ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 966 - (show annotations)
Mon May 14 06:27:21 2012 UTC (8 years, 1 month ago) by zherczeg
File size: 230766 byte(s)
JIT compiler update
1 ChangeLog for PCRE
2 ------------------
4 Version 8.31
5 -----------------------------
7 1. Fixing a wrong JIT test case and some compiler warnings.
9 2. Removed a bashism from the RunTest script.
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)".
15 4. Partial matching support is added to the JIT compiler.
17 5. Fixed several bugs concerned with partial matching of items that consist
18 of more than one character:
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.
24 (b) \R did not give a hard partial match if \r was found at the end of the
25 subject.
27 (c) \X did not give a hard partial match after matching one or more
28 characters at the end of the subject.
30 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
31 a partial match for the string "\r".
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.
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.
40 7. Individual JIT compile options can be set in pcretest by following -s+[+]
41 or /S+[+] with a digit between 1 and 7.
43 8. OP_NOT now supports any UTF character not just single-byte ones.
45 9. (*MARK) control verb is now supported by the JIT compiler.
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.)
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.
58 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
59 call.
61 14. As documented, (*COMMIT) is now confined to within a positive assertion.
63 15. It is now possible to link pcretest with libedit as an alternative to
64 libreadline.
66 16. (*COMMIT) control verb is now supported by the JIT compiler.
68 17. The Unicode data tables have been updated to Unicode 6.1.0.
70 18. Added --file-list option to pcregrep.
72 19. Added binary file support to pcregrep, including the -a, --binary-files,
73 -I, and --text options.
75 20. The madvise function is renamed for posix_madvise for QNX compatibility
76 reasons. Fixed by Giuseppe D'Angelo.
78 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
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.
83 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
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.
90 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
91 was not doing so when it should - probably a typo introduced by SVN 528
92 (change 8.10/14).
94 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
95 \w+ when the character tables indicated that \x{c4} was a word character.
96 There were several related cases, all because the tests for doing a table
97 lookup were testing for characters less than 127 instead of 255.
99 27. If a pattern contains capturing parentheses that are not used in a match,
100 their slots in the ovector are set to -1. For those that are higher than
101 any matched groups, this happens at the end of processing. In the case when
102 there were back references that the ovector was too small to contain
103 (causing temporary malloc'd memory to be used during matching), and the
104 highest capturing number was not used, memory off the end of the ovector
105 was incorrectly being set to -1. (It was using the size of the temporary
106 memory instead of the true size.)
108 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
109 ovector size, it uses memory at the end of the block that it has got.
111 29. Check for an overlong MARK name and give an error at compile time. The
112 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
114 30. JIT compiler update.
117 Version 8.30 04-February-2012
118 -----------------------------
120 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
121 name is defined in ctype.h.
123 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
124 only in quite long subpatterns.
126 3. Removed the function pcre_info(), which has been obsolete and deprecated
127 since it was replaced by pcre_fullinfo() in February 2000.
129 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
130 match a (*MARK), and the match failed at the start of the subject, a
131 reference to memory before the start of the subject could occur. This bug
132 was introduced by fix 17 of release 8.21.
134 5. A reference to an unset group with zero minimum repetition was giving
135 totally wrong answers (in non-JavaScript-compatibility mode). For example,
136 /(another)?(\1?)test/ matched against "hello world test". This bug was
137 introduced in release 8.13.
139 6. Add support for 16-bit character strings (a large amount of work involving
140 many changes and refactorings).
142 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
143 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
144 from a file.
146 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
147 rounding is not applied in this particular case).
149 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
150 if they appear, or are escaped, in patterns.
152 10. Get rid of a number of -Wunused-but-set-variable warnings.
154 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
155 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
156 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
157 also returns the mark "x". This bug applied to capturing parentheses,
158 non-capturing parentheses, and atomic parentheses. It also applied to some
159 assertions.
161 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
162 information out of configure.ac instead of relying on pcre.h.generic, which
163 is not stored in the repository.
165 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
166 -lreadline.
168 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
170 15. Applied Graycode's patch to put the top-level frame on the stack rather
171 than the heap when not using the stack for recursion. This gives a
172 performance improvement in many cases when recursion is not deep.
174 16. Experimental code added to "pcretest -C" to output the stack frame size.
177 Version 8.21 12-Dec-2011
178 ------------------------
180 1. Updating the JIT compiler.
182 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
183 are added as well.
185 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
186 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
187 calling _pcre_jit_exec. Some extra comments are added.
189 4. (*MARK) settings inside atomic groups that do not contain any capturing
190 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
191 was introduced by change 18 for 8.20.
193 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
194 ECMA-262 standard.
196 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
197 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
198 This bug was probably introduced by change 9 of 8.13.
200 7. While fixing 6 above, I noticed that a number of other items were being
201 incorrectly rejected as "not fixed length". This arose partly because newer
202 opcodes had not been added to the fixed-length checking code. I have (a)
203 corrected the bug and added tests for these items, and (b) arranged for an
204 error to occur if an unknown opcode is encountered while checking for fixed
205 length instead of just assuming "not fixed length". The items that were
206 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
207 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
208 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
210 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
211 being incorrectly compiled and would have given unpredicatble results.
213 9. A possessively repeated subpattern with minimum repeat count greater than
214 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
215 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
216 the first (A) could occur when it should not.
218 10. Add a cast and remove a redundant test from the code.
220 11. JIT should use pcre_malloc/pcre_free for allocation.
222 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
223 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
224 is anything other than /usr, -L is still shown).
226 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
228 14. Perl does not support \N without a following name in a [] class; PCRE now
229 also gives an error.
231 15. If a forward reference was repeated with an upper limit of around 2000,
232 it caused the error "internal error: overran compiling workspace". The
233 maximum number of forward references (including repeats) was limited by the
234 internal workspace, and dependent on the LINK_SIZE. The code has been
235 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
236 the default depends on LINK_SIZE. There is a new upper limit (for safety)
237 of around 200,000 forward references. While doing this, I also speeded up
238 the filling in of repeated forward references.
240 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
241 incorrectly expecting the subject to contain another "a" after the start.
243 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
244 in the match, the SKIP should be ignored. This was not happening; instead
245 the SKIP was being treated as NOMATCH. For patterns such as
246 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
247 tested.
249 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
250 now much more compatible with Perl, in particular in cases where the result
251 is a non-match for a non-anchored pattern. For example, if
252 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
253 "m", where previously it did not return a name. A side effect of this
254 change is that for partial matches, the last encountered mark name is
255 returned, as for non matches. A number of tests that were previously not
256 Perl-compatible have been moved into the Perl-compatible test files. The
257 refactoring has had the pleasing side effect of removing one argument from
258 the match() function, thus reducing its stack requirements.
260 19. If the /S+ option was used in pcretest to study a pattern using JIT,
261 subsequent uses of /S (without +) incorrectly behaved like /S+.
263 21. Retrieve executable code size support for the JIT compiler and fixing
264 some warnings.
266 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
267 not work when the shorter character appeared right at the end of the
268 subject string.
270 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
271 systems.
273 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
274 output it when the /M option is used in pcretest.
276 25. The CheckMan script was not being included in the distribution. Also, added
277 an explicit "perl" to run Perl scripts from the PrepareRelease script
278 because this is reportedly needed in Windows.
280 26. If study data was being save in a file and studying had not found a set of
281 "starts with" bytes for the pattern, the data written to the file (though
282 never used) was taken from uninitialized memory and so caused valgrind to
283 complain.
285 27. Updated RunTest.bat as provided by Sheri Pierce.
287 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
289 29. Computation of memory usage for the table of capturing group names was
290 giving an unnecessarily large value.
293 Version 8.20 21-Oct-2011
294 ------------------------
296 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
297 a POSIX class. After further experiments with Perl, which convinced me that
298 Perl has bugs and confusions, a closing square bracket is no longer allowed
299 in a POSIX name. This bug also affected patterns with classes that started
300 with full stops.
302 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
303 captured substring, but while checking the failing first alternative,
304 substring 1 is temporarily captured. If the output vector supplied to
305 pcre_exec() was not big enough for this capture, the yield of the function
306 was still zero ("insufficient space for captured substrings"). This cannot
307 be totally fixed without adding another stack variable, which seems a lot
308 of expense for a edge case. However, I have improved the situation in cases
309 such as /(a)(b)x|abc/ matched against "abc", where the return code
310 indicates that fewer than the maximum number of slots in the ovector have
311 been set.
313 3. Related to (2) above: when there are more back references in a pattern than
314 slots in the output vector, pcre_exec() uses temporary memory during
315 matching, and copies in the captures as far as possible afterwards. It was
316 using the entire output vector, but this conflicts with the specification
317 that only 2/3 is used for passing back captured substrings. Now it uses
318 only the first 2/3, for compatibility. This is, of course, another edge
319 case.
321 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
322 main code base, and can be used by building with --enable-jit. When this is
323 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
324 runtime --no-jit option is given.
326 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
327 ovector, the return from the function was zero, implying that there were
328 other matches that did not fit. The correct "exactly full" value is now
329 returned.
331 6. If a subpattern that was called recursively or as a subroutine contained
332 (*PRUNE) or any other control that caused it to give a non-standard return,
333 invalid errors such as "Error -26 (nested recursion at the same subject
334 position)" or even infinite loops could occur.
336 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
337 computing the minimum length on reaching *ACCEPT, and so ended up with the
338 wrong value of 1 rather than 0. Further investigation indicates that
339 computing a minimum subject length in the presence of *ACCEPT is difficult
340 (think back references, subroutine calls), and so I have changed the code
341 so that no minimum is registered for a pattern that contains *ACCEPT.
343 8. If (*THEN) was present in the first (true) branch of a conditional group,
344 it was not handled as intended. [But see 16 below.]
346 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
347 Sheri Pierce.
349 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
350 the first byte in a match must be "a".
352 11. Change 17 for 8.13 increased the recursion depth for patterns like
353 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
354 pattern contains any instances of (*THEN). If it does not, the old
355 optimizations are restored. It would be nice to do this on a per-group
356 basis, but at the moment that is not feasible.
358 12. In some environments, the output of pcretest -C is CRLF terminated. This
359 broke RunTest's code that checks for the link size. A single white space
360 character after the value is now allowed for.
362 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
363 For "fr", it uses the Windows-specific input and output files.
365 14. If (*THEN) appeared in a group that was called recursively or as a
366 subroutine, it did not work as intended. [But see next item.]
368 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
369 pattern fragments (but not containing any | characters). If A and B are
370 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
371 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
372 D. In other words, Perl considers parentheses that do not contain any |
373 characters to be part of a surrounding alternative, whereas PCRE was
374 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
375 differently. PCRE now behaves in the same way as Perl, except in the case
376 of subroutine/recursion calls such as (?1) which have in any case always
377 been different (but PCRE had them first :-).
379 16. Related to 15 above: Perl does not treat the | in a conditional group as
380 creating alternatives. Such a group is treated in the same way as an
381 ordinary group without any | characters when processing (*THEN). PCRE has
382 been changed to match Perl's behaviour.
384 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
385 RunGrepTest script failed.
387 18. Change 22 for version 13 caused atomic groups to use more stack. This is
388 inevitable for groups that contain captures, but it can lead to a lot of
389 stack use in large patterns. The old behaviour has been restored for atomic
390 groups that do not contain any capturing parentheses.
392 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
393 suppress the check for a minimum subject length at run time. (If it was
394 given to pcre_exec() or pcre_dfa_exec() it did work.)
396 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
397 fail to work when decoding hex characters in data strings in EBCDIC
398 environments.
400 21. It appears that in at least one Mac OS environment, the isxdigit() function
401 is implemented as a macro that evaluates to its argument more than once,
402 contravening the C 90 Standard (I haven't checked a later standard). There
403 was an instance in pcretest which caused it to go wrong when processing
404 \x{...} escapes in subject strings. The has been rewritten to avoid using
405 things like p++ in the argument of isxdigit().
408 Version 8.13 16-Aug-2011
409 ------------------------
411 1. The Unicode data tables have been updated to Unicode 6.0.0.
413 2. Two minor typos in pcre_internal.h have been fixed.
415 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
416 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
417 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
419 4. There were a number of related bugs in the code for matching backrefences
420 caselessly in UTF-8 mode when codes for the characters concerned were
421 different numbers of bytes. For example, U+023A and U+2C65 are an upper
422 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
423 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
424 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
425 2-byte code at the end of the subject (it thought there wasn't enough data
426 left).
428 5. Comprehensive information about what went wrong is now returned by
429 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
430 as the output vector has at least 2 elements. The offset of the start of
431 the failing character and a reason code are placed in the vector.
433 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
434 now returned is for the first byte of the failing character, instead of the
435 last byte inspected. This is an incompatible change, but I hope it is small
436 enough not to be a problem. It makes the returned offset consistent with
437 pcre_exec() and pcre_dfa_exec().
439 7. pcretest now gives a text phrase as well as the error number when
440 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
441 failure, the offset and reason code are output.
443 8. When \R was used with a maximizing quantifier it failed to skip backwards
444 over a \r\n pair if the subsequent match failed. Instead, it just skipped
445 back over a single character (\n). This seems wrong (because it treated the
446 two characters as a single entity when going forwards), conflicts with the
447 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
448 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
449 has been changed.
451 9. Some internal refactoring has changed the processing so that the handling
452 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
453 time (the PCRE_DOTALL option was changed this way some time ago: version
454 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
455 which was always a bit of a fudge. It also means that there is one less
456 argument for the match() function, which reduces its stack requirements
457 slightly. This change also fixes an incompatibility with Perl: the pattern
458 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
460 10. More internal refactoring has drastically reduced the number of recursive
461 calls to match() for possessively repeated groups such as (abc)++ when
462 using pcre_exec().
464 11. While implementing 10, a number of bugs in the handling of groups were
465 discovered and fixed:
467 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
468 (a|)*(?1) gave a compile-time internal error.
469 ((a|)+)+ did not notice that the outer group could match an empty string.
470 (^a|^)+ was not marked as anchored.
471 (.*a|.*)+ was not marked as matching at start or after a newline.
473 12. Yet more internal refactoring has removed another argument from the match()
474 function. Special calls to this function are now indicated by setting a
475 value in a variable in the "match data" data block.
477 13. Be more explicit in pcre_study() instead of relying on "default" for
478 opcodes that mean there is no starting character; this means that when new
479 ones are added and accidentally left out of pcre_study(), testing should
480 pick them up.
482 14. The -s option of pcretest has been documented for ages as being an old
483 synonym of -m (show memory usage). I have changed it to mean "force study
484 for every regex", that is, assume /S for every regex. This is similar to -i
485 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
486 using it. It makes it easier to run collections of tests with and without
487 study enabled, and thereby test pcre_study() more easily. All the standard
488 tests are now run with and without -s (but some patterns can be marked as
489 "never study" - see 20 below).
491 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
492 restoration of the capturing data to the outer values was not happening
493 correctly.
495 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
496 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
497 pattern had matched an empty string, and so incorrectly returned a no
498 match.
500 17. There was optimizing code for the last branch of non-capturing parentheses,
501 and also for the obeyed branch of a conditional subexpression, which used
502 tail recursion to cut down on stack usage. Unfortunately, now that there is
503 the possibility of (*THEN) occurring in these branches, tail recursion is
504 no longer possible because the return has to be checked for (*THEN). These
505 two optimizations have therefore been removed. [But see 8.20/11 above.]
507 18. If a pattern containing \R was studied, it was assumed that \R always
508 matched two bytes, thus causing the minimum subject length to be
509 incorrectly computed because \R can also match just one byte.
511 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
512 was incorrectly computed.
514 20. If /S is present twice on a test pattern in pcretest input, it now
515 *disables* studying, thereby overriding the use of -s on the command line
516 (see 14 above). This is necessary for one or two tests to keep the output
517 identical in both cases.
519 21. When (*ACCEPT) was used in an assertion that matched an empty string and
520 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
522 22. When an atomic group that contained a capturing parenthesis was
523 successfully matched, but the branch in which it appeared failed, the
524 capturing was not being forgotten if a higher numbered group was later
525 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
526 group 1 to "a", when in fact it should be unset. This applied to multi-
527 branched capturing and non-capturing groups, repeated or not, and also to
528 positive assertions (capturing in negative assertions does not happen
529 in PCRE) and also to nested atomic groups.
531 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
532 subject after a captured substring, to make it easier to tell which of a
533 number of identical substrings has been captured.
535 24. The way atomic groups are processed by pcre_exec() has been changed so that
536 if they are repeated, backtracking one repetition now resets captured
537 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
538 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
539 "aaa". Previously, it would have been "a". As part of this code
540 refactoring, the way recursive calls are handled has also been changed.
542 25. If an assertion condition captured any substrings, they were not passed
543 back unless some other capturing happened later. For example, if
544 (?(?=(a))a) was matched against "a", no capturing was returned.
546 26. When studying a pattern that contained subroutine calls or assertions,
547 the code for finding the minimum length of a possible match was handling
548 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
549 group 1 called group 2 while simultaneously a separate group 2 called group
550 1). A stack overflow occurred in this case. I have fixed this by limiting
551 the recursion depth to 10.
553 27. Updated RunTest.bat in the distribution to the version supplied by Tom
554 Fortmann. This supports explicit test numbers on the command line, and has
555 argument validation and error reporting.
557 28. An instance of \X with an unlimited repeat could fail if at any point the
558 first character it looked at was a mark character.
560 29. Some minor code refactoring concerning Unicode properties and scripts
561 should reduce the stack requirement of match() slightly.
563 30. Added the '=' option to pcretest to check the setting of unused capturing
564 slots at the end of the pattern, which are documented as being -1, but are
565 not included in the return count.
567 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
568 compiled something random. Now it gives a compile-time error (as does
569 Perl).
571 32. A *MARK encountered during the processing of a positive assertion is now
572 recorded and passed back (compatible with Perl).
574 33. If --only-matching or --colour was set on a pcregrep call whose pattern
575 had alternative anchored branches, the search for a second match in a line
576 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
577 matched the line "0102" twice. The same bug affected patterns that started
578 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
579 twice.
581 34. Previously, PCRE did not allow quantification of assertions. However, Perl
582 does, and because of capturing effects, quantifying parenthesized
583 assertions may at times be useful. Quantifiers are now allowed for
584 parenthesized assertions.
586 35. A minor code tidy in pcre_compile() when checking options for \R usage.
588 36. \g was being checked for fancy things in a character class, when it should
589 just be a literal "g".
591 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
592 appearance of a nested POSIX class supersedes an apparent external class.
593 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
594 unescaped square brackets may also appear as part of class names. For
595 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
596 more like Perl. (But see 8.20/1 above.)
598 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
599 was because it thought it was \N{name}, which is not supported).
601 39. Add minix to OS list not supporting the -S option in pcretest.
603 40. PCRE tries to detect cases of infinite recursion at compile time, but it
604 cannot analyze patterns in sufficient detail to catch mutual recursions
605 such as ((?1))((?2)). There is now a runtime test that gives an error if a
606 subgroup is called recursively as a subpattern for a second time at the
607 same position in the subject string. In previous releases this might have
608 been caught by the recursion limit, or it might have run out of stack.
610 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
611 happen only once. PCRE was, however incorrectly giving a compile time error
612 "recursive call could loop indefinitely" because it cannot analyze the
613 pattern in sufficient detail. The compile time test no longer happens when
614 PCRE is compiling a conditional subpattern, but actual runaway loops are
615 now caught at runtime (see 40 above).
617 42. It seems that Perl allows any characters other than a closing parenthesis
618 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
619 has been changed to be the same.
621 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
622 as not to get warnings when autogen.sh is called. Also changed
623 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
625 44. To help people who use pcregrep to scan files containing exceedingly long
626 lines, the following changes have been made:
628 (a) The default value of the buffer size parameter has been increased from
629 8K to 20K. (The actual buffer used is three times this size.)
631 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
632 PCRE is built.
634 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
635 to be set at run time.
637 (d) Numerical values in pcregrep options can be followed by K or M, for
638 example --buffer-size=50K.
640 (e) If a line being scanned overflows pcregrep's buffer, an error is now
641 given and the return code is set to 2.
643 45. Add a pointer to the latest mark to the callout data block.
645 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
646 partial match of an empty string instead of no match. This was specific to
647 the use of ".".
649 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
650 complete match instead of a partial match. This bug was dependent on both
651 the PCRE_UTF8 and PCRE_DOTALL options being set.
653 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
654 starting byte set, because \b was not being ignored.
657 Version 8.12 15-Jan-2011
658 ------------------------
660 1. Fixed some typos in the markup of the man pages, and wrote a script that
661 checks for such things as part of the documentation building process.
663 2. On a big-endian 64-bit system, pcregrep did not correctly process the
664 --match-limit and --recursion-limit options (added for 8.11). In
665 particular, this made one of the standard tests fail. (The integer value
666 went into the wrong half of a long int.)
668 3. If the --colour option was given to pcregrep with -v (invert match), it
669 did strange things, either producing crazy output, or crashing. It should,
670 of course, ignore a request for colour when reporting lines that do not
671 match.
673 4. Another pcregrep bug caused similar problems if --colour was specified with
674 -M (multiline) and the pattern match finished with a line ending.
676 5. In pcregrep, when a pattern that ended with a literal newline sequence was
677 matched in multiline mode, the following line was shown as part of the
678 match. This seems wrong, so I have changed it.
680 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
681 the check for further matches in the same line (so they could be coloured)
682 to overrun the end of the current line. If another match was found, it was
683 incorrectly shown (and then shown again when found in the next line).
685 7. If pcregrep was compiled under Windows, there was a reference to the
686 function pcregrep_exit() before it was defined. I am assuming this was
687 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
688 reported by a user. I've moved the definition above the reference.
691 Version 8.11 10-Dec-2010
692 ------------------------
694 1. (*THEN) was not working properly if there were untried alternatives prior
695 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
696 backtracked to try for "b" instead of moving to the next alternative branch
697 at the same level (in this case, to look for "c"). The Perl documentation
698 is clear that when (*THEN) is backtracked onto, it goes to the "next
699 alternative in the innermost enclosing group".
701 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
702 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
703 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
704 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
705 (*THEN).
707 3. If \s appeared in a character class, it removed the VT character from
708 the class, even if it had been included by some previous item, for example
709 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
710 of \s, but is part of the POSIX "space" class.)
712 4. A partial match never returns an empty string (because you can always
713 match an empty string at the end of the subject); however the checking for
714 an empty string was starting at the "start of match" point. This has been
715 changed to the "earliest inspected character" point, because the returned
716 data for a partial match starts at this character. This means that, for
717 example, /(?<=abc)def/ gives a partial match for the subject "abc"
718 (previously it gave "no match").
720 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
721 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
722 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
723 has an implication that the given string is incomplete (because a partial
724 match is preferred over a full match). For this reason, these items now
725 give a partial match in this situation. [Aside: previously, the one case
726 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
727 match rather than a full match, which was wrong by the old rules, but is
728 now correct.]
730 6. There was a bug in the handling of #-introduced comments, recognized when
731 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
732 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
733 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
734 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
735 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
736 places in pcre_compile().
738 7. Related to (6) above, when pcre_compile() was skipping #-introduced
739 comments when looking ahead for named forward references to subpatterns,
740 the only newline sequence it recognized was NL. It now handles newlines
741 according to the set newline convention.
743 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
744 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
745 cater for a lack of strerror(). These oversights have been fixed.
747 9. Added --match-limit and --recursion-limit to pcregrep.
749 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
751 11. When the -o option was used, pcregrep was setting a return code of 1, even
752 when matches were found, and --line-buffered was not being honoured.
754 12. Added an optional parentheses number to the -o and --only-matching options
755 of pcregrep.
757 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
758 can match an empty string. The code to do it in pcretest and pcredemo
759 needed fixing:
761 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
762 only one byte after an empty string match just before CRLF (this case
763 just got forgotten; "any" and "anycrlf" were OK).
765 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
766 mode when an empty string match preceded an ASCII character followed by
767 a non-ASCII character. (The code for advancing by one character rather
768 than one byte was nonsense.)
770 (c) The pcredemo.c sample program did not have any code at all to handle
771 the cases when CRLF is a valid newline sequence.
773 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
774 as a starting offset was within the subject string. There is now a new
775 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
776 negative or greater than the length of the string. In order to test this,
777 pcretest is extended to allow the setting of negative starting offsets.
779 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
780 starting offset points to the beginning of a UTF-8 character was
781 unnecessarily clumsy. I tidied it up.
783 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
784 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
786 17. Nobody had reported that the --include_dir option, which was added in
787 release 7.7 should have been called --include-dir (hyphen, not underscore)
788 for compatibility with GNU grep. I have changed it to --include-dir, but
789 left --include_dir as an undocumented synonym, and the same for
790 --exclude-dir, though that is not available in GNU grep, at least as of
791 release 2.5.4.
793 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
794 characters from a string of bytes) have been redefined so as not to use
795 loops, in order to improve performance in some environments. At the same
796 time, I abstracted some of the common code into auxiliary macros to save
797 repetition (this should not affect the compiled code).
799 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
800 compile-time error is now given if \c is not followed by an ASCII
801 character, that is, a byte less than 128. (In EBCDIC mode, the code is
802 different, and any byte value is allowed.)
804 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
805 START_OPTIMIZE option, which is now allowed at compile time - but just
806 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
807 to pcregrep and other applications that have no direct access to PCRE
808 options. The new /Y option in pcretest sets this option when calling
809 pcre_compile().
811 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
812 back references. Groups containing recursive back references were forced to
813 be atomic by that change, but in the case of named groups, the amount of
814 memory required was incorrectly computed, leading to "Failed: internal
815 error: code overflow". This has been fixed.
817 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
818 pcretest.c, to avoid build problems in some Borland environments.
821 Version 8.10 25-Jun-2010
822 ------------------------
824 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
825 THEN.
827 2. (*ACCEPT) was not working when inside an atomic group.
829 3. Inside a character class, \B is treated as a literal by default, but
830 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
831 causes the error). The code is unchanged, but I tidied the documentation.
833 4. Inside a character class, PCRE always treated \R and \X as literals,
834 whereas Perl faults them if its -w option is set. I have changed PCRE so
835 that it faults them when PCRE_EXTRA is set.
837 5. Added support for \N, which always matches any character other than
838 newline. (It is the same as "." when PCRE_DOTALL is not set.)
840 6. When compiling pcregrep with newer versions of gcc which may have
841 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
842 declared with attribute warn_unused_result" were given. Just casting the
843 result to (void) does not stop the warnings; a more elaborate fudge is
844 needed. I've used a macro to implement this.
846 7. Minor change to pcretest.c to avoid a compiler warning.
848 8. Added four artifical Unicode properties to help with an option to make
849 \s etc use properties (see next item). The new properties are: Xan
850 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
852 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
853 use Unicode properties. (*UCP) at the start of a pattern can be used to set
854 this option. Modified pcretest to add /W to test this facility. Added
855 REG_UCP to make it available via the POSIX interface.
857 10. Added --line-buffered to pcregrep.
859 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
860 studied, and the match started with a letter with a code point greater than
861 127 whose first byte was different to the first byte of the other case of
862 the letter, the other case of this starting letter was not recognized
863 (#976).
865 12. If a pattern that was studied started with a repeated Unicode property
866 test, for example, \p{Nd}+, there was the theoretical possibility of
867 setting up an incorrect bitmap of starting bytes, but fortunately it could
868 not have actually happened in practice until change 8 above was made (it
869 added property types that matched character-matching opcodes).
871 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
872 possible starting bytes for non-anchored patterns.
874 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
875 \R, and also a number of cases that involve Unicode properties, both
876 explicit and implicit when PCRE_UCP is set.
878 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
879 input, it could crash or give wrong results if characters with values
880 greater than 0xc0 were present in the subject string. (Detail: it assumed
881 UTF-8 input when processing these items.)
883 16. Added a lot of (int) casts to avoid compiler warnings in systems where
884 size_t is 64-bit (#991).
886 17. Added a check for running out of memory when PCRE is compiled with
887 --disable-stack-for-recursion (#990).
889 18. If the last data line in a file for pcretest does not have a newline on
890 the end, a newline was missing in the output.
892 19. The default pcre_chartables.c file recognizes only ASCII characters (values
893 less than 128) in its various bitmaps. However, there is a facility for
894 generating tables according to the current locale when PCRE is compiled. It
895 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
896 characters, are recognized by isspace() and therefore were getting set in
897 these tables, and indeed these tables seem to approximate to ISO 8859. This
898 caused a problem in UTF-8 mode when pcre_study() was used to create a list
899 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
900 which of course cannot start UTF-8 characters. I have changed the code so
901 that only real ASCII characters (less than 128) and the correct starting
902 bytes for UTF-8 encodings are set for characters greater than 127 when in
903 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
904 altogether.)
906 20. Added the /T option to pcretest so as to be able to run tests with non-
907 standard character tables, thus making it possible to include the tests
908 used for 19 above in the standard set of tests.
910 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
911 reference to a subpattern the other side of a comment that contains an
912 opening parenthesis caused either an internal compiling error, or a
913 reference to the wrong subpattern.
916 Version 8.02 19-Mar-2010
917 ------------------------
919 1. The Unicode data tables have been updated to Unicode 5.2.0.
921 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
922 configured.
924 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
925 original author of that file, following a query about its status.
927 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
928 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
930 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
931 quantifier applied to a forward-referencing subroutine call, could compile
932 incorrect code or give the error "internal error: previously-checked
933 referenced subpattern not found".
935 6. Both MS Visual Studio and Symbian OS have problems with initializing
936 variables to point to external functions. For these systems, therefore,
937 pcre_malloc etc. are now initialized to local functions that call the
938 relevant global functions.
940 7. There were two entries missing in the vectors called coptable and poptable
941 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
942 I've fixed the data, and added a kludgy way of testing at compile time that
943 the lengths are correct (equal to the number of opcodes).
945 8. Following on from 7, I added a similar kludge to check the length of the
946 eint vector in pcreposix.c.
948 9. Error texts for pcre_compile() are held as one long string to avoid too
949 much relocation at load time. To find a text, the string is searched,
950 counting zeros. There was no check for running off the end of the string,
951 which could happen if a new error number was added without updating the
952 string.
954 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
956 11. \K was not working if it appeared in an atomic group or in a group that
957 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
958 \K is "not well defined" if used in an assertion. PCRE now accepts it if
959 the assertion is positive, but not if it is negative.
961 12. Change 11 fortuitously reduced the size of the stack frame used in the
962 "match()" function of pcre_exec.c by one pointer. Forthcoming
963 implementation of support for (*MARK) will need an extra pointer on the
964 stack; I have reserved it now, so that the stack frame size does not
965 decrease.
967 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
968 item in branch that calls a recursion is a subroutine call - as in the
969 second branch in the above example - was incorrectly given the compile-
970 time error "recursive call could loop indefinitely" because pcre_compile()
971 was not correctly checking the subroutine for matching a non-empty string.
973 14. The checks for overrunning compiling workspace could trigger after an
974 overrun had occurred. This is a "should never occur" error, but it can be
975 triggered by pathological patterns such as hundreds of nested parentheses.
976 The checks now trigger 100 bytes before the end of the workspace.
978 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
981 Version 8.01 19-Jan-2010
982 ------------------------
984 1. If a pattern contained a conditional subpattern with only one branch (in
985 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
986 computed the wrong minimum data length (which is of course zero for such
987 subpatterns). This could cause incorrect "no match" results.
989 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
990 the pattern is reset in the first branch, pcre_compile() failed with
991 "internal error: code overflow at offset...". This happened only when
992 the reset was to the original external option setting. (An optimization
993 abstracts leading options settings into an external setting, which was the
994 cause of this.)
996 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
997 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
998 assertion pattern did not match (meaning that the assertion was true), it
999 was incorrectly treated as false if the SKIP had been reached during the
1000 matching. This also applied to assertions used as conditions.
1002 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1003 assertion subpattern, including such a pattern used as a condition,
1004 unpredictable results occurred, instead of the error return
1007 5. The C++ GlobalReplace function was not working like Perl for the special
1008 situation when an empty string is matched. It now does the fancy magic
1009 stuff that is necessary.
1011 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1012 removed. (These were left over from very, very early versions of PCRE.)
1014 7. Some cosmetic changes to the code to make life easier when compiling it
1015 as part of something else:
1017 (a) Change DEBUG to PCRE_DEBUG.
1019 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1020 called "current" as "current_branch", to prevent a collision with the
1021 Linux macro when compiled as a kernel module.
1023 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1024 prevent a collision with the Linux macro when compiled as a kernel
1025 module.
1027 8. In pcre_compile() there are some checks for integer overflows that used to
1028 cast potentially large values to (double). This has been changed to that
1029 when building, a check for int64_t is made, and if it is found, it is used
1030 instead, thus avoiding the use of floating point arithmetic. (There is no
1031 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1032 double.
1034 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1035 2005 (difference between two addresses compared to an unsigned value).
1037 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1038 custom one, because of the following reported problem in Windows:
1040 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1041 under Win32.
1042 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1043 therefore missing the function definition.
1044 - The compiler thus generates a "C" signature for the test function.
1045 - The linker fails to find the "C" function.
1046 - PCRE fails to configure if asked to do so against libbz2.
1048 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1049 messages were output:
1051 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1052 rerunning libtoolize, to keep the correct libtool macros in-tree.
1053 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1055 I have done both of these things.
1057 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1058 most of the time, it *can* run out if it is given a pattern that contains a
1059 runaway infinite recursion. I updated the discussion in the pcrestack man
1060 page.
1062 13. Now that we have gone to the x.xx style of version numbers, the minor
1063 version may start with zero. Using 08 or 09 is a bad idea because users
1064 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1065 interpreted as invalid octal numbers. I've updated the previous comment in
1066 configure.ac, and also added a check that gives an error if 08 or 09 are
1067 used.
1069 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1070 causing partial matching to fail when the end of the subject matched \W
1071 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1073 15. There were some discrepancies between the declarations in pcre_internal.h
1074 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1075 their definitions. The declarations used "const uschar *" and the
1076 definitions used USPTR. Even though USPTR is normally defined as "const
1077 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1078 reported that: "This difference in casting confuses some C++ compilers, for
1079 example, SunCC recognizes above declarations as different functions and
1080 generates broken code for hbpcre." I have changed the declarations to use
1081 USPTR.
1083 16. GNU libtool is named differently on some systems. The autogen.sh script now
1084 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1085 (FreeBSD).
1087 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1088 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1089 comment: "Figure out how to create a longlong from a string: strtoll and
1090 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1091 instance, but it only takes 2 args instead of 3!"
1093 18. A subtle bug concerned with back references has been fixed by a change of
1094 specification, with a corresponding code fix. A pattern such as
1095 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1096 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1097 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1098 same bug. Such groups have to be quantified to be useful, or contained
1099 inside another quantified group. (If there's no repetition, the reference
1100 can never match.) The problem arises because, having left the group and
1101 moved on to the rest of the pattern, a later failure that backtracks into
1102 the group uses the captured value from the final iteration of the group
1103 rather than the correct earlier one. I have fixed this in PCRE by forcing
1104 any group that contains a reference to itself to be an atomic group; that
1105 is, there cannot be any backtracking into it once it has completed. This is
1106 similar to recursive and subroutine calls.
1109 Version 8.00 19-Oct-09
1110 ----------------------
1112 1. The table for translating pcre_compile() error codes into POSIX error codes
1113 was out-of-date, and there was no check on the pcre_compile() error code
1114 being within the table. This could lead to an OK return being given in
1115 error.
1117 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1118 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1119 in a Windows environment.
1121 3. The pcregrep --count option prints the count for each file even when it is
1122 zero, as does GNU grep. However, pcregrep was also printing all files when
1123 --files-with-matches was added. Now, when both options are given, it prints
1124 counts only for those files that have at least one match. (GNU grep just
1125 prints the file name in this circumstance, but including the count seems
1126 more useful - otherwise, why use --count?) Also ensured that the
1127 combination -clh just lists non-zero counts, with no names.
1129 4. The long form of the pcregrep -F option was incorrectly implemented as
1130 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1131 but it seems right to fix it, and I didn't think it was worth preserving
1132 the old behaviour.
1134 5. The command line items --regex=pattern and --regexp=pattern were not
1135 recognized by pcregrep, which required --regex pattern or --regexp pattern
1136 (with a space rather than an '='). The man page documented the '=' forms,
1137 which are compatible with GNU grep; these now work.
1139 6. No libpcreposix.pc file was created for pkg-config; there was just
1140 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1142 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1143 when UCP support is not needed, by modifying the Python script that
1144 generates it from Unicode data files. This should not matter if the module
1145 is correctly used as a library, but I received one complaint about 50K of
1146 unwanted data. My guess is that the person linked everything into his
1147 program rather than using a library. Anyway, it does no harm.
1149 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1150 was a minimum greater than 1 for a wide character in a possessive
1151 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1152 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1153 character. Chaos in the form of incorrect output or a compiling loop could
1154 result.
1156 9. The restrictions on what a pattern can contain when partial matching is
1157 requested for pcre_exec() have been removed. All patterns can now be
1158 partially matched by this function. In addition, if there are at least two
1159 slots in the offset vector, the offset of the earliest inspected character
1160 for the match and the offset of the end of the subject are set in them when
1161 PCRE_ERROR_PARTIAL is returned.
1163 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1164 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1165 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1166 and may be more useful for multi-segment matching.
1168 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1169 used to be given if ever the end of the subject was reached; now it is
1170 given only if matching could not proceed because another character was
1171 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1172 string "Z", which now yields "no match" instead of "partial match". In the
1173 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1174 final character ended with (*FAIL).
1176 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1177 if the pattern had a "must contain" character that was already found in the
1178 earlier partial match, unless partial matching was again requested. For
1179 example, with the pattern /dog.(body)?/, the "must contain" character is
1180 "g". If the first part-match was for the string "dog", restarting with
1181 "sbody" failed. This bug has been fixed.
1183 13. The string returned by pcre_dfa_exec() after a partial match has been
1184 changed so that it starts at the first inspected character rather than the
1185 first character of the match. This makes a difference only if the pattern
1186 starts with a lookbehind assertion or \b or \B (\K is not supported by
1187 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1188 matching functions compatible, and I think it's the right thing to do.
1190 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1191 so that the demonstration program is easily available in environments where
1192 PCRE has not been installed from source.
1194 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1195 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1196 library.
1198 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1199 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1200 is not the first non-POSIX option to be added. Clearly some people find
1201 these options useful.
1203 17. If a caller to the POSIX matching function regexec() passes a non-zero
1204 value for nmatch with a NULL value for pmatch, the value of
1205 nmatch is forced to zero.
1207 18. RunGrepTest did not have a test for the availability of the -u option of
1208 the diff command, as RunTest does. It now checks in the same way as
1209 RunTest, and also checks for the -b option.
1211 19. If an odd number of negated classes containing just a single character
1212 interposed, within parentheses, between a forward reference to a named
1213 subpattern and the definition of the subpattern, compilation crashed with
1214 an internal error, complaining that it could not find the referenced
1215 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1216 [The bug was that it was starting one character too far in when skipping
1217 over the character class, thus treating the ] as data rather than
1218 terminating the class. This meant it could skip too much.]
1220 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1221 /g option in pcretest when the pattern contains \K, which makes it possible
1222 to have an empty string match not at the start, even when the pattern is
1223 anchored. Updated pcretest and pcredemo to use this option.
1225 21. If the maximum number of capturing subpatterns in a recursion was greater
1226 than the maximum at the outer level, the higher number was returned, but
1227 with unset values at the outer level. The correct (outer level) value is
1228 now given.
1230 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1231 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1232 make it do so. The string so far is captured, making this feature
1233 compatible with Perl.
1235 23. The tests have been re-organized, adding tests 11 and 12, to make it
1236 possible to check the Perl 5.10 features against Perl 5.10.
1238 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1239 pattern matches a fixed length string. PCRE did not allow this; now it
1240 does. Neither allows recursion.
1242 25. I finally figured out how to implement a request to provide the minimum
1243 length of subject string that was needed in order to match a given pattern.
1244 (It was back references and recursion that I had previously got hung up
1245 on.) This code has now been added to pcre_study(); it finds a lower bound
1246 to the length of subject needed. It is not necessarily the greatest lower
1247 bound, but using it to avoid searching strings that are too short does give
1248 some useful speed-ups. The value is available to calling programs via
1249 pcre_fullinfo().
1251 26. While implementing 25, I discovered to my embarrassment that pcretest had
1252 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1253 study optimizations had never been tested with that matching function.
1254 Oops. What is worse, even when it was passed study data, there was a bug in
1255 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1256 were also very few tests of studied patterns with pcre_dfa_exec().
1258 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1259 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1260 on the other side of the coin, they are no longer allowed to have different
1261 names, because these cannot be distinguished in PCRE, and this has caused
1262 confusion. (This is a difference from Perl.)
1264 28. When duplicate subpattern names are present (necessarily with different
1265 numbers, as required by 27 above), and a test is made by name in a
1266 conditional pattern, either for a subpattern having been matched, or for
1267 recursion in such a pattern, all the associated numbered subpatterns are
1268 tested, and the overall condition is true if the condition is true for any
1269 one of them. This is the way Perl works, and is also more like the way
1270 testing by number works.
1273 Version 7.9 11-Apr-09
1274 ---------------------
1276 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1277 (pcretest), all targets were linked against these libraries. This included
1278 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1279 libraries. This caused unwanted dependencies to be created. This problem
1280 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1281 pcretest is linked with readline.
1283 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1284 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1285 moved outside it again, because FALSE and TRUE are already defined in AIX,
1286 but BOOL is not.
1288 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1289 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1291 4. The pcregrep documentation said spaces were inserted as well as colons (or
1292 hyphens) following file names and line numbers when outputting matching
1293 lines. This is not true; no spaces are inserted. I have also clarified the
1294 wording for the --colour (or --color) option.
1296 5. In pcregrep, when --colour was used with -o, the list of matching strings
1297 was not coloured; this is different to GNU grep, so I have changed it to be
1298 the same.
1300 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1301 each matching line was coloured. Now it goes on to look for further matches
1302 of any of the test patterns, which is the same behaviour as GNU grep.
1304 7. A pattern that could match an empty string could cause pcregrep to loop; it
1305 doesn't make sense to accept an empty string match in pcregrep, so I have
1306 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1307 seems to be how GNU grep behaves.
1309 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1310 start or after a newline", because the conditional assertion was not being
1311 correctly handled. The rule now is that both the assertion and what follows
1312 in the first alternative must satisfy the test.
1314 9. If auto-callout was enabled in a pattern with a conditional group whose
1315 condition was an assertion, PCRE could crash during matching, both with
1316 pcre_exec() and pcre_dfa_exec().
1318 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1319 used for matching.
1321 11. Unicode property support in character classes was not working for
1322 characters (bytes) greater than 127 when not in UTF-8 mode.
1324 12. Added the -M command line option to pcretest.
1326 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1328 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1330 16. Added comments and documentation about mis-use of no_arg in the C++
1331 wrapper.
1333 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1334 from Martin Jerabek that uses macro names for all relevant character and
1335 string constants.
1337 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1338 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1339 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1340 these, but not everybody uses configure.
1342 19. A conditional group that had only one branch was not being correctly
1343 recognized as an item that could match an empty string. This meant that an
1344 enclosing group might also not be so recognized, causing infinite looping
1345 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1346 with the subject "ab", where knowledge that the repeated group can match
1347 nothing is needed in order to break the loop.
1349 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1350 exec(), but without supplying a callout function, matching went wrong.
1352 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1353 leak if the size of the offset vector was greater than 30. When the vector
1354 is smaller, the saved offsets during recursion go onto a local stack
1355 vector, but for larger vectors malloc() is used. It was failing to free
1356 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1357 error, in fact).
1359 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1360 heapframe that is used only when UTF-8 support is enabled. This caused no
1361 problem, but was untidy.
1363 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1364 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1365 included within another project.
1367 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1368 slightly modified by me:
1370 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1371 not building pcregrep.
1373 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1374 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1376 25. Forward references, both numeric and by name, in patterns that made use of
1377 duplicate group numbers, could behave incorrectly or give incorrect errors,
1378 because when scanning forward to find the reference group, PCRE was not
1379 taking into account the duplicate group numbers. A pattern such as
1380 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1382 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1383 the feature of a custom pointer more persuasive (as requested by a user).
1385 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1386 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1387 pre-defined.
1389 28. Added support for (*UTF8) at the start of a pattern.
1391 29. Arrange for flags added by the "release type" setting in CMake to be shown
1392 in the configuration summary.
1395 Version 7.8 05-Sep-08
1396 ---------------------
1398 1. Replaced UCP searching code with optimized version as implemented for Ad
1399 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1400 stage table and inline lookup instead of a function, giving speed ups of 2
1401 to 5 times on some simple patterns that I tested. Permission was given to
1402 distribute the MultiStage2.py script that generates the tables (it's not in
1403 the tarball, but is in the Subversion repository).
1405 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1406 scripts.
1408 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1409 a group with a zero qualifier. The result of the study could be incorrect,
1410 or the function might crash, depending on the pattern.
1412 4. Caseless matching was not working for non-ASCII characters in back
1413 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1414 It now works when Unicode Property Support is available.
1416 5. In pcretest, an escape such as \x{de} in the data was always generating
1417 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1418 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1419 truncation.
1421 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1423 7. Added two (int) casts to pcregrep when printing the difference of two
1424 pointers, in case they are 64-bit values.
1426 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1427 test 2 if it fails.
1429 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1430 and a #define of that name to empty if it is not externally set. This is to
1431 allow users of MSVC to set it if necessary.
1433 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1434 the convenience functions in the pcre_get.c source file.
1436 11. An option change at the start of a pattern that had top-level alternatives
1437 could cause overwriting and/or a crash. This command provoked a crash in
1438 some environments:
1440 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1442 This potential security problem was recorded as CVE-2008-2371.
1444 12. For a pattern where the match had to start at the beginning or immediately
1445 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1446 pcre_dfa_exec() could read past the end of the passed subject if there was
1447 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1448 pcretest so that it places the subject at the end of its malloc-ed buffer.
1450 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1451 exec() might read past the end of the data buffer in UTF-8 mode.
1453 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1454 the data contained the byte 0x85 as part of a UTF-8 character within its
1455 first line. This applied both to normal and DFA matching.
1457 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1458 /^[^d]*?$/8 failed to match "abc".
1460 16. Added a missing copyright notice to pcrecpp_internal.h.
1462 17. Make it more clear in the documentation that values returned from
1463 pcre_exec() in ovector are byte offsets, not character counts.
1465 18. Tidied a few places to stop certain compilers from issuing warnings.
1467 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1468 supplied by Stefan Weber. I made a further small update for 7.8 because
1469 there is a change of source arrangements: the pcre_searchfuncs.c module is
1470 replaced by pcre_ucd.c.
1473 Version 7.7 07-May-08
1474 ---------------------
1476 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1477 a string to a long long, pretend we don't even have a long long." This is
1478 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1480 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1481 pre-7.6 versions, which defined a global no_arg variable instead of putting
1482 it in the RE class. (See also #8 below.)
1484 3. Remove a line of dead code, identified by coverity and reported by Nuno
1485 Lopes.
1487 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1489 (1) The include/exclude patterns were being applied to the whole pathnames
1490 of files, instead of just to the final components.
1492 (2) If there was more than one level of directory, the subdirectories were
1493 skipped unless they satisfied the include/exclude conditions. This is
1494 inconsistent with GNU grep (and could even be seen as contrary to the
1495 pcregrep specification - which I improved to make it absolutely clear).
1496 The action now is always to scan all levels of directory, and just
1497 apply the include/exclude patterns to regular files.
1499 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1500 --exclude_dir in the tests to avoid scanning .svn directories.
1502 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1503 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1504 doesn't support NULs in patterns.
1506 7. Added some missing "const"s to declarations of static tables in
1507 pcre_compile.c and pcre_dfa_exec.c.
1509 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1510 caused by fix #2 above. (Subsequently also a second patch to fix the
1511 first patch. And a third patch - this was a messy problem.)
1513 9. Applied Craig's patch to remove the use of push_back().
1515 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1516 matching function regexec().
1518 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1519 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1520 references. PCRE supports relative numbers with this syntax (I don't think
1521 Oniguruma does).
1523 12. Previously, a group with a zero repeat such as (...){0} was completely
1524 omitted from the compiled regex. However, this means that if the group
1525 was called as a subroutine from elsewhere in the pattern, things went wrong
1526 (an internal error was given). Such groups are now left in the compiled
1527 pattern, with a new opcode that causes them to be skipped at execution
1528 time.
1530 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1531 to the way PCRE behaves:
1533 (a) A lone ] character is dis-allowed (Perl treats it as data).
1535 (b) A back reference to an unmatched subpattern matches an empty string
1536 (Perl fails the current match path).
1538 (c) A data ] in a character class must be notated as \] because if the
1539 first data character in a class is ], it defines an empty class. (In
1540 Perl it is not possible to have an empty class.) The empty class []
1541 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1542 The negative empty class [^] matches any one character, independently
1543 of the DOTALL setting.
1545 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1546 non-existent subpattern following a character class starting with ']' and
1547 containing () gave an internal compiling error instead of "reference to
1548 non-existent subpattern". Fortunately, when the pattern did exist, the
1549 compiled code was correct. (When scanning forwards to check for the
1550 existencd of the subpattern, it was treating the data ']' as terminating
1551 the class, so got the count wrong. When actually compiling, the reference
1552 was subsequently set up correctly.)
1554 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1555 it was being rejected as not supported by pcre_dfa_exec(), even though
1556 other assertions are supported. I have made pcre_dfa_exec() support
1557 (*FAIL).
1559 16. The implementation of 13c above involved the invention of a new opcode,
1560 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1561 cannot be changed at match time, I realized I could make a small
1562 improvement to matching performance by compiling OP_ALLANY instead of
1563 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1564 on the OP_ANY path.
1566 17. Compiling pcretest on Windows with readline support failed without the
1567 following two fixes: (1) Make the unistd.h include conditional on
1568 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1570 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1571 ncurses library to be included for pcretest when ReadLine support is
1572 requested, but also to allow for it to be overridden. This patch came from
1573 Daniel Bergström.
1575 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1576 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1577 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1578 spotting this.
1581 Version 7.6 28-Jan-08
1582 ---------------------
1584 1. A character class containing a very large number of characters with
1585 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1586 overflow.
1588 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1589 HAVE_LONG_LONG is not defined.
1591 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1592 bring them up to date and include new features. This patch includes:
1594 - Fixed PH's badly added libz and libbz2 support.
1595 - Fixed a problem with static linking.
1596 - Added pcredemo. [But later removed - see 7 below.]
1597 - Fixed dftables problem and added an option.
1598 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1600 - Added readline support for pcretest.
1601 - Added an listing of the option settings after cmake has run.
1603 4. A user submitted a patch to Makefile that makes it easy to create
1604 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1605 Makefile.am that cause it to include this special target, without
1606 affecting anything else. Note that the same mingw target plus all
1607 the other distribution libraries and programs are now supported
1608 when configuring with CMake (see 6 below) instead of with
1609 Configure/Make.
1611 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1612 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1613 exported in the Windows port". It has not yet been confirmed that the patch
1614 solves the problem, but it does no harm.
1616 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1617 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1618 with CMake, and also correct the comment about stack recursion.
1620 7. Remove the automatic building of pcredemo from the ./configure system and
1621 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1622 of a program that users should build themselves after PCRE is installed, so
1623 building it automatically is not really right. What is more, it gave
1624 trouble in some build environments.
1626 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1629 Version 7.5 10-Jan-08
1630 ---------------------
1632 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1633 values in parens when parsing an RE using the C++ wrapper."
1635 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1636 Characters greater than 255 were excluded from the class instead of being
1637 included.
1639 3. The same bug as (2) above applied to negated POSIX classes such as
1640 [:^space:].
1642 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1643 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1644 I have changed it.
1646 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1647 first named subpattern) and a construct such as (?&a) would reference the
1648 first named subpattern whose name started with "a" (in other words, the
1649 length check was missing). Both these problems are fixed. "Subpattern name
1650 expected" is now given for (?&) (a zero-length name), and this patch also
1651 makes it give the same error for \k'' (previously it complained that that
1652 was a reference to a non-existent subpattern).
1654 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1655 this is right because (?- can be followed by option settings as well as by
1656 digits. I have, however, made the messages clearer.
1658 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1659 than the number used in the conditional) now cause a compile-time error.
1660 This is actually not compatible with Perl, which accepts such patterns, but
1661 treats the conditional as always being FALSE (as PCRE used to), but it
1662 seems to me that giving a diagnostic is better.
1664 8. Change "alphameric" to the more common word "alphanumeric" in comments
1665 and messages.
1667 9. Fix two occurrences of "backslash" in comments that should have been
1668 "backspace".
1670 10. Remove two redundant lines of code that can never be obeyed (their function
1671 was moved elsewhere).
1673 11. The program that makes PCRE's Unicode character property table had a bug
1674 which caused it to generate incorrect table entries for sequences of
1675 characters that have the same character type, but are in different scripts.
1676 It amalgamated them into a single range, with the script of the first of
1677 them. In other words, some characters were in the wrong script. There were
1678 thirteen such cases, affecting characters in the following ranges:
1680 U+002b0 - U+002c1
1681 U+0060c - U+0060d
1682 U+0061e - U+00612
1683 U+0064b - U+0065e
1684 U+0074d - U+0076d
1685 U+01800 - U+01805
1686 U+01d00 - U+01d77
1687 U+01d9b - U+01dbf
1688 U+0200b - U+0200f
1689 U+030fc - U+030fe
1690 U+03260 - U+0327f
1691 U+0fb46 - U+0fbb1
1692 U+10450 - U+1049d
1694 12. The -o option (show only the matching part of a line) for pcregrep was not
1695 compatible with GNU grep in that, if there was more than one match in a
1696 line, it showed only the first of them. It now behaves in the same way as
1697 GNU grep.
1699 13. If the -o and -v options were combined for pcregrep, it printed a blank
1700 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1701 does the same. The return code can be used to tell if there were any
1702 non-matching lines.
1704 14. Added --file-offsets and --line-offsets to pcregrep.
1706 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1707 infinitely looping recursion. The bug was that positive lookaheads were not
1708 being skipped when checking for a possible empty match (negative lookaheads
1709 and both kinds of lookbehind were skipped).
1711 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1712 inclusion of <windows.h> to before rather than after the definition of
1713 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1715 17. Specifying a possessive quantifier with a specific limit for a Unicode
1716 character property caused pcre_compile() to compile bad code, which led at
1717 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1718 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1719 caused the error; without that there was no problem.
1721 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1723 19. Added --enable-pcretest-libreadline.
1725 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1726 RE::GlobalReplace(). As a result, the number of replacements returned was
1727 double what it should be. I removed one of the increments, but Craig sent a
1728 later patch that removed the other one (the right fix) and added unit tests
1729 that check the return values (which was not done before).
1731 21. Several CMake things:
1733 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1734 the names libpcre and libpcreposix, not just pcre and pcreposix.
1736 (2) The above change means that pcretest and pcregrep are now correctly
1737 linked with the newly-built libraries, not previously installed ones.
1741 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1742 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1743 UTF-8 newline character). The key issue is that the pattern starts .*;
1744 this means that the match must be either at the beginning, or after a
1745 newline. The bug was in the code for advancing after a failed match and
1746 checking that the new position followed a newline. It was not taking
1747 account of UTF-8 characters correctly.
1749 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1750 character classes. PCRE was not treating the sequence [:...:] as a
1751 character class unless the ... were all letters. Perl, however, seems to
1752 allow any characters between [: and :], though of course it rejects as
1753 unknown any "names" that contain non-letters, because all the known class
1754 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1755 for example, whereas PCRE did not - it did not recognize a POSIX character
1756 class. This seemed a bit dangerous, so the code has been changed to be
1757 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1758 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1759 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1760 Perl does, and where it didn't before.
1762 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1763 Windows environments %n is disabled by default.
1766 Version 7.4 21-Sep-07
1767 ---------------------
1769 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1770 means that a class such as [\s] counted as "explicit reference to CR or
1771 LF". That isn't really right - the whole point of the change was to try to
1772 help when there was an actual mention of one of the two characters. So now
1773 the change happens only if \r or \n (or a literal CR or LF) character is
1774 encountered.
1776 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1777 of both had grown to the point where there were only 3 bits left.
1778 Fortunately, there was spare space in the data structure, and so I have
1779 moved the internal flags into a new 16-bit field to free up more option
1780 bits.
1782 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1783 but did not set the internal JCHANGED flag - either of these is enough to
1784 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1785 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1786 start sets both bits.
1788 4. Added options (at build time, compile time, exec time) to change \R from
1789 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1791 5. doc/pcresyntax.html was missing from the distribution.
1793 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1794 compatibility, even though it is no longer used.
1796 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1797 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1798 windows.h file is present (where different names are used). [This was
1799 reversed later after testing - see 16 below.]
1801 8. Changed all #include <config.h> to #include "config.h". There were also
1802 some further <pcre.h> cases that I changed to "pcre.h".
1804 9. When pcregrep was used with the --colour option, it missed the line ending
1805 sequence off the lines that it output.
1807 10. It was pointed out to me that arrays of string pointers cause lots of
1808 relocations when a shared library is dynamically loaded. A technique of
1809 using a single long string with a table of offsets can drastically reduce
1810 these. I have refactored PCRE in four places to do this. The result is
1811 dramatic:
1813 Originally: 290
1814 After changing UCP table: 187
1815 After changing error message table: 43
1816 After changing table of "verbs" 36
1817 After changing table of Posix names 22
1819 Thanks to the folks working on Gregex for glib for this insight.
1821 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1822 unicode-properties was also set.
1824 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1826 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1827 checked only for CRLF.
1829 14. Added casts to pcretest.c to avoid compiler warnings.
1831 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1833 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1834 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1835 entirely. This removes changes made in 7 above.
1837 17. The CMake files have been updated, and there is now more information about
1838 building with CMake in the NON-UNIX-USE document.
1841 Version 7.3 28-Aug-07
1842 ---------------------
1844 1. In the rejigging of the build system that eventually resulted in 7.1, the
1845 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1846 brackets there is not right, since it causes compilers to look for an
1847 installed pcre.h, not the version that is in the source that is being
1848 compiled (which of course may be different). I have changed it back to:
1850 #include "pcre.h"
1852 I have a vague recollection that the change was concerned with compiling in
1853 different directories, but in the new build system, that is taken care of
1854 by the VPATH setting the Makefile.
1856 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1857 when the subject happened to end in the byte 0x85 (e.g. if the last
1858 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1859 characters but of course it shouldn't be taken as a newline when it is part
1860 of another character. The bug was that, for an unlimited repeat of . in
1861 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1862 characters when looking for a newline.
1864 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1866 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1867 in debug output.
1869 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1870 long printing in the pcrecpp unittest when running under MinGW.
1872 6. ESC_K was left out of the EBCDIC table.
1874 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1875 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1876 limit also applies to "virtual nesting" when a pattern is recursive, and in
1877 this case 1000 isn't so big. I have been able to remove this limit at the
1878 expense of backing off one optimization in certain circumstances. Normally,
1879 when pcre_exec() would call its internal match() function recursively and
1880 immediately return the result unconditionally, it uses a "tail recursion"
1881 feature to save stack. However, when a subpattern that can match an empty
1882 string has an unlimited repetition quantifier, it no longer makes this
1883 optimization. That gives it a stack frame in which to save the data for
1884 checking that an empty string has been matched. Previously this was taken
1885 from the 1000-entry workspace that had been reserved. So now there is no
1886 explicit limit, but more stack is used.
1888 8. Applied Daniel's patches to solve problems with the import/export magic
1889 syntax that is required for Windows, and which was going wrong for the
1890 pcreposix and pcrecpp parts of the library. These were overlooked when this
1891 problem was solved for the main library.
1893 9. There were some crude static tests to avoid integer overflow when computing
1894 the size of patterns that contain repeated groups with explicit upper
1895 limits. As the maximum quantifier is 65535, the maximum group length was
1896 set at 30,000 so that the product of these two numbers did not overflow a
1897 32-bit integer. However, it turns out that people want to use groups that
1898 are longer than 30,000 bytes (though not repeat them that many times).
1899 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1900 made it possible to implement the integer overflow checks in a much more
1901 dynamic way, which I have now done. The artificial limitation on group
1902 length has been removed - we now have only the limit on the total length of
1903 the compiled pattern, which depends on the LINK_SIZE setting.
1905 10. Fixed a bug in the documentation for get/copy named substring when
1906 duplicate names are permitted. If none of the named substrings are set, the
1907 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1908 empty string.
1910 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1911 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1912 because the ] is interpreted as the first data character and the
1913 terminating ] is not found. PCRE has been made compatible with Perl in this
1914 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1915 cause memory overwriting.
1917 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1918 string has been matched (to stop an infinite loop). It was not recognizing
1919 a conditional subpattern that could match an empty string if that
1920 subpattern was within another subpattern. For example, it looped when
1921 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1922 condition was not nested. This bug has been fixed.
1924 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1925 past the start of the subject in the presence of bytes with the top bit
1926 set, for example "\x8aBCD".
1928 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1929 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1931 14. Optimized (?!) to (*FAIL).
1933 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1934 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1935 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1936 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1937 does: it's just the validity check that is more restrictive.
1939 16. Inserted checks for integer overflows during escape sequence (backslash)
1940 processing, and also fixed erroneous offset values for syntax errors during
1941 backslash processing.
1943 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1944 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1946 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1947 caused an overrun.
1949 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1950 something other than just ASCII characters) inside a group that had an
1951 unlimited repeat caused a loop at compile time (while checking to see
1952 whether the group could match an empty string).
1954 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1955 [\P{Any}] did so. (Error in the code for printing property names.)
1957 21. An orphan \E inside a character class could cause a crash.
1959 22. A repeated capturing bracket such as (A)? could cause a wild memory
1960 reference during compilation.
1962 23. There are several functions in pcre_compile() that scan along a compiled
1963 expression for various reasons (e.g. to see if it's fixed length for look
1964 behind). There were bugs in these functions when a repeated \p or \P was
1965 present in the pattern. These operators have additional parameters compared
1966 with \d, etc, and these were not being taken into account when moving along
1967 the compiled data. Specifically:
1969 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1970 length.
1972 (b) An item such as \pL+ within a repeated group could cause crashes or
1973 loops.
1975 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1976 "reference to non-existent subpattern" error.
1978 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1980 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1981 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1983 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1985 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1986 character were causing crashes (broken optimization).
1988 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1989 \p or \P) caused a compile-time loop.
1991 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1992 break. For example, the unstudied pattern [\r\n]A does not match the string
1993 "\r\nA" because change 7.0/46 below moves the current point on by two
1994 characters after failing to match at the start. However, the pattern \nA
1995 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1996 the same is true. There doesn't seem any very clean way out of this, but
1997 what I have chosen to do makes the common cases work: PCRE now takes note
1998 of whether there can be an explicit match for \r or \n anywhere in the
1999 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2000 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2001 pattern has explicit CR or LF references.
2003 29. Added (*CR) etc for changing newline setting at start of pattern.
2006 Version 7.2 19-Jun-07
2007 ---------------------
2009 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2010 which is apparently normally available under Windows.
2012 2. Re-jig the pcregrep tests with different newline settings in an attempt
2013 to make them independent of the local environment's newline setting.
2015 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2017 4. Some of the "internals" tests were previously cut out when the link size
2018 was not 2, because the output contained actual offsets. The recent new
2019 "Z" feature of pcretest means that these can be cut out, making the tests
2020 usable with all link sizes.
2022 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2023 stack recursion. This gives a massive performance boost under BSD, but just
2024 a small improvement under Linux. However, it saves one field in the frame
2025 in all cases.
2027 6. Added more features from the forthcoming Perl 5.10:
2029 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2030 recursion call. It refers to the nth most recently opened parentheses.
2032 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2033 to be opened parentheses.
2035 (c) Conditions that refer to capturing parentheses can be specified
2036 relatively, for example, (?(-2)... or (?(+3)...
2038 (d) \K resets the start of the current match so that everything before
2039 is not part of it.
2041 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2043 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2044 reference syntax.
2046 (g) (?| introduces a group in which the numbering of parentheses in each
2047 alternative starts with the same number.
2049 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2051 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2054 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2055 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2056 for detecting groups that can match an empty string.
2058 9. A pattern with a very large number of alternatives (more than several
2059 hundred) was running out of internal workspace during the pre-compile
2060 phase, where pcre_compile() figures out how much memory will be needed. A
2061 bit of new cunning has reduced the workspace needed for groups with
2062 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2063 workspace instead of running out of the 4096 that are available.
2065 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2067 11. Applied patch from Google to remove an optimization that didn't quite work.
2068 The report of the bug said:
2070 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2071 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2072 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2074 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2075 it matched the wrong number of bytes.
2078 Version 7.1 24-Apr-07
2079 ---------------------
2081 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2082 that is more "standard", making use of automake and other Autotools. There
2083 is some re-arrangement of the files and adjustment of comments consequent
2084 on this.
2086 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2087 for recursive directory scanning broke on some systems because the files
2088 are not scanned in any specific order and on different systems the order
2089 was different. A call to "sort" has been inserted into RunGrepTest for the
2090 approprate test as a short-term fix. In the longer term there may be an
2091 alternative.
2093 3. I had an email from Eric Raymond about problems translating some of PCRE's
2094 man pages to HTML (despite the fact that I distribute HTML pages, some
2095 people do their own conversions for various reasons). The problems
2096 concerned the use of low-level troff macros .br and .in. I have therefore
2097 removed all such uses from the man pages (some were redundant, some could
2098 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2099 HTML has been updated to handle .nf/.fi and to complain if it encounters
2100 .br or .in.
2102 4. Updated comments in configure.ac that get placed in config.h.in and also
2103 arranged for config.h to be included in the distribution, with the name
2104 config.h.generic, for the benefit of those who have to compile without
2105 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2107 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2108 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2109 makevp.bat for the current PCRE, using the additional files
2110 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2112 6. A Windows user reported a minor discrepancy with test 2, which turned out
2113 to be caused by a trailing space on an input line that had got lost in his
2114 copy. The trailing space was an accident, so I've just removed it.
2116 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2117 that is needed.
2119 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2120 as "const" (a) because they are and (b) because it helps the PHP
2121 maintainers who have recently made a script to detect big data structures
2122 in the php code that should be moved to the .rodata section. I remembered
2123 to update Builducptable as well, so it won't revert if ucptable.h is ever
2124 re-created.
2126 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2127 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2128 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2129 support is not required. This saves 1.5-2K of code, which is important in
2130 some applications.
2132 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2133 so as not to refer to the tables, even though these functions will never be
2134 called when UTF-8 support is disabled. Otherwise there are problems with a
2135 shared library.
2137 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2139 (a) It was defining its arguments as char * instead of void *.
2141 (b) It was assuming that all moves were upwards in memory; this was true
2142 a long time ago when I wrote it, but is no longer the case.
2144 The emulated memove() is provided for those environments that have neither
2145 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2146 is clearly not the case, as these two bugs were recently reported.
2148 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2149 and Detrail to create the HTML documentation, the .txt form of the man
2150 pages, and it removes trailing spaces from listed files. It also creates
2151 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2152 case, it wraps all the #defines with #ifndefs. This script should be run
2153 before "make dist".
2155 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2156 with Unicode property support.
2158 (a) For a maximizing quantifier, if the two different cases of the
2159 character were of different lengths in their UTF-8 codings (there are
2160 some cases like this - I found 11), and the matching function had to
2161 back up over a mixture of the two cases, it incorrectly assumed they
2162 were both the same length.
2164 (b) When PCRE was configured to use the heap rather than the stack for
2165 recursion during matching, it was not correctly preserving the data for
2166 the other case of a UTF-8 character when checking ahead for a match
2167 while processing a minimizing repeat. If the check also involved
2168 matching a wide character, but failed, corruption could cause an
2169 erroneous result when trying to check for a repeat of the original
2170 character.
2172 13. Some tidying changes to the testing mechanism:
2174 (a) The RunTest script now detects the internal link size and whether there
2175 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2176 values substituted by "configure". (The RunGrepTest script already did
2177 this for UTF-8.) The configure.ac script no longer substitutes the
2178 relevant variables.
2180 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2181 with length and offset values. This means that the output is different
2182 for different internal link sizes. Test 2 is skipped for link sizes
2183 other than 2 because of this, bypassing the problem. Unfortunately,
2184 there was also a test in test 3 (the locale tests) that used /B and
2185 failed for link sizes other than 2. Rather than cut the whole test out,
2186 I have added a new /Z option to pcretest that replaces the length and
2187 offset values with spaces. This is now used to make test 3 independent
2188 of link size. (Test 2 will be tidied up later.)
2190 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2191 segmentation fault instead of returning the appropriate error message.
2193 15. In multiline mode when the newline sequence was set to "any", the pattern
2194 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2195 This doesn't seem right; it now treats the CRLF combination as the line
2196 ending, and so does not match in that case. It's only a pattern such as ^$
2197 that would hit this one: something like ^ABC$ would have failed after \r
2198 and then tried again after \r\n.
2200 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2201 in an attempt to make files that differ only in their line terminators
2202 compare equal. This works on Linux.
2204 17. Under certain error circumstances pcregrep might try to free random memory
2205 as it exited. This is now fixed, thanks to valgrind.
2207 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2208 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2209 was because its rules for how to advance for /g after matching an empty
2210 string at the end of a line did not allow for this case. They now check for
2211 it specially.
2213 20. pcretest is supposed to handle patterns and data of any length, by
2214 extending its buffers when necessary. It was getting this wrong when the
2215 buffer for a data line had to be extended.
2217 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2218 CRLF as a newline sequence.
2220 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2221 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2222 I have nevertheless tidied it up.
2224 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2226 24. Added a man page for pcre-config.
2229 Version 7.0 19-Dec-06
2230 ---------------------
2232 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2233 moving to gcc 4.1.1.
2235 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2236 sys/time.h, which is documented as needed for this function. It doesn't
2237 seem to matter on Linux, but it showed up on some releases of OS X.
2239 3. It seems that there are systems where bytes whose values are greater than
2240 127 match isprint() in the "C" locale. The "C" locale should be the
2241 default when a C program starts up. In most systems, only ASCII printing
2242 characters match isprint(). This difference caused the output from pcretest
2243 to vary, making some of the tests fail. I have changed pcretest so that:
2245 (a) When it is outputting text in the compiled version of a pattern, bytes
2246 other than 32-126 are always shown as hex escapes.
2248 (b) When it is outputting text that is a matched part of a subject string,
2249 it does the same, unless a different locale has been set for the match
2250 (using the /L modifier). In this case, it uses isprint() to decide.
2252 4. Fixed a major bug that caused incorrect computation of the amount of memory
2253 required for a compiled pattern when options that changed within the
2254 pattern affected the logic of the preliminary scan that determines the
2255 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2256 that the computed length was too small. The symptoms of this bug were
2257 either the PCRE error "internal error: code overflow" from pcre_compile(),
2258 or a glibc crash with a message such as "pcretest: free(): invalid next
2259 size (fast)". Examples of patterns that provoked this bug (shown in
2260 pcretest format) are:
2262 /(?-x: )/x
2263 /(?x)(?-x: \s*#\s*)/
2264 /((?i)[\x{c0}])/8
2265 /(?i:[\x{c0}])/8
2267 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2268 is now done differently.
2270 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2271 wrapper classes; (b) implement a new function in the C++ scanner that is
2272 more efficient than the old way of doing things because it avoids levels of
2273 recursion in the regex matching; (c) add a paragraph to the documentation
2274 for the FullMatch() function.
2276 6. The escape sequence \n was being treated as whatever was defined as
2277 "newline". Not only was this contrary to the documentation, which states
2278 that \n is character 10 (hex 0A), but it also went horribly wrong when
2279 "newline" was defined as CRLF. This has been fixed.
2281 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2282 was being set to -1 for the "end of line" case (supposedly a value that no
2283 character can have). Though this value is never used (the check for end of
2284 line is "zero bytes in current character"), it caused compiler complaints.
2285 I've changed it to 0xffffffff.
2287 8. In pcre_version.c, the version string was being built by a sequence of
2288 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2289 string (as it is for production releases) called a macro with an empty
2290 argument. The C standard says the result of this is undefined. The gcc
2291 compiler treats it as an empty string (which was what was wanted) but it is
2292 reported that Visual C gives an error. The source has been hacked around to
2293 avoid this problem.
2295 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2296 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2297 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2298 of them did).
2300 10. Originally, pcretest opened its input and output without "b"; then I was
2301 told that "b" was needed in some environments, so it was added for release
2302 5.0 to both the input and output. (It makes no difference on Unix-like
2303 systems.) Later I was told that it is wrong for the input on Windows. I've
2304 now abstracted the modes into two macros, to make it easier to fiddle with
2305 them, and removed "b" from the input mode under Windows.
2307 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2309 12. Added -help and --help to pcretest as an official way of being reminded
2310 of the options.
2312 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2313 and pcrecpp.cc because they annoy compilers at high warning levels.
2315 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2317 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2318 scripts are not C programs :-) and which was not noticed because it works
2319 on Linux.
2321 16. pcretest is supposed to handle any length of pattern and data line (as one
2322 line or as a continued sequence of lines) by extending its input buffer if
2323 necessary. This feature was broken for very long pattern lines, leading to
2324 a string of junk being passed to pcre_compile() if the pattern was longer
2325 than about 50K.
2327 17. I have done a major re-factoring of the way pcre_compile() computes the
2328 amount of memory needed for a compiled pattern. Previously, there was code
2329 that made a preliminary scan of the pattern in order to do this. That was
2330 OK when PCRE was new, but as the facilities have expanded, it has become
2331 harder and harder to keep it in step with the real compile phase, and there
2332 have been a number of bugs (see for example, 4 above). I have now found a
2333 cunning way of running the real compile function in a "fake" mode that
2334 enables it to compute how much memory it would need, while actually only
2335 ever using a few hundred bytes of working memory and without too many
2336 tests of the mode. This should make future maintenance and development
2337 easier. A side effect of this work is that the limit of 200 on the nesting
2338 depth of parentheses has been removed (though this was never a serious
2339 limitation, I suspect). However, there is a downside: pcre_compile() now
2340 runs more slowly than before (30% or more, depending on the pattern). I
2341 hope this isn't a big issue. There is no effect on runtime performance.
2343 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2344 newline (only possible for the last line of a file) and it was a
2345 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2347 19. Added additional timing features to pcretest. (1) The -tm option now times
2348 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2349 separate command line item, by a number that specifies the number of
2350 repeats to use when timing. The default is 50000; this gives better
2351 precision, but takes uncomfortably long for very large patterns.
2353 20. Extended pcre_study() to be more clever in cases where a branch of a
2354 subpattern has no definite first character. For example, (a*|b*)[cd] would
2355 previously give no result from pcre_study(). Now it recognizes that the
2356 first character must be a, b, c, or d.
2358 21. There was an incorrect error "recursive call could loop indefinitely" if
2359 a subpattern (or the entire pattern) that was being tested for matching an
2360 empty string contained only one non-empty item after a nested subpattern.
2361 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2362 incorrectly, because the \d was being skipped in the check.
2364 22. The pcretest program now has a new pattern option /B and a command line
2365 option -b, which is equivalent to adding /B to every pattern. This causes
2366 it to show the compiled bytecode, without the additional information that
2367 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2368 is the same as /B/I).
2370 23. A new optimization is now able automatically to treat some sequences such
2371 as a*b as a*+b. More specifically, if something simple (such as a character
2372 or a simple class like \d) has an unlimited quantifier, and is followed by
2373 something that cannot possibly match the quantified thing, the quantifier
2374 is automatically "possessified".
2376 24. A recursive reference to a subpattern whose number was greater than 39
2377 went wrong under certain circumstances in UTF-8 mode. This bug could also
2378 have affected the operation of pcre_study().
2380 25. Realized that a little bit of performance could be had by replacing
2381 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2383 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2385 27. Possessive quantifiers such as a++ were previously implemented by turning
2386 them into atomic groups such as ($>a+). Now they have their own opcodes,
2387 which improves performance. This includes the automatically created ones
2388 from 23 above.
2390 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2391 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2392 the first matched character to be a colon. This applied both to named and
2393 numbered groups.
2395 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2397 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2398 building PCRE on Windows easier, so I have included it in the distribution.
2400 31. There is now a check in pcretest against a ridiculously large number being
2401 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2402 loop, the loop is abandoned.
2404 32. Forward references to subpatterns in conditions such as (?(2)...) where
2405 subpattern 2 is defined later cause pcre_compile() to search forwards in
2406 the pattern for the relevant set of parentheses. This search went wrong
2407 when there were unescaped parentheses in a character class, parentheses
2408 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2410 33. "Subroutine" calls and backreferences were previously restricted to
2411 referencing subpatterns earlier in the regex. This restriction has now
2412 been removed.
2414 34. Added a number of extra features that are going to be in Perl 5.10. On the
2415 whole, these are just syntactic alternatives for features that PCRE had
2416 previously implemented using the Python syntax or my own invention. The
2417 other formats are all retained for compatibility.
2419 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2420 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2421 also .NET compatible.
2423 (b) A recursion or subroutine call to a named group can now be defined as
2424 (?&name) as well as (?P>name).
2426 (c) A backreference to a named group can now be defined as \k<name> or
2427 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2428 5.10, are also .NET compatible.
2430 (d) A conditional reference to a named group can now use the syntax
2431 (?(<name>) or (?('name') as well as (?(name).
2433 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2434 groups (named and numbered) that are never evaluated inline, but can be
2435 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2436 is always false. There may be only one alternative in such a group.
2438 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2439 as the simple (?(R). The condition is true only if the most recent
2440 recursion is that of the given number or name. It does not search out
2441 through the entire recursion stack.
2443 (g) The escape \gN or \g{N} has been added, where N is a positive or
2444 negative number, specifying an absolute or relative reference.
2446 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2447 some "unreachable code" warnings.
2449 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2450 things, this adds five new scripts.
2452 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2453 There were also incompatibilities regarding the handling of \Q..\E inside
2454 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2455 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2457 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2458 matches an empty string, and forcibly breaks the loop. There were bugs in
2459 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2460 against aaaa the result was just "a" rather than "aaaa", for example. Two
2461 separate and independent bugs (that affected different cases) have been
2462 fixed.
2464 39. Refactored the code to abolish the use of different opcodes for small
2465 capturing bracket numbers. This is a tidy that I avoided doing when I
2466 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2467 The new approach is not only tidier, it makes it possible to reduce the
2468 memory needed to fix the previous bug (38).
2470 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2471 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2472 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2473 mode.
2475 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2476 report.
2478 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2479 copy construction and assignment in the C++ wrapper.
2481 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2482 couple of bugs that could have given wrong results in the "--newline=crlf"
2483 case.
2485 44. Added a number of casts and did some reorganization of signed/unsigned int
2486 variables following suggestions from Dair Grant. Also renamed the variable
2487 "this" as "item" because it is a C++ keyword.
2489 45. Arranged for dftables to add
2491 #include "pcre_internal.h"
2493 to pcre_chartables.c because without it, gcc 4.x may remove the array
2494 definition from the final binary if PCRE is built into a static library and
2495 dead code stripping is activated.
2497 46. For an unanchored pattern, if a match attempt fails at the start of a
2498 newline sequence, and the newline setting is CRLF or ANY, and the next two
2499 characters are CRLF, advance by two characters instead of one.
2502 Version 6.7 04-Jul-06
2503 ---------------------
2505 1. In order to handle tests when input lines are enormously long, pcretest has
2506 been re-factored so that it automatically extends its buffers when
2507 necessary. The code is crude, but this _is_ just a test program. The
2508 default size has been increased from 32K to 50K.
2510 2. The code in pcre_study() was using the value of the re argument before
2511 testing it for NULL. (Of course, in any sensible call of the function, it
2512 won't be NULL.)
2514 3. The memmove() emulation function in pcre_internal.h, which is used on
2515 systems that lack both memmove() and bcopy() - that is, hardly ever -
2516 was missing a "static" storage class specifier.
2518 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2519 containing an extended class (one that cannot be represented by a bitmap
2520 because it contains high-valued characters or Unicode property items, e.g.
2521 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2522 pattern, but PCRE should not loop if you do not (it no longer does).
2523 [Detail: two cases were found: (a) a repeated subpattern containing an
2524 extended class; (b) a recursive reference to a subpattern that followed a
2525 previous extended class. It wasn't skipping over the extended class
2526 correctly when UTF-8 mode was not set.]
2528 5. A negated single-character class was not being recognized as fixed-length
2529 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2530 compile error "lookbehind assertion is not fixed length".
2532 6. The RunPerlTest auxiliary script was showing an unexpected difference
2533 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2534 write a Perl script that can interpret lines of an input file either as
2535 byte characters or as UTF-8, which is what "perltest" was being required to
2536 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2537 can't do is switch easily at run time between having the "use utf8;" pragma
2538 or not. In the end, I fudged it by using the RunPerlTest script to insert
2539 "use utf8;" explicitly for the UTF-8 tests.
2541 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2542 the end of the subject string, contrary to the documentation and to what
2543 Perl does. This was true of both matching functions. Now it matches only at
2544 the start of the subject and immediately after *internal* newlines.
2546 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2547 a pointer to an int instead of a pointer to an unsigned long int. This
2548 caused problems on 64-bit systems.
2550 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2551 instance of the 'standard' template library not being so standard".
2553 10. There was no check on the number of named subpatterns nor the maximum
2554 length of a subpattern name. The product of these values is used to compute
2555 the size of the memory block for a compiled pattern. By supplying a very
2556 long subpattern name and a large number of named subpatterns, the size
2557 computation could be caused to overflow. This is now prevented by limiting
2558 the length of names to 32 characters, and the number of named subpatterns
2559 to 10,000.
2561 11. Subpatterns that are repeated with specific counts have to be replicated in
2562 the compiled pattern. The size of memory for this was computed from the
2563 length of the subpattern and the repeat count. The latter is limited to
2564 65535, but there was no limit on the former, meaning that integer overflow
2565 could in principle occur. The compiled length of a repeated subpattern is
2566 now limited to 30,000 bytes in order to prevent this.
2568 12. Added the optional facility to have named substrings with the same name.
2570 13. Added the ability to use a named substring as a condition, using the
2571 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2572 are numbers (not recommended). Forward references are permitted.
2574 14. Added forward references in named backreferences (if you see what I mean).
2576 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2577 pattern could run off the end of the subject. For example, the pattern
2578 "(?s)(.{1,5})"8 did this with the subject "ab".
2580 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2581 PCRE_CASELESS was set when matching characters that were quantified with ?
2582 or *.
2584 17. A character class other than a single negated character that had a minimum
2585 but no maximum quantifier - for example [ab]{6,} - was not handled
2586 correctly by pce_dfa_exec(). It would match only one character.
2588 18. A valid (though odd) pattern that looked like a POSIX character
2589 class but used an invalid character after [ (for example [[,abc,]]) caused
2590 pcre_compile() to give the error "Failed: internal error: code overflow" or
2591 in some cases to crash with a glibc free() error. This could even happen if
2592 the pattern terminated after [[ but there just happened to be a sequence of
2593 letters, a binary zero, and a closing ] in the memory that followed.
2595 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2596 over the years. Originally (before any Unicode support), just the bottom 8
2597 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2598 output from "man perlunicode" includes this:
2600 The regular expression compiler produces polymorphic opcodes. That
2601 is, the pattern adapts to the data and automatically switches to
2602 the Unicode character scheme when presented with Unicode data--or
2603 instead uses a traditional byte scheme when presented with byte
2604 data.
2606 Sadly, a wide octal escape does not cause a switch, and in a string with
2607 no other multibyte characters, these octal escapes are treated as before.
2608 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2609 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2610 Unicode string.
2612 I have not perpetrated such confusion in PCRE. Up till now, it took just
2613 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2614 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2615 translate to the appropriate multibyte character.
2617 29. Applied some refactoring to reduce the number of warnings from Microsoft
2618 and Borland compilers. This has included removing the fudge introduced
2619 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2620 a warning about an unused variable.
2622 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2623 characters since release 4.0, because Perl (from release 5.004) does not.
2624 [Or at least, is documented not to: some releases seem to be in conflict
2625 with the documentation.] However, when a pattern was studied with
2626 pcre_study() and all its branches started with \s, PCRE still included VT
2627 as a possible starting character. Of course, this did no harm; it just
2628 caused an unnecessary match attempt.
2630 22. Removed a now-redundant internal flag bit that recorded the fact that case
2631 dependency changed within the pattern. This was once needed for "required
2632 byte" processing, but is no longer used. This recovers a now-scarce options
2633 bit. Also moved the least significant internal flag bit to the most-
2634 significant bit of the word, which was not previously used (hangover from
2635 the days when it was an int rather than a uint) to free up another bit for
2636 the future.
2638 23. Added support for CRLF line endings as well as CR and LF. As well as the
2639 default being selectable at build time, it can now be changed at runtime
2640 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2641 specify that it is scanning data with non-default line endings.
2643 24. Changed the definition of CXXLINK to make it agree with the definition of
2644 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2646 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2647 recursions. This makes a big different to stack usage for some patterns.
2649 26. If a subpattern containing a named recursion or subroutine reference such
2650 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2651 the space required for the compiled pattern went wrong and gave too small a
2652 value. Depending on the environment, this could lead to "Failed: internal
2653 error: code overflow at offset 49" or "glibc detected double free or
2654 corruption" errors.
2656 27. Applied patches from Google (a) to support the new newline modes and (b) to
2657 advance over multibyte UTF-8 characters in GlobalReplace.
2659 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2660 difference for some implementation of PCRE in some Windows version.
2662 29. Added some extra testing facilities to pcretest:
2664 \q<number> in a data line sets the "match limit" value
2665 \Q<number> in a data line sets the "match recursion limt" value
2666 -S <number> sets the stack size, where <number> is in megabytes
2668 The -S option isn't available for Windows.
2671 Version 6.6 06-Feb-06
2672 ---------------------
2674 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2675 in pcreposix.h. I have copied the definition from pcre.h.
2677 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2678 because pcre.h is no longer a built file.
2680 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2681 not normally included in the compiled code.
2684 Version 6.5 01-Feb-06
2685 ---------------------
2687 1. When using the partial match feature with pcre_dfa_exec(), it was not
2688 anchoring the second and subsequent partial matches at the new starting
2689 point. This could lead to incorrect results. For example, with the pattern
2690 /1234/, partially matching against "123" and then "a4" gave a match.
2692 2. Changes to pcregrep:
2694 (a) All non-match returns from pcre_exec() were being treated as failures
2695 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2696 error message is output. Some extra information is given for the
2698 probably the only errors that are likely to be caused by users (by
2699 specifying a regex that has nested indefinite repeats, for instance).
2700 If there are more than 20 of these errors, pcregrep is abandoned.
2702 (b) A binary zero was treated as data while matching, but terminated the
2703 output line if it was written out. This has been fixed: binary zeroes
2704 are now no different to any other data bytes.
2706 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2707 used to set a locale for matching. The --locale=xxxx long option has
2708 been added (no short equivalent) to specify a locale explicitly on the
2709 pcregrep command, overriding the environment variables.
2711 (d) When -B was used with -n, some line numbers in the output were one less
2712 than they should have been.
2714 (e) Added the -o (--only-matching) option.
2716 (f) If -A or -C was used with -c (count only), some lines of context were
2717 accidentally printed for the final match.
2719 (g) Added the -H (--with-filename) option.
2721 (h) The combination of options -rh failed to suppress file names for files
2722 that were found from directory arguments.
2724 (i) Added the -D (--devices) and -d (--directories) options.
2726 (j) Added the -F (--fixed-strings) option.
2728 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2730 (l) Added the --colo(u)r option.
2732 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2733 is not present by default.
2735 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2736 items such as (?R) or (?1), when the recursion could match a number of
2737 alternatives. If it matched one of the alternatives, but subsequently,
2738 outside the recursion, there was a failure, the code tried to back up into
2739 the recursion. However, because of the way PCRE is implemented, this is not
2740 possible, and the result was an incorrect result from the match.
2742 In order to prevent this happening, the specification of recursion has
2743 been changed so that all such subpatterns are automatically treated as
2744 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2746 4. I had overlooked the fact that, in some locales, there are characters for
2747 which isalpha() is true but neither isupper() nor islower() are true. In
2748 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2749 and ordfeminine) are like this. This affected the treatment of \w and \W
2750 when they appeared in character classes, but not when they appeared outside
2751 a character class. The bit map for "word" characters is now created
2752 separately from the results of isalnum() instead of just taking it from the
2753 upper, lower, and digit maps. (Plus the underscore character, of course.)
2755 5. The above bug also affected the handling of POSIX character classes such as
2756 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2757 permanent tables. Instead, the bit maps for such a class were previously
2758 created as the appropriate unions of the upper, lower, and digit bitmaps.
2759 Now they are created by subtraction from the [[:word:]] class, which has
2760 its own bitmap.
2762 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2763 It is created by subtracting the vertical space characters (\x09, \x0a,
2764 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2765 subtraction was done in the overall bitmap for a character class, meaning
2766 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2767 be recognized. This bug has been fixed.
2769 7. Patches from the folks at Google:
2771 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2772 real life, but is still worth protecting against".
2774 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2775 regular expressions".
2777 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2778 have it.
2780 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2781 "configure" and the latter not, in order to fix a problem somebody had
2782 with compiling the Arg class on HP-UX.
2784 (e) Improve the error-handling of the C++ wrapper a little bit.
2786 (f) New tests for checking recursion limiting.
2788 8. The pcre_memmove() function, which is used only if the environment does not
2789 have a standard memmove() function (and is therefore rarely compiled),
2790 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2791 returning a result (though PCRE never actually uses the result).
2793 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2794 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2795 returned instead of calling malloc() with an overflowing number that would
2796 most likely cause subsequent chaos.
2798 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2800 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2801 with this option is matched, the nmatch and pmatch options of regexec() are
2802 ignored.
2804 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2805 provided in case anyone wants to the the POSIX interface with UTF-8
2806 strings.
2808 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2809 C++ linking (needed for some HP-UX environments).
2811 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2812 (unused parameter) and in the pcre_printint() function (omitted "default"
2813 switch label when the default is to do nothing).
2815 15. Added some code to make it possible, when PCRE is compiled as a C++
2816 library, to replace subject pointers for pcre_exec() with a smart pointer
2817 class, thus making it possible to process discontinuous strings.
2819 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2820 much the same function. They were added by different people who were trying
2821 to make PCRE easy to compile on non-Unix systems. It has been suggested
2822 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2823 for compiling on Windows systems. I have therefore replaced it with
2824 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2825 defaults to "extern" for C or "extern C" for C++, which works fine on
2826 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2827 SCOPE with something explicit in config.h. In addition:
2829 (a) pcreposix.h still had just "extern" instead of either of these macros;
2830 I have replaced it with PCRE_DATA_SCOPE.
2832 (b) Functions such as _pcre_xclass(), which are internal to the library,
2833 but external in the C sense, all had PCRE_EXPORT in their definitions.
2834 This is apparently wrong for the Windows case, so I have removed it.
2835 (It makes no difference on Unix-like systems.)
2837 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2838 of recursive calls to match(). This is different to MATCH_LIMIT because
2839 that limits the total number of calls to match(), not all of which increase
2840 the depth of recursion. Limiting the recursion depth limits the amount of
2841 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2842 when PCRE is compiled, and changed at run time. A patch from Google adds
2843 this functionality to the C++ interface.
2845 18. Changes to the handling of Unicode character properties:
2847 (a) Updated the table to Unicode 4.1.0.
2849 (b) Recognize characters that are not in the table as "Cn" (undefined).
2851 (c) I revised the way the table is implemented to a much improved format
2852 which includes recognition of ranges. It now supports the ranges that
2853 are defined in UnicodeData.txt, and it also amalgamates other
2854 characters into ranges. This has reduced the number of entries in the
2855 table from around 16,000 to around 3,000, thus reducing its size
2856 considerably. I realized I did not need to use a tree structure after
2857 all - a binary chop search is just as efficient. Having reduced the
2858 number of entries, I extended their size from 6 bytes to 8 bytes to
2859 allow for more data.
2861 (d) Added support for Unicode script names via properties such as \p{Han}.
2863 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2864 matching that character.
2866 20. When matching a repeated Unicode property with a minimum greater than zero,
2867 (for example \pL{2,}), PCRE could look past the end of the subject if it
2868 reached it while seeking the minimum number of characters. This could
2869 happen only if some of the characters were more than one byte long, because
2870 there is a check for at least the minimum number of bytes.
2872 21. Refactored the implementation of \p and \P so as to be more general, to
2873 allow for more different types of property in future. This has changed the
2874 compiled form incompatibly. Anybody with saved compiled patterns that use
2875 \p or \P will have to recompile them.
2877 22. Added "Any" and "L&" to the supported property types.
2879 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2880 but give a compile time error if the value is greater than 0xff.
2882 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2883 accidentally not being installed or uninstalled.
2885 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2886 made were to insert the current release number. This seemed silly, because
2887 it made things harder for people building PCRE on systems that don't run
2888 "configure". I have turned pcre.h into a distributed file, no longer built
2889 by "configure", with the version identification directly included. There is
2890 no longer a pcre.h.in file.
2892 However, this change necessitated a change to the pcre-config script as
2893 well. It is built from pcre-config.in, and one of the substitutions was the
2894 release number. I have updated configure.ac so that ./configure now finds
2895 the release number by grepping pcre.h.
2897 26. Added the ability to run the tests under valgrind.
2900 Version 6.4 05-Sep-05
2901 ---------------------
2903 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2904 "--" to be printed when multiple files were scanned, even when none of the
2905 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2906 consider it to be a bug, and have restored the previous behaviour.
2908 2. A couple of code tidies to get rid of compiler warnings.
2910 3. The pcretest program used to cheat by referring to symbols in the library
2911 whose names begin with _pcre_. These are internal symbols that are not
2912 really supposed to be visible externally, and in some environments it is
2913 possible to suppress them. The cheating is now confined to including
2914 certain files from the library's source, which is a bit cleaner.
2916 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2917 file's purpose clearer.
2919 5. Reorganized pcre_ucp_findchar().
2922 Version 6.3 15-Aug-05
2923 ---------------------
2925 1. The file libpcre.pc.in did not have general read permission in the tarball.
2927 2. There were some problems when building without C++ support:
2929 (a) If C++ support was not built, "make install" and "make test" still
2930 tried to test it.
2932 (b) There were problems when the value of CXX was explicitly set. Some
2933 changes have been made to try to fix these, and ...
2935 (c) --disable-cpp can now be used to explicitly disable C++ support.
2937 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2938 backslash in a target when C++ was disabled. This confuses some
2939 versions of "make", apparently. Using an intermediate variable solves
2940 this. (Same for CPP_LOBJ.)
2942 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2943 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2944 necessary on certain architectures.
2946 4. Added a setting of -export-symbols-regex to the link command to remove
2947 those symbols that are exported in the C sense, but actually are local
2948 within the library, and not documented. Their names all begin with
2949 "_pcre_". This is not a perfect job, because (a) we have to except some
2950 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2951 available (and never for static libraries). I have made a note to try to
2952 find a way round (a) in the future.
2955 Version 6.2 01-Aug-05
2956 ---------------------
2958 1. There was no test for integer overflow of quantifier values. A construction
2959 such as {1111111111111111} would give undefined results. What is worse, if
2960 a minimum quantifier for a parenthesized subpattern overflowed and became
2961 negative, the calculation of the memory size went wrong. This could have
2962 led to memory overwriting.
2964 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2966 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2967 operating environments where this matters.
2969 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2970 PCRE options from within the C++ wrapper.
2972 5. Named capturing subpatterns were not being correctly counted when a pattern
2973 was compiled. This caused two problems: (a) If there were more than 100
2974 such subpatterns, the calculation of the memory needed for the whole
2975 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2976 back references of the form \12, where the number was greater than 9, were
2977 not recognized as back references, even though there were sufficient
2978 previous subpatterns.
2980 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2981 versions of gcc, e.g. 2.95.4.
2984 Version 6.1 21-Jun-05
2985 ---------------------
2987 1. There was one reference to the variable "posix" in pcretest.c that was not
2988 surrounded by "#if !defined NOPOSIX".
2990 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2991 the cross-check on the old pcre_info() function, for the benefit of the
2992 cut-down version of PCRE that is currently imported into Exim.
2994 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2995 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2996 patterns that start (?i) but I don't think that matters. The use of (?i) is
2997 just an example; this all applies to the other options as well.
2999 4. Since libtool seems to echo the compile commands it is issuing, the output
3000 from "make" can be reduced a bit by putting "@" in front of each libtool
3001 compile command.
3003 5. Patch from the folks at Google for configure.in to be a bit more thorough
3004 in checking for a suitable C++ installation before trying to compile the
3005 C++ stuff. This should fix a reported problem when a compiler was present,
3006 but no suitable headers.
3008 6. The man pages all had just "PCRE" as their title. I have changed them to
3009 be the relevant file name. I have also arranged that these names are
3010 retained in the file doc/pcre.txt, which is a concatenation in text format
3011 of all the man pages except the little individual ones for each function.
3013 7. The NON-UNIX-USE file had not been updated for the different set of source
3014 files that come with release 6. I also added a few comments about the C++
3015 wrapper.
3018 Version 6.0 07-Jun-05
3019 ---------------------
3021 1. Some minor internal re-organization to help with my DFA experiments.
3023 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3024 didn't matter for the library itself when fully configured, but did matter
3025 when compiling without UCP support, or within Exim, where the ucp files are
3026 not imported.
3028 3. Refactoring of the library code to split up the various functions into
3029 different source modules. The addition of the new DFA matching code (see
3030 below) to a single monolithic source would have made it really too
3031 unwieldy, quite apart from causing all the code to be include in a
3032 statically linked application, when only some functions are used. This is
3033 relevant even without the DFA addition now that patterns can be compiled in
3034 one application and matched in another.
3036 The downside of splitting up is that there have to be some external
3037 functions and data tables that are used internally in different modules of
3038 the library but which are not part of the API. These have all had their
3039 names changed to start with "_pcre_" so that they are unlikely to clash
3040 with other external names.
3042 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3043 a different (DFA) algorithm. Although it is slower than the original
3044 function, it does have some advantages for certain types of matching
3045 problem.
3047 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3048 including restarting after a partial match.
3050 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3051 defined when compiling for Windows was sent to me. I have put it into the
3052 code, though I have no means of testing or verifying it.
3054 7. Added the pcre_refcount() auxiliary function.
3056 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3057 match before or at the first newline in the subject string. In pcretest,
3058 the /f option on a pattern can be used to set this.
3060 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3061 would behave wrongly. This has been present in PCRE since release 4.0.
3063 10. A number of changes to the pcregrep command:
3065 (a) Refactored how -x works; insert ^(...)$ instead of setting
3066 PCRE_ANCHORED and checking the length, in preparation for adding
3067 something similar for -w.
3069 (b) Added the -w (match as a word) option.
3071 (c) Refactored the way lines are read and buffered so as to have more
3072 than one at a time available.
3074 (d) Implemented a pcregrep test script.
3076 (e) Added the -M (multiline match) option. This allows patterns to match
3077 over several lines of the subject. The buffering ensures that at least
3078 8K, or the rest of the document (whichever is the shorter) is available
3079 for matching (and similarly the previous 8K for lookbehind assertions).
3081 (f) Changed the --help output so that it now says
3083 -w, --word-regex(p)
3085 instead of two lines, one with "regex" and the other with "regexp"
3086 because that confused at least one person since the short forms are the
3087 same. (This required a bit of code, as the output is generated
3088 automatically from a table. It wasn't just a text change.)
3090 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3091 option but starts with a hyphen. Could be a pattern or a path name
3092 starting with a hyphen, for instance.
3094 (h) "-" can be given as a file name to represent stdin.
3096 (i) When file names are being printed, "(standard input)" is used for
3097 the standard input, for compatibility with GNU grep. Previously
3098 "<stdin>" was used.
3100 (j) The option --label=xxx can be used to supply a name to be used for
3101 stdin when file names are being printed. There is no short form.
3103 (k) Re-factored the options decoding logic because we are going to add
3104 two more options that take data. Such options can now be given in four
3105 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3107 (l) Added the -A, -B, and -C options for requesting that lines of context
3108 around matches be printed.
3110 (m) Added the -L option to print the names of files that do not contain
3111 any matching lines, that is, the complement of -l.
3113 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3114 continue to scan other files.
3116 (o) The -s option was incorrectly implemented. For compatibility with other
3117 greps, it now suppresses the error message for a non-existent or non-
3118 accessible file (but not the return code). There is a new option called
3119 -q that suppresses the output of matching lines, which was what -s was
3120 previously doing.
3122 (p) Added --include and --exclude options to specify files for inclusion
3123 and exclusion when recursing.
3125 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3126 Hopefully, it now does.
3128 12. Missing cast in pcre_study().
3130 13. Added an "uninstall" target to the makefile.
3132 14. Replaced "extern" in the function prototypes in Makefile.in with
3133 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3134 world, but is set differently for Windows.
3136 15. Added a second compiling function called pcre_compile2(). The only
3137 difference is that it has an extra argument, which is a pointer to an
3138 integer error code. When there is a compile-time failure, this is set
3139 non-zero, in addition to the error test pointer being set to point to an
3140 error message. The new argument may be NULL if no error number is required
3141 (but then you may as well call pcre_compile(), which is now just a
3142 wrapper). This facility is provided because some applications need a
3143 numeric error indication, but it has also enabled me to tidy up the way
3144 compile-time errors are handled in the POSIX wrapper.
3146 16. Added VPATH=.libs to the makefile; this should help when building with one
3147 prefix path and installing with another. (Or so I'm told by someone who
3148 knows more about this stuff than I do.)
3150 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3151 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3152 match everything, including newlines. This is not POSIX-compatible, but
3153 somebody wanted the feature. From pcretest it can be activated by using
3154 both the P and the s flags.
3156 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3158 19. libpcre.pc was being incorrectly installed as executable.
3160 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3161 it now also looks for '\r' so that it will work unmodified on Windows.
3163 21. Added Google's contributed C++ wrapper to the distribution.
3165 22. Added some untidy missing memory free() calls in pcretest, to keep
3166 Electric Fence happy when testing.
3170 Version 5.0 13-Sep-04
3171 ---------------------
3173 1. Internal change: literal characters are no longer packed up into items
3174 containing multiple characters in a single byte-string. Each character
3175 is now matched using a separate opcode. However, there may be more than one
3176 byte in the character in UTF-8 mode.
3178 2. The pcre_callout_block structure has two new fields: pattern_position and
3179 next_item_length. These contain the offset in the pattern to the next match
3180 item, and its length, respectively.
3182 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3183 insertion of callouts before each pattern item. Added the /C option to
3184 pcretest to make use of this.
3186 4. On the advice of a Windows user, the lines
3188 #if defined(_WIN32) || defined(WIN32)
3189 _setmode( _fileno( stdout ), 0x8000 );
3190 #endif /* defined(_WIN32) || defined(WIN32) */
3192 have been added to the source of pcretest. This apparently does useful
3193 magic in relation to line terminators.
3195 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3196 for the benefit of those environments where the "b" makes a difference.
3198 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3199 to know about it. I have put a hack into configure.in that adds in code
3200 to set GCC=yes if CC=icc. This seems to end up at a point in the
3201 generated configure script that is early enough to affect the setting of
3202 compiler options, which is what is needed, but I have no means of testing
3203 whether it really works. (The user who reported this had patched the
3204 generated configure script, which of course I cannot do.)
3206 LATER: After change 22 below (new libtool files), the configure script
3207 seems to know about icc (and also ecc). Therefore, I have commented out
3208 this hack in configure.in.
3210 7. Added support for pkg-config (2 patches were sent in).
3212 8. Negated POSIX character classes that used a combination of internal tables
3213 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3214 [[:^ascii]]. Typically, they would match almost any characters. The other
3215 POSIX classes were not broken in this way.
3217 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3218 to find the match, as PCRE was deluded into thinking that the match had to
3219 start at the start point or following a newline. The same bug applied to
3220 patterns with negative forward assertions or any backward assertions
3221 preceding ".*" at the start, unless the pattern required a fixed first
3222 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3224 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3225 starting at the last subject character, bytes beyond the end of the subject
3226 string were read.
3228 11. Renamed the variable "class" as "classbits" to make life easier for C++
3229 users. (Previously there was a macro definition, but it apparently wasn't
3230 enough.)
3232 12. Added the new field "tables" to the extra data so that tables can be passed
3233 in at exec time, or the internal tables can be re-selected. This allows
3234 a compiled regex to be saved and re-used at a later time by a different
3235 program that might have everything at different addresses.
3237 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3238 -R library as well as a -L library.
3240 14. The debugging options of pcretest (-d on the command line or D on a
3241 pattern) showed incorrect output for anything following an extended class
3242 that contained multibyte characters and which was followed by a quantifier.
3244 15. Added optional support for general category Unicode character properties
3245 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3246 support. It adds about 90K to the size of the library. The meanings of the
3247 inbuilt class escapes such as \d and \s have NOT been changed.
3249 16. Updated pcredemo.c to include calls to free() to release the memory for the
3250 compiled pattern.
3252 17. The generated file chartables.c was being created in the source directory
3253 instead of in the building directory. This caused the build to fail if the
3254 source directory was different from the building directory, and was
3255 read-only.
3257 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3258 file. No doubt somebody will tell me if they don't make sense... Also added
3259 Dan Mooney's comments about building on OpenVMS.
3261 19. Added support for partial matching via the PCRE_PARTIAL option for
3262 pcre_exec() and the \P data escape in pcretest.
3264 20. Extended pcretest with 3 new pattern features:
3266 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3267 write the compiled pattern to the file whose name is "rest-of-line".
3268 This is a straight binary dump of the data, with the saved pointer to
3269 the character tables forced to be NULL. The study data, if any, is
3270 written too. After writing, pcretest reads a new pattern.
3272 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3273 compiled pattern from the given file. There must not be any
3274 occurrences of "<" in the file name (pretty unlikely); if there are,
3275 pcretest will instead treat the initial "<" as a pattern delimiter.
3276 After reading in the pattern, pcretest goes on to read data lines as
3277 usual.
3279 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3280 and 16-bit fields in a compiled pattern, to simulate a pattern that
3281 was compiled on a host of opposite endianness.
3283 21. The pcre-exec() function can now cope with patterns that were compiled on
3284 hosts of opposite endianness, with this restriction:
3286 As for any compiled expression that is saved and used later, the tables
3287 pointer field cannot be preserved; the extra_data field in the arguments
3288 to pcre_exec() should be used to pass in a tables address if a value
3289 other than the default internal tables were used at compile time.
3291 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3292 now diagnosed as an error. Previously, most of the time, a negative number
3293 would have been treated as zero, but if in addition "ovector" was passed as
3294 NULL, a crash could occur.
3296 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3297 new versions from the libtool 1.5 distribution (the last one is a copy of
3298 a file called libtool.m4). This seems to have fixed the need to patch
3299 "configure" to support Darwin 1.3 (which I used to do). However, I still
3300 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3301 workstation).
3303 24. Changed the PCRE licence to be the more standard "BSD" licence.
3306 Version 4.5 01-Dec-03
3307 ---------------------
3309 1. There has been some re-arrangement of the code for the match() function so
3310 that it can be compiled in a version that does not call itself recursively.
3311 Instead, it keeps those local variables that need separate instances for
3312 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3313 needs to "recurse". Keeping track of where control must go is done by means
3314 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3315 hide most of the details from the main code, and operates only if
3316 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3317 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3318 operating.
3320 To make it easier for callers to provide specially tailored get/free
3321 functions for this usage, two new functions, pcre_stack_malloc, and
3322 pcre_stack_free, are used. They are always called in strict stacking order,
3323 and the size of block requested is always the same.
3325 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3326 PCRE has been compiled to use the stack or the heap for recursion. The
3327 -C option of pcretest uses this to show which version is compiled.
3329 A new data escape \S, is added to pcretest; it causes the amounts of store
3330 obtained and freed by both kinds of malloc/free at match time to be added
3331 to the output.
3333 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3334 what's available on my current Linux desktop machine.
3336 3. When matching a UTF-8 string, the test for a valid string at the start has
3337 been extended. If start_offset is not zero, PCRE now checks that it points
3338 to a byte that is the start of a UTF-8 character. If not, it returns
3339 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3340 this is necessary because there may be backward assertions in the pattern.
3341 When matching the same subject several times, it may save resources to use
3342 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3344 4. The code for checking the validity of UTF-8 strings has been tightened so
3345 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3346 containing "overlong sequences".
3348 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3349 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3350 should have been "&", but it just so happened that all the cases this let
3351 through by mistake were picked up later in the function.
3353 6. I had used a variable called "isblank" - this is a C99 function, causing
3354 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3356 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3357 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3358 so the version is shown; (c) stop "make test" echoing "./RunTest".
3360 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3362 9. The prototype for memmove() for systems that don't have it was using
3363 size_t, but the inclusion of the header that defines size_t was later. I've
3364 moved the #includes for the C headers earlier to avoid this.
3366 10. Added some adjustments to the code to make it easier to compiler on certain
3367 special systems:
3369 (a) Some "const" qualifiers were missing.
3370 (b) Added the macro EXPORT before all exported functions; by default this
3371 is defined to be empty.
3372 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3373 that it reads its output file name as an argument instead of writing
3374 to the standard output and assuming this can be redirected.
3376 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3377 class containing characters with values greater than 255, PCRE compilation
3378 went into a loop.
3380 12. A recursive reference to a subpattern that was within another subpattern
3381 that had a minimum quantifier of zero caused PCRE to crash. For example,
3382 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3383 recursion. If the recursively-called subpattern itself had a zero repeat,
3384 that was OK.
3386 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3387 buffer into which it was copied (for escape processing) was still set at
3388 1024, so long lines caused crashes.
3390 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3391 "internal error: code overflow...". This applied to any character class
3392 that was followed by a possessive quantifier.
3394 15. Modified the Makefile to add libpcre.la as a prerequisite for
3395 libpcreposix.la because I was told this is needed for a parallel build to
3396 work.
3398 16. If a pattern that contained .* following optional items at the start was
3399 studied, the wrong optimizing data was generated, leading to matching
3400 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3401 matching string must start with a or b or c. The correct conclusion for
3402 this pattern is that a match can start with any character.
3405 Version 4.4 13-Aug-03
3406 ---------------------
3408 1. In UTF-8 mode, a character class containing characters with values between
3409 127 and 255 was not handled correctly if the compiled pattern was studied.
3410 In fixing this, I have also improved the studying algorithm for such
3411 classes (slightly).
3413 2. Three internal functions had redundant arguments passed to them. Removal
3414 might give a very teeny performance improvement.
3416 3. Documentation bug: the value of the capture_top field in a callout is *one
3417 more than* the number of the hightest numbered captured substring.
3419 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3420 in incorrectly linking with a previously installed version. They now link
3421 explicitly with libpcre.la.
3423 5. configure.in no longer needs to recognize Cygwin specially.
3425 6. A problem in pcre.in for Windows platforms is fixed.
3427 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3428 pcretest, it used to include the size of the study block as part of its
3429 output. Unfortunately, the structure contains a field that has a different
3430 size on different hardware architectures. This meant that the tests that
3431 showed this size failed. As the block is currently always of a fixed size,
3432 this information isn't actually particularly useful in pcretest output, so
3433 I have just removed it.
3435 8. Three pre-processor statements accidentally did not start in column 1.
3436 Sadly, there are *still* compilers around that complain, even though
3437 standard C has not required this for well over a decade. Sigh.
3439 9. In pcretest, the code for checking callouts passed small integers in the
3440 callout_data field, which is a void * field. However, some picky compilers
3441 complained about the casts involved for this on 64-bit systems. Now
3442 pcretest passes the address of the small integer instead, which should get
3443 rid of the warnings.
3445 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3446 both compile and run time, and gives an error if an invalid UTF-8 sequence
3447 is found. There is a option for disabling this check in cases where the
3448 string is known to be correct and/or the maximum performance is wanted.
3450 11. In response to a bug report, I changed one line in Makefile.in from
3452 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3453 to
3454 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3456 to look similar to other lines, but I have no way of telling whether this
3457 is the right thing to do, as I do not use Windows. No doubt I'll get told
3458 if it's wrong...
3461 Version 4.3 21-May-03
3462 ---------------------
3464 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3465 Makefile.
3467 2. Some refactoring to improve the quality of the code:
3469 (i) The utf8_table... variables are now declared "const".
3471 (ii) The code for \cx, which used the "case flipping" table to upper case
3472 lower case letters, now just substracts 32. This is ASCII-specific,
3473 but the whole concept of \cx is ASCII-specific, so it seems
3474 reasonable.
3476 (iii) PCRE was using its character types table to recognize decimal and
3477 hexadecimal digits in the pattern. This is silly, because it handles
3478 only 0-9, a-f, and A-F, but the character types table is locale-
3479 specific, which means strange things might happen. A private
3480 table is now used for this - though it costs 256 bytes, a table is
3481 much faster than multiple explicit tests. Of course, the standard
3482 character types table is still used for matching digits in subject
3483 strings against \d.
3485 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3486 ending in _t are). So I've renamed it as ESC_tee.
3488 3. The first argument for regexec() in the POSIX wrapper should have been
3489 defined as "const".
3491 4. Changed pcretest to use malloc() for its buffers so that they can be
3492 Electric Fenced for debugging.
3494 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3495 to read one or more bytes before the start of the subject string. Often this
3496 had no effect on PCRE's behaviour, but in some circumstances it could
3497 provoke a segmentation fault.
3499 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3500 to try to read one or more bytes before the start of the subject string.
3502 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3503 UTF-8 support could misbehave in various ways if the subject string
3504 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3505 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3506 back over UTF-8 characters.)
3509 Version 4.2 14-Apr-03
3510 ---------------------
3512 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3514 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3515 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3516 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3517 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3518 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3520 Note: automatic setting of the BUILD variables is not yet working
3521 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3522 compile-time but not at link-time
3523 [LINK]: use for linking executables only
3524 make different versions for Windows and non-Windows
3525 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3526 libraries
3527 [LINK_FOR_BUILD]: new variable
3528 [OBJEXT]: use throughout
3529 [EXEEXT]: use throughout
3530 <winshared>: new target
3531 <wininstall>: new target
3532 <dftables.o>: use native compiler
3533 <dftables>: use native linker
3534 <install>: handle Windows platform correctly
3535 <clean>: ditto
3536 <check>: ditto
3537 copy DLL to top builddir before testing
3539 As part of these changes, -no-undefined was removed again. This was reported
3540 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3541 in any case.
3543 3. Some tidies to get rid of compiler warnings:
3545 . In the match_data structure, match_limit was an unsigned long int, whereas
3546 match_call_count was an int. I've made them both unsigned long ints.
3548 . In pcretest the fact that a const uschar * doesn't automatically cast to
3549 a void * provoked a warning.
3551 . Turning on some more compiler warnings threw up some "shadow" variables
3552 and a few more missing casts.
3554 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3555 option, a class that contained a single character with a value between 128
3556 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3558 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3559 option, a class that contained several characters, but with at least one
3560 whose value was between 128 and 255 caused PCRE to crash.
3563 Version 4.1 12-Mar-03
3564 ---------------------
3566 1. Compiling with gcc -pedantic found a couple of places where casts were
3567 needed, and a string in dftables.c that was longer than standard compilers are
3568 required to support.
3570 2. Compiling with Sun's compiler found a few more places where the code could
3571 be tidied up in order to avoid warnings.
3573 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3574 first of these names is deprecated in the latest Autoconf in favour of the name
3575 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3576 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3577 analogy I have changed it to CFLAGS_FOR_BUILD.
3579 4. Added -no-undefined to the linking command in the Makefile, because this is
3580 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3581 linking step for the pcreposix library.
3583 5. PCRE was failing to diagnose the case of two named groups with the same
3584 name.
3586 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3587 literal character that is needed in the subject for a match, and scans along to
3588 ensure that it is present before embarking on the full matching process. This
3589 saves time in cases of nested unlimited repeats that are never going to match.
3590 Problem: the scan can take a lot of time if the subject is very long (e.g.
3591 megabytes), thus penalizing straightforward matches. It is now done only if the
3592 amount of subject to be scanned is less than 1000 bytes.
3594 7. A lesser problem with the same optimization is that it was recording the
3595 first character of an anchored pattern as "needed", thus provoking a search
3596 right along the subject, even when the first match of the pattern was going to
3597 fail. The "needed" character is now not set for anchored patterns, unless it
3598 follows something in the pattern that is of non-fixed length. Thus, it still
3599 fulfils its original purpose of finding quick non-matches in cases of nested
3600 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3603 Version 4.0 17-Feb-03
3604 ---------------------
3606 1. If a comment in an extended regex that started immediately after a meta-item
3607 extended to the end of string, PCRE compiled incorrect data. This could lead to
3608 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3610 2. Moved to autoconf 2.53 and libtool 1.4.2.
3612 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3613 the special perltest8 script is no longer needed - all the tests can be run
3614 from a single perltest script.
3616 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3617 by \s. It has now been removed in PCRE. This means it isn't recognized as
3618 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3619 class [:space:] *does* include VT, thereby creating a mess.
3621 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3622 space and tab.
3624 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3625 its new features into the main test script, reducing the number of scripts.
3627 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3628 were backward compatible, and made the (?i) apply to the whole pattern, as if
3629 /i were given. Now it behaves more logically, and applies the option setting
3630 only to what follows. PCRE has been changed to follow suit. However, if it
3631 finds options settings right at the start of the pattern, it extracts them into
3632 the global options, as before. Thus, they show up in the info data.
3634 8. Added support for the \Q...\E escape sequence. Characters in between are
3635 treated as literals. This is slightly different from Perl in that $ and @ are
3636 also handled as literals inside the quotes. In Perl, they will cause variable
3637 interpolation. Note the following examples:
3639 Pattern PCRE matches Perl matches
3641 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3642 \Qabc\$xyz\E abc\$xyz abc\$xyz
3643 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3645 For compatibility with Perl, \Q...\E sequences are recognized inside character
3646 classes as well as outside them.
3648 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3649 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3650 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3651 signed/unsigned warnings.
3653 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3654 option for pcretest, so I've replaced it by a simple function that does just
3655 that job.
3657 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3658 "pcregrep -".
3660 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3661 Java package. This provides some syntactic sugar for simple cases of what my
3662 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3663 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3664 item, you can use this simplified notation. Note that only makes sense with
3665 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3666 greediness, whatever the setting of the PCRE_UNGREEDY option.
3668 13. A change of greediness default within a pattern was not taking effect at
3669 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3670 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3671 was abstracted outside.
3673 14. PCRE now supports the \G assertion. It is true when the current matching
3674 position is at the start point of the match. This differs from \A when the
3675 starting offset is non-zero. Used with the /g option of pcretest (or similar
3676 code), it works in the same way as it does for Perl's /g option. If all
3677 alternatives of a regex begin with \G, the expression is anchored to the start
3678 match position, and the "anchored" flag is set in the compiled expression.
3680 15. Some bugs concerning the handling of certain option changes within patterns
3681 have been fixed. These applied to options other than (?ims). For example,
3682 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3683 been the other way round. Some of this was related to change 7 above.
3685 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3686 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3687 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3688 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3690 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3691 mode. Unlike ".", it always matches newline, whatever the setting of
3692 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3693 assertions. Perl allows it, but it doesn't (in general) work because it can't
3694 calculate the length of the lookbehind. At least, that's the case for Perl
3695 5.8.0 - I've been told they are going to document that it doesn't work in
3696 future.
3698 18. Added an error diagnosis for escapes that PCRE does not support: these are
3699 \L, \l, \N, \P, \p, \U, \u, and \X.
3701 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3702 reading past the end of the pattern in cases such as /[abcd/.
3704 20. PCRE was getting more memory than necessary for patterns with classes that
3705 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3707 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3708 compiling PCRE for use with Virtual Pascal.
3710 22. Small fix to the Makefile to make it work properly if the build is done
3711 outside the source tree.
3713 23. Added a new extension: a condition to go with recursion. If a conditional
3714 subpattern starts with (?(R) the "true" branch is used if recursion has
3715 happened, whereas the "false" branch is used only at the top level.
3717 24. When there was a very long string of literal characters (over 255 bytes
3718 without UTF support, over 250 bytes with UTF support), the computation of how
3719 much memory was required could be incorrect, leading to segfaults or other
3720 strange effects.
3722 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3723 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3724 there was a subsequent back reference to those brackets. This meant that, for
3725 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3726 possible to check for precisely this case. All we can do is abandon the
3727 optimization if .* occurs inside capturing brackets when there are any back
3728 references whatsoever. (See below for a better fix that came later.)
3730 26. The handling of the optimization for finding the first character of a
3731 non-anchored pattern, and for finding a character that is required later in the
3732 match were failing in some cases. This didn't break the matching; it just
3733 failed to optimize when it could. The way this is done has been re-implemented.
3735 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3737 28. Added a new feature that provides some of the functionality that Perl
3738 provides with (?{...}). The facility is termed a "callout". The way it is done
3739 in PCRE is for the caller to provide an optional function, by setting
3740 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3741 global variable. By default it is unset, which disables all calling out. To get
3742 the function called, the regex must include (?C) at appropriate points. This
3743 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3744 This provides a means of identifying different callout points. When PCRE
3745 reaches such a point in the regex, if pcre_callout has been set, the external
3746 function is called. It is provided with data in a structure called
3747 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3748 matching continues; if it returns a non-zero value, the match at the current
3749 point fails. However, backtracking will occur if possible. [This was changed
3750 later and other features added - see item 49 below.]
3752 29. pcretest is upgraded to test the callout functionality. It provides a
3753 callout function that displays information. By default, it shows the start of
3754 the match and the current position in the text. There are some new data escapes
3755 to vary what happens:
3757 \C+ in addition, show current contents of captured substrings
3758 \C- do not supply a callout function
3759 \C!n return 1 when callout number n is reached
3760 \C!n!m return 1 when callout number n is reached for the mth time
3762 30. If pcregrep was called with the -l option and just a single file name, it
3763 output "<stdin>" if a match was found, instead of the file name.
3765 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3766 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3767 pcre_exec(). This saves a malloc/free per call. The default value of
3768 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3769 when configuring.
3771 32. The default maximum size of a compiled pattern is 64K. There have been a
3772 few cases of people hitting this limit. The code now uses macros to handle the
3773 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3774 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3775 configuring. Tests 2 and 5 work only with 2-byte links because they output
3776 debugging information about compiled patterns.
3778 33. Internal code re-arrangements:
3780 (a) Moved the debugging function for printing out a compiled regex into
3781 its own source file (printint.c) and used #include to pull it into
3782 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3783 separate copies.
3785 (b) Defined the list of op-code names for debugging as a macro in
3786 internal.h so that it is next to the definition of the opcodes.
3788 (c) Defined a table of op-code lengths for simpler skipping along compiled
3789 code. This is again a macro in internal.h so that it is next to the
3790 definition of the opcodes.
3792 34. Added support for recursive calls to individual subpatterns, along the
3793 lines of Robin Houston's patch (but implemented somewhat differently).
3795 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3796 allow it to read and process whole directories in Win32. This code was
3797 contributed by Lionel Fourquaux; it has not been tested by me.
3799 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3800 used to name a group. Names consist of alphanumerics and underscores, and must
3801 be unique. Back references use the syntax (?P=name) and recursive calls use
3802 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3803 numbers. The function pcre_fullinfo() can be used after compilation to extract
3804 a name/number map. There are three relevant calls:
3806 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3807 PCRE_INFO_NAMECOUNT yields the number of entries
3808 PCRE_INFO_NAMETABLE yields a pointer to the map.
3810 The map is a vector of fixed-size entries. The size of each entry depends on
3811 the length of the longest name used. The first two bytes of each entry are the
3812 group number, most significant byte first. There follows the corresponding
3813 name, zero terminated. The names are in alphabetical order.
3815 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3816 case instead of 255. Making it the same both with and without UTF-8 support
3817 means that the same test output works with both.
3819 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3820 calling malloc() with a zero argument.
3822 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3823 optimization. I've improved things by keeping a bitmap of backreferences with
3824 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3825 fact referenced, the optimization can be applied. It is unlikely that a
3826 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3827 the match to follow \n) will appear inside brackets with a number greater than
3828 31, but if it does, any back reference > 31 suppresses the optimization.
3830 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3831 of disabling numbered capturing parentheses. Any opening parenthesis that is
3832 not followed by ? behaves as if it were followed by ?: but named parentheses
3833 can still be used for capturing (and they will acquire numbers in the usual
3834 way).
3836 41. Redesigned the return codes from the match() function into yes/no/error so
3837 that errors can be passed back from deep inside the nested calls. A malloc
3838 failure while inside a recursive subpattern call now causes the
3839 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3841 42. It is now possible to set a limit on the number of times the match()
3842 function is called in a call to pcre_exec(). This facility makes it possible to
3843 limit the amount of recursion and backtracking, though not in a directly
3844 obvious way, because the match() function is used in a number of different
3845 circumstances. The count starts from zero for each position in the subject
3846 string (for non-anchored patterns). The default limit is, for compatibility, a
3847 large number, namely 10 000 000. You can change this in two ways:
3849 (a) When configuring PCRE before making, you can use --with-match-limit=n
3850 to set a default value for the compiled library.
3852 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3853 a different value is set. See 45 below.
3855 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3857 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3858 of things that can be changed at compile time. The first argument specifies
3859 what is wanted and the second points to where the information is to be placed.
3860 The current list of available information is:
3864 The output is an integer that is set to one if UTF-8 support is available;
3865 otherwise it is set to zero.
3869 The output is an integer that it set to the value of the code that is used for
3870 newline. It is either LF (10) or CR (13).
3874 The output is an integer that contains the number of bytes used for internal
3875 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3879 The output is an integer that contains the threshold above which the POSIX
3880 interface uses malloc() for output vectors. See item 31 above.
3884 The output is an unsigned integer that contains the default limit of the number
3885 of match() calls in a pcre_exec() execution. See 42 above.
3887 44. pcretest has been upgraded by the addition of the -C option. This causes it
3888 to extract all the available output from the new pcre_config() function, and to
3889 output it. The program then exits immediately.
3891 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3892 order to support additional features. One way would have been to define
3893 pcre_exec2() (for example) with extra arguments, but this would not have been
3894 extensible, and would also have required all calls to the original function to
3895 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3896 is used for passing in "extra" data from pcre_study().
3898 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3899 contains the following fields:
3901 flags a bitmap indicating which of the following fields are set
3902 study_data opaque data from pcre_study()
3903 match_limit a way of specifying a limit on match() calls for a specific
3904 call to pcre_exec()
3905 callout_data data for callouts (see 49 below)
3907 The flag bits are also defined in pcre.h, and are
3913 The pcre_study() function now returns one of these new pcre_extra blocks, with
3914 the actual study data pointed to by the study_data field, and the
3915 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3916 before. That is, this change is entirely upwards-compatible and requires no
3917 change to existing code.
3919 If you want to pass in additional data to pcre_exec(), you can either place it
3920 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3921 block.
3923 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3924 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3925 times with different match limits, until it finds the minimum value needed for
3926 pcre_exec() to complete. The value is then output. This can be instructive; for
3927 most simple matches the number is quite small, but for pathological cases it
3928 gets very large very quickly.
3930 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3931 returns the size of the data block pointed to by the study_data field in a
3932 pcre_extra block, that is, the value that was passed as the argument to
3933 pcre_malloc() when PCRE was getting memory in which to place the information
3934 created by pcre_study(). The fourth argument should point to a size_t variable.
3935 pcretest has been extended so that this information is shown after a successful
3936 pcre_study() call when information about the compiled regex is being displayed.
3938 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3939 because what follows is always an absolute path. (Later: it turns out that this
3940 is more than cosmetic for MinGW, because it doesn't like empty path
3941 components.)
3943 49. Some changes have been made to the callout feature (see 28 above):
3945 (i) A callout function now has three choices for what it returns:
3947 0 => success, carry on matching
3948 > 0 => failure at this point, but backtrack if possible
3949 < 0 => serious error, return this value from pcre_exec()
3951 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3952 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3953 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3954 use by callout functions. It will never be used by PCRE itself.
3956 (ii) The pcre_extra structure (see 45 above) has a void * field called
3957 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3958 pcre_callout_block structure has a field of the same name. The contents of
3959 the field passed in the pcre_extra structure are passed to the callout
3960 function in the corresponding field in the callout block. This makes it
3961 easier to use the same callout-containing regex from multiple threads. For
3962 testing, the pcretest program has a new data escape
3964 \C*n pass the number n (may be negative) as callout_data
3966 If the callout function in pcretest receives a non-zero value as
3967 callout_data, it returns that value.
3969 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3970 there were some redundant $(CFLAGS) in commands that are now specified as
3971 $(LINK), which already includes $(CFLAGS).
3973 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3974 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3975 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3976 one-byte characters throughout. Note that case-insensitive matching applies
3977 only to characters whose values are less than 256. PCRE doesn't support the
3978 notion of cases for higher-valued characters.
3980 (i) A character class whose characters are all within 0-255 is handled as
3981 a bit map, and the map is inverted for negative classes. Previously, a
3982 character > 255 always failed to match such a class; however it should
3983 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3985 (ii) A negated character class with a single character < 255 is coded as
3986 "not this character" (OP_NOT). This wasn't working properly when the test
3987 character was multibyte, either singly or repeated.
3989 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3990 mode, for example: \x{100}{2,3}.
3992 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3993 singly or repeated) now correctly test multibyte characters. However,
3994 PCRE doesn't recognize any characters with values greater than 255 as
3995 digits, spaces, or word characters. Such characters always match \D, \S,
3996 and \W, and never match \d, \s, or \w.
3998 (v) Classes may now contain characters and character ranges with values
3999 greater than 255. For example: [ab\x{100}-\x{400}].
4001 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4002 PCRE in UTF-8 mode.
4004 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4005 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4006 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4007 value.)
4009 53. The single man page has become too large. I have therefore split it up into
4010 a number of separate man pages. These also give rise to individual HTML pages;
4011 these are now put in a separate directory, and there is an index.html page that
4012 lists them all. Some hyperlinking between the pages has been installed.
4014 54. Added convenience functions for handling named capturing parentheses.
4016 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4017 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4018 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4019 are faulted.
4021 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4022 calling configure. These values are used when compiling the dftables.c program
4023 which is run to generate the source of the default character tables. They
4024 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4025 you will need to set these values.
4027 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4030 Version 3.9 02-Jan-02
4031 ---------------------
4033 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4035 2. If --disable-static was given, the building process failed when trying to
4036 build pcretest and pcregrep. (For some reason it was using libtool to compile
4037 them, which is not right, as they aren't part of the library.)
4040 Version 3.8 18-Dec-01
4041 ---------------------
4043 1. The experimental UTF-8 code was completely screwed up. It was packing the
4044 bytes in the wrong order. How dumb can you get?
4047 Version 3.7 29-Oct-01
4048 ---------------------
4050 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4051 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4052 this didn't happen under Solaris 8, where I normally test things.
4054 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4055 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4056 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4057 this fix an hour or so after the initial 3.7 release.)
4060 Version 3.6 23-Oct-01
4061 ---------------------
4063 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4064 offsets passed as NULL with zero offset count.
4066 2. The config.guess and config.sub files had not been updated when I moved to
4067 the latest autoconf.
4070 Version 3.5 15-Aug-01
4071 ---------------------
4073 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4074 had been forgotten.
4076 2. By using declared but undefined structures, we can avoid using "void"
4077 definitions in pcre.h while keeping the internal definitions of the structures
4078 private.
4080 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4081 user point of view, this means that both static and shared libraries are built
4082 by default, but this can be individually controlled. More of the work of
4083 handling this static/shared cases is now inside libtool instead of PCRE's make
4084 file.
4086 4. The pcretest utility is now installed along with pcregrep because it is
4087 useful for users (to test regexs) and by doing this, it automatically gets
4088 relinked by libtool. The documentation has been turned into a man page, so
4089 there are now .1, .txt, and .html versions in /doc.
4091 5. Upgrades to pcregrep:
4092 (i) Added long-form option names like gnu grep.
4093 (ii) Added --help to list all options with an explanatory phrase.
4094 (iii) Added -r, --recursive to recurse into sub-directories.
4095 (iv) Added -f, --file to read patterns from a file.
4097 6. pcre_exec() was referring to its "code" argument before testing that
4098 argument for NULL (and giving an error if it was NULL).
4100 7. Upgraded Makefile.in to allow for compiling in a different directory from
4101 the source directory.
4103 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4104 options bits, the pointer it was passed was to an int instead of to an unsigned
4105 long int. This mattered only on 64-bit systems.
4107 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4108 generated) instead of pcre.in, which it its source. Also made the same change
4109 in several of the .c files.
4111 10. A new release of gcc defines printf() as a macro, which broke pcretest
4112 because it had an ifdef in the middle of a string argument for printf(). Fixed
4113 by using separate calls to printf().
4115 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4116 script, to force use of CR or LF instead of \n in the source. On non-Unix
4117 systems, the value can be set in config.h.
4119 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4120 absolute limit. Changed the text of the error message to make this clear, and
4121 likewise updated the man page.
4123 13. The limit of 99 on the number of capturing subpatterns has been removed.
4124 The new limit is 65535, which I hope will not be a "real" limit.
4127 Version 3.4 22-Aug-00
4128 ---------------------
4130 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4132 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4135 Version 3.3 01-Aug-00
4136 ---------------------
4138 1. If an octal character was given, but the value was greater than \377, it
4139 was not getting masked to the least significant bits, as documented. This could
4140 lead to crashes in some systems.
4142 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4143 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4145 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4146 These just pass their arguments on to (pcre_free)(), but they are provided
4147 because some uses of PCRE bind it to non-C systems that can call its functions,
4148 but cannot call free() or pcre_free() directly.
4150 4. Add "make test" as a synonym for "make check". Corrected some comments in
4151 the Makefile.
4153 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4154 Makefile.
4156 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4157 command called pgrep for grepping around the active processes.
4159 7. Added the beginnings of support for UTF-8 character strings.
4161 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4162 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4163 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4164 out for the ar command.)
4167 Version 3.2 12-May-00
4168 ---------------------
4170 This is purely a bug fixing release.
4172 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4173 of ZA. This was just one example of several cases that could provoke this bug,
4174 which was introduced by change 9 of version 2.00. The code for breaking
4175 infinite loops after an iteration that matches an empty string was't working
4176 correctly.
4178 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4179 when matched against abbab (for example). After matching an empty string, it
4180 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4181 caused it to match further down the string than it should.
4183 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4184 was there because it doesn't seem to be needed, and it causes trouble on some
4185 systems, as it is not a Standard C header. It has been removed.
4187 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4188 were reported on the Macintosh. The changes were from
4190 while ((c = *(++ptr)) != 0 && c != '\n');
4191 to
4192 while ((c = *(++ptr)) != 0 && c != '\n') ;
4194 Totally extraordinary, but if that's what it takes...
4196 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4197 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4198 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4199 assumes the way PCRE uses memmove() (always moving upwards).
4201 6. PCRE is being used in one environment where strchr() is not available. There
4202 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4203 faster code anyway.
4206 Version 3.1 09-Feb-00
4207 ---------------------
4209 The only change in this release is the fixing of some bugs in Makefile.in for
4210 the "install" target:
4212 (1) It was failing to install pcreposix.h.
4214 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4217 Version 3.0 01-Feb-00
4218 ---------------------
4220 1. Add support for the /+ modifier to perltest (to output $` like it does in
4221 pcretest).
4223 2. Add support for the /g modifier to perltest.
4225 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4226 matches null strings.
4228 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4229 pattern. Perl treats empty patterns specially - it reuses the most recent
4230 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4231 effect.
4233 5. The POSIX interface was broken in that it was just handing over the POSIX
4234 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4235 required a bigger vector, with some working space on the end. This means that
4236 the POSIX wrapper now has to get and free some memory, and copy the results.
4238 6. Added some simple autoconf support, placing the test data and the
4239 documentation in separate directories, re-organizing some of the
4240 information files, and making it build pcre-config (a GNU standard). Also added
4241 libtool support for building PCRE as a shared library, which is now the
4242 default.
4244 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4245 09 are not valid octal constants. Single digits will be used for minor values
4246 less than 10.
4248 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4249 existing programs that set these in the POSIX interface can use PCRE without
4250 modification.
4252 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4253 return all that pcre_info() returns, plus additional data. The pcre_info()
4254 function is retained for compatibility, but is considered to be obsolete.
4256 10. Added experimental recursion feature (?R) to handle one common case that
4257 Perl 5.6 will be able to do with (?p{...}).
4259 11. Added support for POSIX character classes like [:alpha:], which Perl is
4260 adopting.
4263 Version 2.08 31-Aug-99
4264 ----------------------
4266 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4267 trying to match at the startoffset position, but instead was moving forward to
4268 the next newline as if a previous match had failed.
4270 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4271 and could get into a loop if a null string was matched other than at the start
4272 of the subject.
4274 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4275 be distinguished at compile time, and for completeness also added PCRE_DATE.
4277 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4278 in GnuWin32 environments.
4281 Version 2.07 29-Jul-99
4282 ----------------------
4284 1. The documentation is now supplied in plain text form and HTML as well as in
4285 the form of man page sources.
4287 2. C++ compilers don't like assigning (void *) values to other pointer types.
4288 In particular this affects malloc(). Although there is no problem in Standard
4289 C, I've put in casts to keep C++ compilers happy.
4291 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4292 should be (const char *).
4294 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4295 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4296 However, I haven't made this a standard facility. The documentation doesn't
4297 mention it, and the Makefile doesn't support it.
4299 5. The Makefile now contains an "install" target, with editable destinations at
4300 the top of the file. The pcretest program is not installed.
4302 6. pgrep -V now gives the PCRE version number and date.
4304 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4305 causing the entire string to be ignored, instead of just the last character.
4307 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4308 non-matching string, it can take a very, very long time, even for strings of
4309 quite modest length, because of the nested recursion. PCRE now does better in
4310 some of these cases. It does this by remembering the last required literal
4311 character in the pattern, and pre-searching the subject to ensure it is present
4312 before running the real match. In other words, it applies a heuristic to detect
4313 some types of certain failure quickly, and in the above example, if presented
4314 with a string that has no trailing " it gives "no match" very quickly.
4316 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4317 other alternatives are tried instead.
4320 Version 2.06 09-Jun-99
4321 ----------------------
4323 1. Change pcretest's output for amount of store used to show just the code
4324 space, because the remainder (the data block) varies in size between 32-bit and
4325 64-bit systems.
4327 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4328 start matching at. This allows lookbehinds to work when searching for multiple
4329 occurrences in a string.
4331 3. Added additional options to pcretest for testing multiple occurrences:
4333 /+ outputs the rest of the string that follows a match
4334 /g loops for multiple occurrences, using the new startoffset argument
4335 /G loops for multiple occurrences by passing an incremented pointer
4337 4. PCRE wasn't doing the "first character" optimization for patterns starting
4338 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4339 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4340 the letter 'x'. On long subject strings, this gives a significant speed-up.
4343 Version 2.05 21-Apr-99
4344 ----------------------
4346 1. Changed the type of magic_number from int to long int so that it works
4347 properly on 16-bit systems.
4349 2. Fixed a bug which caused patterns starting with .* not to work correctly
4350 when the subject string contained newline characters. PCRE was assuming
4351 anchoring for such patterns in all cases, which is not correct because .* will
4352 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4353 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4354 must be retried after every newline in the subject.
4357 Version 2.04 18-Feb-99
4358 ----------------------
4360 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4361 computation of the store needed to hold the pattern was incorrect (too large).
4362 If such patterns were nested a few deep, this could multiply and become a real
4363 problem.
4365 2. Added /M option to pcretest to show the memory requirement of a specific
4366 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4368 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4369 compiled in such a way that the backtracking after subsequent failure was
4370 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4371 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4374 Version 2.03 02-Feb-99
4375 ----------------------
4377 1. Fixed typo and small mistake in man page.
4379 2. Added 4th condition (GPL supersedes if conflict) and created separate
4380 LICENCE file containing the conditions.
4382 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4383 Perl, that is the internal \ allows the delimiter to be included in the
4384 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4385 the final delimiter, add \ to the end of the pattern (to test the error).
4387 4. Added the convenience functions for extracting substrings after a successful
4388 match. Updated pcretest to make it able to test these functions.
4391 Version 2.02 14-Jan-99
4392 ----------------------
4394 1. Initialized the working variables associated with each extraction so that
4395 their saving and restoring doesn't refer to uninitialized store.
4397 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4398 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4399 fix the problem.
4401 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4402 calls, and wasn't printing the correct value for compiling calls. Increased the
4403 default value of LOOPREPEAT, and the number of significant figures in the
4404 times.
4406 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4408 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4409 a building problem on Windows NT with a FAT file system.
4412 Version 2.01 21-Oct-98
4413 ----------------------
4415 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4416 to character tables built by pcre_maketables() in the current locale. If NULL
4417 is passed, the default tables are used.
4420 Version 2.00 24-Sep-98
4421 ----------------------
4423 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4424 it any more.
4426 2. Allow quantification of (?>) groups, and make it work correctly.
4428 3. The first character computation wasn't working for (?>) groups.
4430 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4431 end of the subject) and add 5.005's \z, which really does match only at the
4432 very end of the subject.
4434 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4436 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4437 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4438 localized options. All options to pcre_study() were also removed.
4440 7. Add other new features from 5.005:
4442 $(?<= positive lookbehind
4443 $(?<! negative lookbehind
4444 (?imsx-imsx) added the unsetting capability
4445 such a setting is global if at outer level; local otherwise
4446 (?imsx-imsx:) non-capturing groups with option setting
4447 (?(cond)re|re) conditional pattern matching
4449 A backreference to itself in a repeated group matches the previous
4450 captured string.
4452 8. General tidying up of studying (both automatic and via "study")
4453 consequential on the addition of new assertions.
4455 9. As in 5.005, unlimited repeated groups that could match an empty substring
4456 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4457 runtime if any iteration does actually match an empty substring.
4459 10. Include the RunTest script in the distribution.
4461 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4462 discrepancies, some of which were old and were also with respect to 5.004. They
4463 have now been fixed.
4466 Version 1.09 28-Apr-98
4467 ----------------------
4469 1. A negated single character class followed by a quantifier with a minimum
4470 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4471 program crashes, or just wrong answers. This did not apply to negated classes
4472 containing more than one character, or to minima other than one.
4475 Version 1.08 27-Mar-98
4476 ----------------------
4478 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4480 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4481 latter must appear before anything that relies on it in the pattern.
4484 Version 1.07 16-Feb-98
4485 ----------------------
4487 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4488 repeat of a potentially empty string).
4491 Version 1.06 23-Jan-98
4492 ----------------------
4494 1. Added Markus Oberhumer's little patches for C++.
4496 2. Literal strings longer than 255 characters were broken.
4499 Version 1.05 23-Dec-97
4500 ----------------------
4502 1. Negated character classes containing more than one character were failing if
4503 PCRE_CASELESS was set at run time.
4506 Version 1.04 19-Dec-97
4507 ----------------------
4509 1. Corrected the man page, where some "const" qualifiers had been omitted.
4511 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4512 input syntax.
4514 3. Fixed memory leak which occurred when a regex with back references was
4515 matched with an offsets vector that wasn't big enough. The temporary memory
4516 that is used in this case wasn't being freed if the match failed.
4518 4. Tidied pcretest to ensure it frees memory that it gets.
4520 5. Temporary memory was being obtained in the case where the passed offsets
4521 vector was exactly big enough.
4523 6. Corrected definition of offsetof() from change 5 below.
4525 7. I had screwed up change 6 below and broken the rules for the use of
4526 setjmp(). Now fixed.
4529 Version 1.03 18-Dec-97
4530 ----------------------
4532 1. A erroneous regex with a missing opening parenthesis was correctly
4533 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4534 on some systems.
4536 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4537 it was reported that one broken compiler failed on the former because "code" is
4538 also an independent variable.
4540 3. The erroneous regex a[]b caused an array overrun reference.
4542 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4543 fail on data ending with that character. (It was going on too far, and checking
4544 the next character, typically a binary zero.) This was specific to the
4545 optimized code for single-character negative classes.
4547 5. Added a contributed patch from the TIN world which does the following:
4549 + Add an undef for memmove, in case the the system defines a macro for it.
4551 + Add a definition of offsetof(), in case there isn't one. (I don't know
4552 the reason behind this - offsetof() is part of the ANSI standard - but
4553 it does no harm).
4555 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4556 most of the places where whitespace preceded '#'. I have given up and
4557 allowed the remaining 2 cases to be at the margin.
4559 + Rename some variables in pcre to eliminate shadowing. This seems very
4560 pedantic, but does no harm, of course.
4562 6. Moved the call to setjmp() into its own function, to get rid of warnings
4563 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4565 7. Constructs such as \d{8,} were compiling into the equivalent of
4566 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4567 outcome, but in this particular case used more store than had been allocated,
4568 which caused the bug to be discovered because it threw up an internal error.
4570 8. The debugging code in both pcre and pcretest for outputting the compiled
4571 form of a regex was going wrong in the case of back references followed by
4572 curly-bracketed repeats.
4575 Version 1.02 12-Dec-97
4576 ----------------------
4578 1. Typos in pcre.3 and comments in the source fixed.
4580 2. Applied a contributed patch to get rid of places where it used to remove
4581 'const' from variables, and fixed some signed/unsigned and uninitialized
4582 variable warnings.
4584 3. Added the "runtest" target to Makefile.
4586 4. Set default compiler flag to -O2 rather than just -O.
4589 Version 1.01 19-Nov-97
4590 ----------------------
4592 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4593 like /([ab]*)*/, that is, for classes with more than one character in them.
4595 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4596 as /((?>a*))*/ (a PCRE_EXTRA facility).
4599 Version 1.00 18-Nov-97
4600 ----------------------
4602 1. Added compile-time macros to support systems such as SunOS4 which don't have
4603 memmove() or strerror() but have other things that can be used instead.
4605 2. Arranged that "make clean" removes the executables.
4608 Version 0.99 27-Oct-97
4609 ----------------------
4611 1. Fixed bug in code for optimizing classes with only one character. It was
4612 initializing a 32-byte map regardless, which could cause it to run off the end
4613 of the memory it had got.
4615 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4618 Version 0.98 22-Oct-97
4619 ----------------------
4621 1. Fixed bug in code for handling temporary memory usage when there are more
4622 back references than supplied space in the ovector. This could cause segfaults.
4625 Version 0.97 21-Oct-97
4626 ----------------------
4628 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4630 2. Optimized negated single characters not to use a bit map.
4632 3. Brought error texts together as macro definitions; clarified some of them;
4633 fixed one that was wrong - it said "range out of order" when it meant "invalid
4634 escape sequence".
4636 4. Changed some char * arguments to const char *.
4638 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4640 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4641 pcretest.
4644 Version 0.96 16-Oct-97
4645 ----------------------
4647 1. Added a simple "pgrep" utility to the distribution.
4649 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4650 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4651 where "ddd" means "one or more decimal digits".
4653 3. Fixed serious bug. If a pattern had a back reference, but the call to
4654 pcre_exec() didn't supply a large enough ovector to record the related
4655 identifying subpattern, the match always failed. PCRE now remembers the number
4656 of the largest back reference, and gets some temporary memory in which to save
4657 the offsets during matching if necessary, in order to ensure that
4658 backreferences always work.
4660 4. Increased the compatibility with Perl in a number of ways:
4662 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4663 to request this handling. The option can be set at compile or exec time.
4665 (b) $ matches before a terminating newline by default; an option
4666 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4667 mode). The option can be set at compile or exec time.
4669 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4670 the same as Perl's. If the decimal number it represents is less than 10
4671 or there aren't that many previous left capturing parentheses, an octal
4672 escape is read. Inside a character class, it's always an octal escape,
4673 even if it is a single digit.
4675 (d) An escaped but undefined alphabetic character is taken as a literal,
4676 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4677 escapes.
4679 (e) {0} is now permitted. (The previous item is removed from the compiled
4680 pattern).
4682 5. Changed all the names of code files so that the basic parts are no longer
4683 than 10 characters, and abolished the teeny "globals.c" file.
4685 6. Changed the handling of character classes; they are now done with a 32-byte
4686 bit map always.
4688 7. Added the -d and /D options to pcretest to make it possible to look at the
4689 internals of compilation without having to recompile pcre.
4692 Version 0.95 23-Sep-97
4693 ----------------------
4695 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4696 \x20 at the start of a run of normal characters. These were being treated as
4697 real characters, instead of the source characters being re-checked.
4700 Version 0.94 18-Sep-97
4701 ----------------------
4703 1. The functions are now thread-safe, with the caveat that the global variables
4704 containing pointers to malloc() and free() or alternative functions are the
4705 same for all threads.
4707 2. Get pcre_study() to generate a bitmap of initial characters for non-
4708 anchored patterns when this is possible, and use it if passed to pcre_exec().
4711 Version 0.93 15-Sep-97
4712 ----------------------
4714 1. /(b)|(:+)/ was computing an incorrect first character.
4716 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4717 but not actually doing anything yet.
4719 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4720 as Perl does (e.g. [-az] or [az-]).
4722 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4723 all possible positions.
4725 5. Split up into different modules to avoid including unneeded functions in a
4726 compiled binary. However, compile and exec are still in one module. The "study"
4727 function is split off.
4729 6. The character tables are now in a separate module whose source is generated
4730 by an auxiliary program - but can then be edited by hand if required. There are
4731 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4732 toupper() in the code.
4734 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4735 make them global. Abolish the function for setting them, as the caller can now
4736 set them directly.
4739 Version 0.92 11-Sep-97
4740 ----------------------
4742 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4743 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4745 2. Caseless matching was not working in character classes if the characters in
4746 the pattern were in upper case.
4748 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4750 4. Make PCRE_ANCHORED public and accept as a compile option.
4752 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4753 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4754 pass them.
4756 6. Give an error if bad option bits passed at compile or run time.
4758 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4759 pcretest to cause it to pass that flag.
4761 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4762 options, and the first character, if set.
4764 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4767 Version 0.91 10-Sep-97
4768 ----------------------
4770 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4771 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4773 2. PCRE was looping on encountering an indefinitely repeated back reference to
4774 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4775 Perl does - treats the match as successful.
4777 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5