugging function for printing out a compiled regex into
3755 its own source file (printint.c) and used #include to pull it into
3756 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3757 separate copies.
3758
3759 (b) Defined the list of op-code names for debugging as a macro in
3760 internal.h so that it is next to the definition of the opcodes.
3761
3762 (c) Defined a table of op-code lengths for simpler skipping along compiled
3763 code. This is again a macro in internal.h so that it is next to the
3764 definition of the opcodes.
3765
3766 34. Added support for recursive calls to individual subpatterns, along the
3767 lines of Robin Houston's patch (but implemented somewhat differently).
3768
3769 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3770 allow it to read and process whole directories in Win32. This code was
3771 contributed by Lionel Fourquaux; it has not been tested by me.
3772
3773 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3774 used to name a group. Names consist of alphanumerics and underscores, and must
3775 be unique. Back references use the syntax (?P=name) and recursive calls use
3776 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3777 numbers. The function pcre_fullinfo() can be used after compilation to extract
3778 a name/number map. There are three relevant calls:
3779
3780 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3781 PCRE_INFO_NAMECOUNT yields the number of entries
3782 PCRE_INFO_NAMETABLE yields a pointer to the map.
3783
3784 The map is a vector of fixed-size entries. The size of each entry depends on
3785 the length of the longest name used. The first two bytes of each entry are the
3786 group number, most significant byte first. There follows the corresponding
3787 name, zero terminated. The names are in alphabetical order.
3788
3789 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3790 case instead of 255. Making it the same both with and without UTF-8 support
3791 means that the same test output works with both.
3792
3793 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3794 calling malloc() with a zero argument.
3795
3796 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3797 optimization. I've improved things by keeping a bitmap of backreferences with
3798 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3799 fact referenced, the optimization can be applied. It is unlikely that a
3800 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3801 the match to follow \n) will appear inside brackets with a number greater than
3802 31, but if it does, any back reference > 31 suppresses the optimization.
3803
3804 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3805 of disabling numbered capturing parentheses. Any opening parenthesis that is
3806 not followed by ? behaves as if it were followed by ?: but named parentheses
3807 can still be used for capturing (and they will acquire numbers in the usual
3808 way).
3809
3810 41. Redesigned the return codes from the match() function into yes/no/error so
3811 that errors can be passed back from deep inside the nested calls. A malloc
3812 failure while inside a recursive subpattern call now causes the
3813 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3814
3815 42. It is now possible to set a limit on the number of times the match()
3816 function is called in a call to pcre_exec(). This facility makes it possible to
3817 limit the amount of recursion and backtracking, though not in a directly
3818 obvious way, because the match() function is used in a number of different
3819 circumstances. The count starts from zero for each position in the subject
3820 string (for non-anchored patterns). The default limit is, for compatibility, a
3821 large number, namely 10 000 000. You can change this in two ways:
3822
3823 (a) When configuring PCRE before making, you can use --with-match-limit=n
3824 to set a default value for the compiled library.
3825
3826 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3827 a different value is set. See 45 below.
3828
3829 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3830
3831 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3832 of things that can be changed at compile time. The first argument specifies
3833 what is wanted and the second points to where the information is to be placed.
3834 The current list of available information is:
3835
3836 PCRE_CONFIG_UTF8
3837
3838 The output is an integer that is set to one if UTF-8 support is available;
3839 otherwise it is set to zero.
3840
3841 PCRE_CONFIG_NEWLINE
3842
3843 The output is an integer that it set to the value of the code that is used for
3844 newline. It is either LF (10) or CR (13).
3845
3846 PCRE_CONFIG_LINK_SIZE
3847
3848 The output is an integer that contains the number of bytes used for internal
3849 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3850
3851 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3852
3853 The output is an integer that contains the threshold above which the POSIX
3854 interface uses malloc() for output vectors. See item 31 above.
3855
3856 PCRE_CONFIG_MATCH_LIMIT
3857
3858 The output is an unsigned integer that contains the default limit of the number
3859 of match() calls in a pcre_exec() execution. See 42 above.
3860
3861 44. pcretest has been upgraded by the addition of the -C option. This causes it
3862 to extract all the available output from the new pcre_config() function, and to
3863 output it. The program then exits immediately.
3864
3865 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3866 order to support additional features. One way would have been to define
3867 pcre_exec2() (for example) with extra arguments, but this would not have been
3868 extensible, and would also have required all calls to the original function to
3869 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3870 is used for passing in "extra" data from pcre_study().
3871
3872 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3873 contains the following fields:
3874
3875 flags a bitmap indicating which of the following fields are set
3876 study_data opaque data from pcre_study()
3877 match_limit a way of specifying a limit on match() calls for a specific
3878 call to pcre_exec()
3879 callout_data data for callouts (see 49 below)
3880
3881 The flag bits are also defined in pcre.h, and are
3882
3883 PCRE_EXTRA_STUDY_DATA
3884 PCRE_EXTRA_MATCH_LIMIT
3885 PCRE_EXTRA_CALLOUT_DATA
3886
3887 The pcre_study() function now returns one of these new pcre_extra blocks, with
3888 the actual study data pointed to by the study_data field, and the
3889 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3890 before. That is, this change is entirely upwards-compatible and requires no
3891 change to existing code.
3892
3893 If you want to pass in additional data to pcre_exec(), you can either place it
3894 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3895 block.
3896
3897 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3898 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3899 times with different match limits, until it finds the minimum value needed for
3900 pcre_exec() to complete. The value is then output. This can be instructive; for
3901 most simple matches the number is quite small, but for pathological cases it
3902 gets very large very quickly.
3903
3904 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3905 returns the size of the data block pointed to by the study_data field in a
3906 pcre_extra block, that is, the value that was passed as the argument to
3907 pcre_malloc() when PCRE was getting memory in which to place the information
3908 created by pcre_study(). The fourth argument should point to a size_t variable.
3909 pcretest has been extended so that this information is shown after a successful
3910 pcre_study() call when information about the compiled regex is being displayed.
3911
3912 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3913 because what follows is always an absolute path. (Later: it turns out that this
3914 is more than cosmetic for MinGW, because it doesn't like empty path
3915 components.)
3916
3917 49. Some changes have been made to the callout feature (see 28 above):
3918
3919 (i) A callout function now has three choices for what it returns:
3920
3921 0 => success, carry on matching
3922 > 0 => failure at this point, but backtrack if possible
3923 < 0 => serious error, return this value from pcre_exec()
3924
3925 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3926 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3927 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3928 use by callout functions. It will never be used by PCRE itself.
3929
3930 (ii) The pcre_extra structure (see 45 above) has a void * field called
3931 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3932 pcre_callout_block structure has a field of the same name. The contents of
3933 the field passed in the pcre_extra structure are passed to the callout
3934 function in the corresponding field in the callout block. This makes it
3935 easier to use the same callout-containing regex from multiple threads. For
3936 testing, the pcretest program has a new data escape
3937
3938 \C*n pass the number n (may be negative) as callout_data
3939
3940 If the callout function in pcretest receives a non-zero value as
3941 callout_data, it returns that value.
3942
3943 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3944 there were some redundant $(CFLAGS) in commands that are now specified as
3945 $(LINK), which already includes $(CFLAGS).
3946
3947 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3948 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3949 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3950 one-byte characters throughout. Note that case-insensitive matching applies
3951 only to characters whose values are less than 256. PCRE doesn't support the
3952 notion of cases for higher-valued characters.
3953
3954 (i) A character class whose characters are all within 0-255 is handled as
3955 a bit map, and the map is inverted for negative classes. Previously, a
3956 character > 255 always failed to match such a class; however it should
3957 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3958
3959 (ii) A negated character class with a single character < 255 is coded as
3960 "not this character" (OP_NOT). This wasn't working properly when the test
3961 character was multibyte, either singly or repeated.
3962
3963 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3964 mode, for example: \x{100}{2,3}.
3965
3966 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3967 singly or repeated) now correctly test multibyte characters. However,
3968 PCRE doesn't recognize any characters with values greater than 255 as
3969 digits, spaces, or word characters. Such characters always match \D, \S,
3970 and \W, and never match \d, \s, or \w.
3971
3972 (v) Classes may now contain characters and character ranges with values
3973 greater than 255. For example: [ab\x{100}-\x{400}].
3974
3975 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3976 PCRE in UTF-8 mode.
3977
3978 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3979 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3980 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3981 value.)
3982
3983 53. The single man page has become too large. I have therefore split it up into
3984 a number of separate man pages. These also give rise to individual HTML pages;
3985 these are now put in a separate directory, and there is an index.html page that
3986 lists them all. Some hyperlinking between the pages has been installed.
3987
3988 54. Added convenience functions for handling named capturing parentheses.
3989
3990 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3991 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3992 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3993 are faulted.
3994
3995 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3996 calling configure. These values are used when compiling the dftables.c program
3997 which is run to generate the source of the default character tables. They
3998 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3999 you will need to set these values.
4000
4001 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4002
4003
4004 Version 3.9 02-Jan-02
4005 ---------------------
4006
4007 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4008
4009 2. If --disable-static was given, the building process failed when trying to
4010 build pcretest and pcregrep. (For some reason it was using libtool to compile
4011 them, which is not right, as they aren't part of the library.)
4012
4013
4014 Version 3.8 18-Dec-01
4015 ---------------------
4016
4017 1. The experimental UTF-8 code was completely screwed up. It was packing the
4018 bytes in the wrong order. How dumb can you get?
4019
4020
4021 Version 3.7 29-Oct-01
4022 ---------------------
4023
4024 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4025 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4026 this didn't happen under Solaris 8, where I normally test things.
4027
4028 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4029 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4030 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4031 this fix an hour or so after the initial 3.7 release.)
4032
4033
4034 Version 3.6 23-Oct-01
4035 ---------------------
4036
4037 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4038 offsets passed as NULL with zero offset count.
4039
4040 2. The config.guess and config.sub files had not been updated when I moved to
4041 the latest autoconf.
4042
4043
4044 Version 3.5 15-Aug-01
4045 ---------------------
4046
4047 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4048 had been forgotten.
4049
4050 2. By using declared but undefined structures, we can avoid using "void"
4051 definitions in pcre.h while keeping the internal definitions of the structures
4052 private.
4053
4054 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4055 user point of view, this means that both static and shared libraries are built
4056 by default, but this can be individually controlled. More of the work of
4057 handling this static/shared cases is now inside libtool instead of PCRE's make
4058 file.
4059
4060 4. The pcretest utility is now installed along with pcregrep because it is
4061 useful for users (to test regexs) and by doing this, it automatically gets
4062 relinked by libtool. The documentation has been turned into a man page, so
4063 there are now .1, .txt, and .html versions in /doc.
4064
4065 5. Upgrades to pcregrep:
4066 (i) Added long-form option names like gnu grep.
4067 (ii) Added --help to list all options with an explanatory phrase.
4068 (iii) Added -r, --recursive to recurse into sub-directories.
4069 (iv) Added -f, --file to read patterns from a file.
4070
4071 6. pcre_exec() was referring to its "code" argument before testing that
4072 argument for NULL (and giving an error if it was NULL).
4073
4074 7. Upgraded Makefile.in to allow for compiling in a different directory from
4075 the source directory.
4076
4077 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4078 options bits, the pointer it was passed was to an int instead of to an unsigned
4079 long int. This mattered only on 64-bit systems.
4080
4081 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4082 generated) instead of pcre.in, which it its source. Also made the same change
4083 in several of the .c files.
4084
4085 10. A new release of gcc defines printf() as a macro, which broke pcretest
4086 because it had an ifdef in the middle of a string argument for printf(). Fixed
4087 by using separate calls to printf().
4088
4089 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4090 script, to force use of CR or LF instead of \n in the source. On non-Unix
4091 systems, the value can be set in config.h.
4092
4093 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4094 absolute limit. Changed the text of the error message to make this clear, and
4095 likewise updated the man page.
4096
4097 13. The limit of 99 on the number of capturing subpatterns has been removed.
4098 The new limit is 65535, which I hope will not be a "real" limit.
4099
4100
4101 Version 3.4 22-Aug-00
4102 ---------------------
4103
4104 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4105
4106 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4107
4108
4109 Version 3.3 01-Aug-00
4110 ---------------------
4111
4112 1. If an octal character was given, but the value was greater than \377, it
4113 was not getting masked to the least significant bits, as documented. This could
4114 lead to crashes in some systems.
4115
4116 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4117 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4118
4119 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4120 These just pass their arguments on to (pcre_free)(), but they are provided
4121 because some uses of PCRE bind it to non-C systems that can call its functions,
4122 but cannot call free() or pcre_free() directly.
4123
4124 4. Add "make test" as a synonym for "make check". Corrected some comments in
4125 the Makefile.
4126
4127 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4128 Makefile.
4129
4130 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4131 command called pgrep for grepping around the active processes.
4132
4133 7. Added the beginnings of support for UTF-8 character strings.
4134
4135 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4136 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4137 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4138 out for the ar command.)
4139
4140
4141 Version 3.2 12-May-00
4142 ---------------------
4143
4144 This is purely a bug fixing release.
4145
4146 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4147 of ZA. This was just one example of several cases that could provoke this bug,
4148 which was introduced by change 9 of version 2.00. The code for breaking
4149 infinite loops after an iteration that matches an empty string was't working
4150 correctly.
4151
4152 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4153 when matched against abbab (for example). After matching an empty string, it
4154 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4155 caused it to match further down the string than it should.
4156
4157 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4158 was there because it doesn't seem to be needed, and it causes trouble on some
4159 systems, as it is not a Standard C header. It has been removed.
4160
4161 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4162 were reported on the Macintosh. The changes were from
4163
4164 while ((c = *(++ptr)) != 0 && c != '\n');
4165 to
4166 while ((c = *(++ptr)) != 0 && c != '\n') ;
4167
4168 Totally extraordinary, but if that's what it takes...
4169
4170 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4171 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4172 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4173 assumes the way PCRE uses memmove() (always moving upwards).
4174
4175 6. PCRE is being used in one environment where strchr() is not available. There
4176 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4177 faster code anyway.
4178
4179
4180 Version 3.1 09-Feb-00
4181 ---------------------
4182
4183 The only change in this release is the fixing of some bugs in Makefile.in for
4184 the "install" target:
4185
4186 (1) It was failing to install pcreposix.h.
4187
4188 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4189
4190
4191 Version 3.0 01-Feb-00
4192 ---------------------
4193
4194 1. Add support for the /+ modifier to perltest (to output $` like it does in
4195 pcretest).
4196
4197 2. Add support for the /g modifier to perltest.
4198
4199 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4200 matches null strings.
4201
4202 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4203 pattern. Perl treats empty patterns specially - it reuses the most recent
4204 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4205 effect.
4206
4207 5. The POSIX interface was broken in that it was just handing over the POSIX
4208 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4209 required a bigger vector, with some working space on the end. This means that
4210 the POSIX wrapper now has to get and free some memory, and copy the results.
4211
4212 6. Added some simple autoconf support, placing the test data and the
4213 documentation in separate directories, re-organizing some of the
4214 information files, and making it build pcre-config (a GNU standard). Also added
4215 libtool support for building PCRE as a shared library, which is now the
4216 default.
4217
4218 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4219 09 are not valid octal constants. Single digits will be used for minor values
4220 less than 10.
4221
4222 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4223 existing programs that set these in the POSIX interface can use PCRE without
4224 modification.
4225
4226 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4227 return all that pcre_info() returns, plus additional data. The pcre_info()
4228 function is retained for compatibility, but is considered to be obsolete.
4229
4230 10. Added experimental recursion feature (?R) to handle one common case that
4231 Perl 5.6 will be able to do with (?p{...}).
4232
4233 11. Added support for POSIX character classes like [:alpha:], which Perl is
4234 adopting.
4235
4236
4237 Version 2.08 31-Aug-99
4238 ----------------------
4239
4240 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4241 trying to match at the startoffset position, but instead was moving forward to
4242 the next newline as if a previous match had failed.
4243
4244 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4245 and could get into a loop if a null string was matched other than at the start
4246 of the subject.
4247
4248 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4249 be distinguished at compile time, and for completeness also added PCRE_DATE.
4250
4251 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4252 in GnuWin32 environments.
4253
4254
4255 Version 2.07 29-Jul-99
4256 ----------------------
4257
4258 1. The documentation is now supplied in plain text form and HTML as well as in
4259 the form of man page sources.
4260
4261 2. C++ compilers don't like assigning (void *) values to other pointer types.
4262 In particular this affects malloc(). Although there is no problem in Standard
4263 C, I've put in casts to keep C++ compilers happy.
4264
4265 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4266 should be (const char *).
4267
4268 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4269 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4270 However, I haven't made this a standard facility. The documentation doesn't
4271 mention it, and the Makefile doesn't support it.
4272
4273 5. The Makefile now contains an "install" target, with editable destinations at
4274 the top of the file. The pcretest program is not installed.
4275
4276 6. pgrep -V now gives the PCRE version number and date.
4277
4278 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4279 causing the entire string to be ignored, instead of just the last character.
4280
4281 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4282 non-matching string, it can take a very, very long time, even for strings of
4283 quite modest length, because of the nested recursion. PCRE now does better in
4284 some of these cases. It does this by remembering the last required literal
4285 character in the pattern, and pre-searching the subject to ensure it is present
4286 before running the real match. In other words, it applies a heuristic to detect
4287 some types of certain failure quickly, and in the above example, if presented
4288 with a string that has no trailing " it gives "no match" very quickly.
4289
4290 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4291 other alternatives are tried instead.
4292
4293
4294 Version 2.06 09-Jun-99
4295 ----------------------
4296
4297 1. Change pcretest's output for amount of store used to show just the code
4298 space, because the remainder (the data block) varies in size between 32-bit and
4299 64-bit systems.
4300
4301 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4302 start matching at. This allows lookbehinds to work when searching for multiple
4303 occurrences in a string.
4304
4305 3. Added additional options to pcretest for testing multiple occurrences:
4306
4307 /+ outputs the rest of the string that follows a match
4308 /g loops for multiple occurrences, using the new startoffset argument
4309 /G loops for multiple occurrences by passing an incremented pointer
4310
4311 4. PCRE wasn't doing the "first character" optimization for patterns starting
4312 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4313 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4314 the letter 'x'. On long subject strings, this gives a significant speed-up.
4315
4316
4317 Version 2.05 21-Apr-99
4318 ----------------------
4319
4320 1. Changed the type of magic_number from int to long int so that it works
4321 properly on 16-bit systems.
4322
4323 2. Fixed a bug which caused patterns starting with .* not to work correctly
4324 when the subject string contained newline characters. PCRE was assuming
4325 anchoring for such patterns in all cases, which is not correct because .* will
4326 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4327 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4328 must be retried after every newline in the subject.
4329
4330
4331 Version 2.04 18-Feb-99
4332 ----------------------
4333
4334 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4335 computation of the store needed to hold the pattern was incorrect (too large).
4336 If such patterns were nested a few deep, this could multiply and become a real
4337 problem.
4338
4339 2. Added /M option to pcretest to show the memory requirement of a specific
4340 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4341
4342 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4343 compiled in such a way that the backtracking after subsequent failure was
4344 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4345 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4346
4347
4348 Version 2.03 02-Feb-99
4349 ----------------------
4350
4351 1. Fixed typo and small mistake in man page.
4352
4353 2. Added 4th condition (GPL supersedes if conflict) and created separate
4354 LICENCE file containing the conditions.
4355
4356 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4357 Perl, that is the internal \ allows the delimiter to be included in the
4358 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4359 the final delimiter, add \ to the end of the pattern (to test the error).
4360
4361 4. Added the convenience functions for extracting substrings after a successful
4362 match. Updated pcretest to make it able to test these functions.
4363
4364
4365 Version 2.02 14-Jan-99
4366 ----------------------
4367
4368 1. Initialized the working variables associated with each extraction so that
4369 their saving and restoring doesn't refer to uninitialized store.
4370
4371 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4372 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4373 fix the problem.
4374
4375 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4376 calls, and wasn't printing the correct value for compiling calls. Increased the
4377 default value of LOOPREPEAT, and the number of significant figures in the
4378 times.
4379
4380 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4381
4382 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4383 a building problem on Windows NT with a FAT file system.
4384
4385
4386 Version 2.01 21-Oct-98
4387 ----------------------
4388
4389 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4390 to character tables built by pcre_maketables() in the current locale. If NULL
4391 is passed, the default tables are used.
4392
4393
4394 Version 2.00 24-Sep-98
4395 ----------------------
4396
4397 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4398 it any more.
4399
4400 2. Allow quantification of (?>) groups, and make it work correctly.
4401
4402 3. The first character computation wasn't working for (?>) groups.
4403
4404 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4405 end of the subject) and add 5.005's \z, which really does match only at the
4406 very end of the subject.
4407
4408 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4409
4410 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4411 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4412 localized options. All options to pcre_study() were also removed.
4413
4414 7. Add other new features from 5.005:
4415
4416 $(?<= positive lookbehind
4417 $(?<! negative lookbehind
4418 (?imsx-imsx) added the unsetting capability
4419 such a setting is global if at outer level; local otherwise
4420 (?imsx-imsx:) non-capturing groups with option setting
4421 (?(cond)re|re) conditional pattern matching
4422
4423 A backreference to itself in a repeated group matches the previous
4424 captured string.
4425
4426 8. General tidying up of studying (both automatic and via "study")
4427 consequential on the addition of new assertions.
4428
4429 9. As in 5.005, unlimited repeated groups that could match an empty substring
4430 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4431 runtime if any iteration does actually match an empty substring.
4432
4433 10. Include the RunTest script in the distribution.
4434
4435 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4436 discrepancies, some of which were old and were also with respect to 5.004. They
4437 have now been fixed.
4438
4439
4440 Version 1.09 28-Apr-98
4441 ----------------------
4442
4443 1. A negated single character class followed by a quantifier with a minimum
4444 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4445 program crashes, or just wrong answers. This did not apply to negated classes
4446 containing more than one character, or to minima other than one.
4447
4448
4449 Version 1.08 27-Mar-98
4450 ----------------------
4451
4452 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4453
4454 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4455 latter must appear before anything that relies on it in the pattern.
4456
4457
4458 Version 1.07 16-Feb-98
4459 ----------------------
4460
4461 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4462 repeat of a potentially empty string).
4463
4464
4465 Version 1.06 23-Jan-98
4466 ----------------------
4467
4468 1. Added Markus Oberhumer's little patches for C++.
4469
4470 2. Literal strings longer than 255 characters were broken.
4471
4472
4473 Version 1.05 23-Dec-97
4474 ----------------------
4475
4476 1. Negated character classes containing more than one character were failing if
4477 PCRE_CASELESS was set at run time.
4478
4479
4480 Version 1.04 19-Dec-97
4481 ----------------------
4482
4483 1. Corrected the man page, where some "const" qualifiers had been omitted.
4484
4485 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4486 input syntax.
4487
4488 3. Fixed memory leak which occurred when a regex with back references was
4489 matched with an offsets vector that wasn't big enough. The temporary memory
4490 that is used in this case wasn't being freed if the match failed.
4491
4492 4. Tidied pcretest to ensure it frees memory that it gets.
4493
4494 5. Temporary memory was being obtained in the case where the passed offsets
4495 vector was exactly big enough.
4496
4497 6. Corrected definition of offsetof() from change 5 below.
4498
4499 7. I had screwed up change 6 below and broken the rules for the use of
4500 setjmp(). Now fixed.
4501
4502
4503 Version 1.03 18-Dec-97
4504 ----------------------
4505
4506 1. A erroneous regex with a missing opening parenthesis was correctly
4507 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4508 on some systems.
4509
4510 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4511 it was reported that one broken compiler failed on the former because "code" is
4512 also an independent variable.
4513
4514 3. The erroneous regex a[]b caused an array overrun reference.
4515
4516 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4517 fail on data ending with that character. (It was going on too far, and checking
4518 the next character, typically a binary zero.) This was specific to the
4519 optimized code for single-character negative classes.
4520
4521 5. Added a contributed patch from the TIN world which does the following:
4522
4523 + Add an undef for memmove, in case the the system defines a macro for it.
4524
4525 + Add a definition of offsetof(), in case there isn't one. (I don't know
4526 the reason behind this - offsetof() is part of the ANSI standard - but
4527 it does no harm).
4528
4529 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4530 most of the places where whitespace preceded '#'. I have given up and
4531 allowed the remaining 2 cases to be at the margin.
4532
4533 + Rename some variables in pcre to eliminate shadowing. This seems very
4534 pedantic, but does no harm, of course.
4535
4536 6. Moved the call to setjmp() into its own function, to get rid of warnings
4537 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4538
4539 7. Constructs such as \d{8,} were compiling into the equivalent of
4540 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4541 outcome, but in this particular case used more store than had been allocated,
4542 which caused the bug to be discovered because it threw up an internal error.
4543
4544 8. The debugging code in both pcre and pcretest for outputting the compiled
4545 form of a regex was going wrong in the case of back references followed by
4546 curly-bracketed repeats.
4547
4548
4549 Version 1.02 12-Dec-97
4550 ----------------------
4551
4552 1. Typos in pcre.3 and comments in the source fixed.
4553
4554 2. Applied a contributed patch to get rid of places where it used to remove
4555 'const' from variables, and fixed some signed/unsigned and uninitialized
4556 variable warnings.
4557
4558 3. Added the "runtest" target to Makefile.
4559
4560 4. Set default compiler flag to -O2 rather than just -O.
4561
4562
4563 Version 1.01 19-Nov-97
4564 ----------------------
4565
4566 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4567 like /([ab]*)*/, that is, for classes with more than one character in them.
4568
4569 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4570 as /((?>a*))*/ (a PCRE_EXTRA facility).
4571
4572
4573 Version 1.00 18-Nov-97
4574 ----------------------
4575
4576 1. Added compile-time macros to support systems such as SunOS4 which don't have
4577 memmove() or strerror() but have other things that can be used instead.
4578
4579 2. Arranged that "make clean" removes the executables.
4580
4581
4582 Version 0.99 27-Oct-97
4583 ----------------------
4584
4585 1. Fixed bug in code for optimizing classes with only one character. It was
4586 initializing a 32-byte map regardless, which could cause it to run off the end
4587 of the memory it had got.
4588
4589 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4590
4591
4592 Version 0.98 22-Oct-97
4593 ----------------------
4594
4595 1. Fixed bug in code for handling temporary memory usage when there are more
4596 back references than supplied space in the ovector. This could cause segfaults.
4597
4598
4599 Version 0.97 21-Oct-97
4600 ----------------------
4601
4602 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4603
4604 2. Optimized negated single characters not to use a bit map.
4605
4606 3. Brought error texts together as macro definitions; clarified some of them;
4607 fixed one that was wrong - it said "range out of order" when it meant "invalid
4608 escape sequence".
4609
4610 4. Changed some char * arguments to const char *.
4611
4612 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4613
4614 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4615 pcretest.
4616
4617
4618 Version 0.96 16-Oct-97
4619 ----------------------
4620
4621 1. Added a simple "pgrep" utility to the distribution.
4622
4623 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4624 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4625 where "ddd" means "one or more decimal digits".
4626
4627 3. Fixed serious bug. If a pattern had a back reference, but the call to
4628 pcre_exec() didn't supply a large enough ovector to record the related
4629 identifying subpattern, the match always failed. PCRE now remembers the number
4630 of the largest back reference, and gets some temporary memory in which to save
4631 the offsets during matching if necessary, in order to ensure that
4632 backreferences always work.
4633
4634 4. Increased the compatibility with Perl in a number of ways:
4635
4636 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4637 to request this handling. The option can be set at compile or exec time.
4638
4639 (b) $ matches before a terminating newline by default; an option
4640 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4641 mode). The option can be set at compile or exec time.
4642
4643 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4644 the same as Perl's. If the decimal number it represents is less than 10
4645 or there aren't that many previous left capturing parentheses, an octal
4646 escape is read. Inside a character class, it's always an octal escape,
4647 even if it is a single digit.
4648
4649 (d) An escaped but undefined alphabetic character is taken as a literal,
4650 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4651 escapes.
4652
4653 (e) {0} is now permitted. (The previous item is removed from the compiled
4654 pattern).
4655
4656 5. Changed all the names of code files so that the basic parts are no longer
4657 than 10 characters, and abolished the teeny "globals.c" file.
4658
4659 6. Changed the handling of character classes; they are now done with a 32-byte
4660 bit map always.
4661
4662 7. Added the -d and /D options to pcretest to make it possible to look at the
4663 internals of compilation without having to recompile pcre.
4664
4665
4666 Version 0.95 23-Sep-97
4667 ----------------------
4668
4669 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4670 \x20 at the start of a run of normal characters. These were being treated as
4671 real characters, instead of the source characters being re-checked.
4672
4673
4674 Version 0.94 18-Sep-97
4675 ----------------------
4676
4677 1. The functions are now thread-safe, with the caveat that the global variables
4678 containing pointers to malloc() and free() or alternative functions are the
4679 same for all threads.
4680
4681 2. Get pcre_study() to generate a bitmap of initial characters for non-
4682 anchored patterns when this is possible, and use it if passed to pcre_exec().
4683
4684
4685 Version 0.93 15-Sep-97
4686 ----------------------
4687
4688 1. /(b)|(:+)/ was computing an incorrect first character.
4689
4690 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4691 but not actually doing anything yet.
4692
4693 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4694 as Perl does (e.g. [-az] or [az-]).
4695
4696 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4697 all possible positions.
4698
4699 5. Split up into different modules to avoid including unneeded functions in a
4700 compiled binary. However, compile and exec are still in one module. The "study"
4701 function is split off.
4702
4703 6. The character tables are now in a separate module whose source is generated
4704 by an auxiliary program - but can then be edited by hand if required. There are
4705 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4706 toupper() in the code.
4707
4708 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4709 make them global. Abolish the function for setting them, as the caller can now
4710 set them directly.
4711
4712
4713 Version 0.92 11-Sep-97
4714 ----------------------
4715
4716 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4717 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4718
4719 2. Caseless matching was not working in character classes if the characters in
4720 the pattern were in upper case.
4721
4722 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4723
4724 4. Make PCRE_ANCHORED public and accept as a compile option.
4725
4726 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4727 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4728 pass them.
4729
4730 6. Give an error if bad option bits passed at compile or run time.
4731
4732 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4733 pcretest to cause it to pass that flag.
4734
4735 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4736 options, and the first character, if set.
4737
4738 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4739
4740
4741 Version 0.91 10-Sep-97
4742 ----------------------
4743
4744 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4745 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4746
4747 2. PCRE was looping on encountering an indefinitely repeated back reference to
4748 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4749 Perl does - treats the match as successful.
4750
4751 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5