ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 962 - (show annotations)
Fri Apr 20 17:28:23 2012 UTC (7 years, 11 months ago) by ph10
File size: 229841 byte(s)
Error occurred while calculating annotation data.
Fix auto-possessifying bugs when PCRE_UCP is not set, but character tables 
specify characters in the range 127-255 are letters, spaces, etc.
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.
100 Version 8.30 04-February-2012
101 -----------------------------
103 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
104 name is defined in ctype.h.
106 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
107 only in quite long subpatterns.
109 3. Removed the function pcre_info(), which has been obsolete and deprecated
110 since it was replaced by pcre_fullinfo() in February 2000.
112 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
113 match a (*MARK), and the match failed at the start of the subject, a
114 reference to memory before the start of the subject could occur. This bug
115 was introduced by fix 17 of release 8.21.
117 5. A reference to an unset group with zero minimum repetition was giving
118 totally wrong answers (in non-JavaScript-compatibility mode). For example,
119 /(another)?(\1?)test/ matched against "hello world test". This bug was
120 introduced in release 8.13.
122 6. Add support for 16-bit character strings (a large amount of work involving
123 many changes and refactorings).
125 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
126 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
127 from a file.
129 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
130 rounding is not applied in this particular case).
132 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
133 if they appear, or are escaped, in patterns.
135 10. Get rid of a number of -Wunused-but-set-variable warnings.
137 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
138 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
139 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
140 also returns the mark "x". This bug applied to capturing parentheses,
141 non-capturing parentheses, and atomic parentheses. It also applied to some
142 assertions.
144 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
145 information out of configure.ac instead of relying on pcre.h.generic, which
146 is not stored in the repository.
148 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
149 -lreadline.
151 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
153 15. Applied Graycode's patch to put the top-level frame on the stack rather
154 than the heap when not using the stack for recursion. This gives a
155 performance improvement in many cases when recursion is not deep.
157 16. Experimental code added to "pcretest -C" to output the stack frame size.
160 Version 8.21 12-Dec-2011
161 ------------------------
163 1. Updating the JIT compiler.
165 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
166 are added as well.
168 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
169 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
170 calling _pcre_jit_exec. Some extra comments are added.
172 4. (*MARK) settings inside atomic groups that do not contain any capturing
173 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
174 was introduced by change 18 for 8.20.
176 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
177 ECMA-262 standard.
179 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
180 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
181 This bug was probably introduced by change 9 of 8.13.
183 7. While fixing 6 above, I noticed that a number of other items were being
184 incorrectly rejected as "not fixed length". This arose partly because newer
185 opcodes had not been added to the fixed-length checking code. I have (a)
186 corrected the bug and added tests for these items, and (b) arranged for an
187 error to occur if an unknown opcode is encountered while checking for fixed
188 length instead of just assuming "not fixed length". The items that were
189 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
190 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
191 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
193 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
194 being incorrectly compiled and would have given unpredicatble results.
196 9. A possessively repeated subpattern with minimum repeat count greater than
197 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
198 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
199 the first (A) could occur when it should not.
201 10. Add a cast and remove a redundant test from the code.
203 11. JIT should use pcre_malloc/pcre_free for allocation.
205 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
206 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
207 is anything other than /usr, -L is still shown).
209 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
211 14. Perl does not support \N without a following name in a [] class; PCRE now
212 also gives an error.
214 15. If a forward reference was repeated with an upper limit of around 2000,
215 it caused the error "internal error: overran compiling workspace". The
216 maximum number of forward references (including repeats) was limited by the
217 internal workspace, and dependent on the LINK_SIZE. The code has been
218 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
219 the default depends on LINK_SIZE. There is a new upper limit (for safety)
220 of around 200,000 forward references. While doing this, I also speeded up
221 the filling in of repeated forward references.
223 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
224 incorrectly expecting the subject to contain another "a" after the start.
226 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
227 in the match, the SKIP should be ignored. This was not happening; instead
228 the SKIP was being treated as NOMATCH. For patterns such as
229 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
230 tested.
232 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
233 now much more compatible with Perl, in particular in cases where the result
234 is a non-match for a non-anchored pattern. For example, if
235 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
236 "m", where previously it did not return a name. A side effect of this
237 change is that for partial matches, the last encountered mark name is
238 returned, as for non matches. A number of tests that were previously not
239 Perl-compatible have been moved into the Perl-compatible test files. The
240 refactoring has had the pleasing side effect of removing one argument from
241 the match() function, thus reducing its stack requirements.
243 19. If the /S+ option was used in pcretest to study a pattern using JIT,
244 subsequent uses of /S (without +) incorrectly behaved like /S+.
246 21. Retrieve executable code size support for the JIT compiler and fixing
247 some warnings.
249 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
250 not work when the shorter character appeared right at the end of the
251 subject string.
253 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
254 systems.
256 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
257 output it when the /M option is used in pcretest.
259 25. The CheckMan script was not being included in the distribution. Also, added
260 an explicit "perl" to run Perl scripts from the PrepareRelease script
261 because this is reportedly needed in Windows.
263 26. If study data was being save in a file and studying had not found a set of
264 "starts with" bytes for the pattern, the data written to the file (though
265 never used) was taken from uninitialized memory and so caused valgrind to
266 complain.
268 27. Updated RunTest.bat as provided by Sheri Pierce.
270 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
272 29. Computation of memory usage for the table of capturing group names was
273 giving an unnecessarily large value.
276 Version 8.20 21-Oct-2011
277 ------------------------
279 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
280 a POSIX class. After further experiments with Perl, which convinced me that
281 Perl has bugs and confusions, a closing square bracket is no longer allowed
282 in a POSIX name. This bug also affected patterns with classes that started
283 with full stops.
285 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
286 captured substring, but while checking the failing first alternative,
287 substring 1 is temporarily captured. If the output vector supplied to
288 pcre_exec() was not big enough for this capture, the yield of the function
289 was still zero ("insufficient space for captured substrings"). This cannot
290 be totally fixed without adding another stack variable, which seems a lot
291 of expense for a edge case. However, I have improved the situation in cases
292 such as /(a)(b)x|abc/ matched against "abc", where the return code
293 indicates that fewer than the maximum number of slots in the ovector have
294 been set.
296 3. Related to (2) above: when there are more back references in a pattern than
297 slots in the output vector, pcre_exec() uses temporary memory during
298 matching, and copies in the captures as far as possible afterwards. It was
299 using the entire output vector, but this conflicts with the specification
300 that only 2/3 is used for passing back captured substrings. Now it uses
301 only the first 2/3, for compatibility. This is, of course, another edge
302 case.
304 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
305 main code base, and can be used by building with --enable-jit. When this is
306 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
307 runtime --no-jit option is given.
309 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
310 ovector, the return from the function was zero, implying that there were
311 other matches that did not fit. The correct "exactly full" value is now
312 returned.
314 6. If a subpattern that was called recursively or as a subroutine contained
315 (*PRUNE) or any other control that caused it to give a non-standard return,
316 invalid errors such as "Error -26 (nested recursion at the same subject
317 position)" or even infinite loops could occur.
319 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
320 computing the minimum length on reaching *ACCEPT, and so ended up with the
321 wrong value of 1 rather than 0. Further investigation indicates that
322 computing a minimum subject length in the presence of *ACCEPT is difficult
323 (think back references, subroutine calls), and so I have changed the code
324 so that no minimum is registered for a pattern that contains *ACCEPT.
326 8. If (*THEN) was present in the first (true) branch of a conditional group,
327 it was not handled as intended. [But see 16 below.]
329 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
330 Sheri Pierce.
332 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
333 the first byte in a match must be "a".
335 11. Change 17 for 8.13 increased the recursion depth for patterns like
336 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
337 pattern contains any instances of (*THEN). If it does not, the old
338 optimizations are restored. It would be nice to do this on a per-group
339 basis, but at the moment that is not feasible.
341 12. In some environments, the output of pcretest -C is CRLF terminated. This
342 broke RunTest's code that checks for the link size. A single white space
343 character after the value is now allowed for.
345 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
346 For "fr", it uses the Windows-specific input and output files.
348 14. If (*THEN) appeared in a group that was called recursively or as a
349 subroutine, it did not work as intended. [But see next item.]
351 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
352 pattern fragments (but not containing any | characters). If A and B are
353 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
354 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
355 D. In other words, Perl considers parentheses that do not contain any |
356 characters to be part of a surrounding alternative, whereas PCRE was
357 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
358 differently. PCRE now behaves in the same way as Perl, except in the case
359 of subroutine/recursion calls such as (?1) which have in any case always
360 been different (but PCRE had them first :-).
362 16. Related to 15 above: Perl does not treat the | in a conditional group as
363 creating alternatives. Such a group is treated in the same way as an
364 ordinary group without any | characters when processing (*THEN). PCRE has
365 been changed to match Perl's behaviour.
367 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
368 RunGrepTest script failed.
370 18. Change 22 for version 13 caused atomic groups to use more stack. This is
371 inevitable for groups that contain captures, but it can lead to a lot of
372 stack use in large patterns. The old behaviour has been restored for atomic
373 groups that do not contain any capturing parentheses.
375 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
376 suppress the check for a minimum subject length at run time. (If it was
377 given to pcre_exec() or pcre_dfa_exec() it did work.)
379 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
380 fail to work when decoding hex characters in data strings in EBCDIC
381 environments.
383 21. It appears that in at least one Mac OS environment, the isxdigit() function
384 is implemented as a macro that evaluates to its argument more than once,
385 contravening the C 90 Standard (I haven't checked a later standard). There
386 was an instance in pcretest which caused it to go wrong when processing
387 \x{...} escapes in subject strings. The has been rewritten to avoid using
388 things like p++ in the argument of isxdigit().
391 Version 8.13 16-Aug-2011
392 ------------------------
394 1. The Unicode data tables have been updated to Unicode 6.0.0.
396 2. Two minor typos in pcre_internal.h have been fixed.
398 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
399 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
400 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
402 4. There were a number of related bugs in the code for matching backrefences
403 caselessly in UTF-8 mode when codes for the characters concerned were
404 different numbers of bytes. For example, U+023A and U+2C65 are an upper
405 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
406 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
407 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
408 2-byte code at the end of the subject (it thought there wasn't enough data
409 left).
411 5. Comprehensive information about what went wrong is now returned by
412 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
413 as the output vector has at least 2 elements. The offset of the start of
414 the failing character and a reason code are placed in the vector.
416 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
417 now returned is for the first byte of the failing character, instead of the
418 last byte inspected. This is an incompatible change, but I hope it is small
419 enough not to be a problem. It makes the returned offset consistent with
420 pcre_exec() and pcre_dfa_exec().
422 7. pcretest now gives a text phrase as well as the error number when
423 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
424 failure, the offset and reason code are output.
426 8. When \R was used with a maximizing quantifier it failed to skip backwards
427 over a \r\n pair if the subsequent match failed. Instead, it just skipped
428 back over a single character (\n). This seems wrong (because it treated the
429 two characters as a single entity when going forwards), conflicts with the
430 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
431 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
432 has been changed.
434 9. Some internal refactoring has changed the processing so that the handling
435 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
436 time (the PCRE_DOTALL option was changed this way some time ago: version
437 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
438 which was always a bit of a fudge. It also means that there is one less
439 argument for the match() function, which reduces its stack requirements
440 slightly. This change also fixes an incompatibility with Perl: the pattern
441 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
443 10. More internal refactoring has drastically reduced the number of recursive
444 calls to match() for possessively repeated groups such as (abc)++ when
445 using pcre_exec().
447 11. While implementing 10, a number of bugs in the handling of groups were
448 discovered and fixed:
450 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
451 (a|)*(?1) gave a compile-time internal error.
452 ((a|)+)+ did not notice that the outer group could match an empty string.
453 (^a|^)+ was not marked as anchored.
454 (.*a|.*)+ was not marked as matching at start or after a newline.
456 12. Yet more internal refactoring has removed another argument from the match()
457 function. Special calls to this function are now indicated by setting a
458 value in a variable in the "match data" data block.
460 13. Be more explicit in pcre_study() instead of relying on "default" for
461 opcodes that mean there is no starting character; this means that when new
462 ones are added and accidentally left out of pcre_study(), testing should
463 pick them up.
465 14. The -s option of pcretest has been documented for ages as being an old
466 synonym of -m (show memory usage). I have changed it to mean "force study
467 for every regex", that is, assume /S for every regex. This is similar to -i
468 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
469 using it. It makes it easier to run collections of tests with and without
470 study enabled, and thereby test pcre_study() more easily. All the standard
471 tests are now run with and without -s (but some patterns can be marked as
472 "never study" - see 20 below).
474 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
475 restoration of the capturing data to the outer values was not happening
476 correctly.
478 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
479 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
480 pattern had matched an empty string, and so incorrectly returned a no
481 match.
483 17. There was optimizing code for the last branch of non-capturing parentheses,
484 and also for the obeyed branch of a conditional subexpression, which used
485 tail recursion to cut down on stack usage. Unfortunately, now that there is
486 the possibility of (*THEN) occurring in these branches, tail recursion is
487 no longer possible because the return has to be checked for (*THEN). These
488 two optimizations have therefore been removed. [But see 8.20/11 above.]
490 18. If a pattern containing \R was studied, it was assumed that \R always
491 matched two bytes, thus causing the minimum subject length to be
492 incorrectly computed because \R can also match just one byte.
494 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
495 was incorrectly computed.
497 20. If /S is present twice on a test pattern in pcretest input, it now
498 *disables* studying, thereby overriding the use of -s on the command line
499 (see 14 above). This is necessary for one or two tests to keep the output
500 identical in both cases.
502 21. When (*ACCEPT) was used in an assertion that matched an empty string and
503 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
505 22. When an atomic group that contained a capturing parenthesis was
506 successfully matched, but the branch in which it appeared failed, the
507 capturing was not being forgotten if a higher numbered group was later
508 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
509 group 1 to "a", when in fact it should be unset. This applied to multi-
510 branched capturing and non-capturing groups, repeated or not, and also to
511 positive assertions (capturing in negative assertions does not happen
512 in PCRE) and also to nested atomic groups.
514 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
515 subject after a captured substring, to make it easier to tell which of a
516 number of identical substrings has been captured.
518 24. The way atomic groups are processed by pcre_exec() has been changed so that
519 if they are repeated, backtracking one repetition now resets captured
520 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
521 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
522 "aaa". Previously, it would have been "a". As part of this code
523 refactoring, the way recursive calls are handled has also been changed.
525 25. If an assertion condition captured any substrings, they were not passed
526 back unless some other capturing happened later. For example, if
527 (?(?=(a))a) was matched against "a", no capturing was returned.
529 26. When studying a pattern that contained subroutine calls or assertions,
530 the code for finding the minimum length of a possible match was handling
531 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
532 group 1 called group 2 while simultaneously a separate group 2 called group
533 1). A stack overflow occurred in this case. I have fixed this by limiting
534 the recursion depth to 10.
536 27. Updated RunTest.bat in the distribution to the version supplied by Tom
537 Fortmann. This supports explicit test numbers on the command line, and has
538 argument validation and error reporting.
540 28. An instance of \X with an unlimited repeat could fail if at any point the
541 first character it looked at was a mark character.
543 29. Some minor code refactoring concerning Unicode properties and scripts
544 should reduce the stack requirement of match() slightly.
546 30. Added the '=' option to pcretest to check the setting of unused capturing
547 slots at the end of the pattern, which are documented as being -1, but are
548 not included in the return count.
550 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
551 compiled something random. Now it gives a compile-time error (as does
552 Perl).
554 32. A *MARK encountered during the processing of a positive assertion is now
555 recorded and passed back (compatible with Perl).
557 33. If --only-matching or --colour was set on a pcregrep call whose pattern
558 had alternative anchored branches, the search for a second match in a line
559 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
560 matched the line "0102" twice. The same bug affected patterns that started
561 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
562 twice.
564 34. Previously, PCRE did not allow quantification of assertions. However, Perl
565 does, and because of capturing effects, quantifying parenthesized
566 assertions may at times be useful. Quantifiers are now allowed for
567 parenthesized assertions.
569 35. A minor code tidy in pcre_compile() when checking options for \R usage.
571 36. \g was being checked for fancy things in a character class, when it should
572 just be a literal "g".
574 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
575 appearance of a nested POSIX class supersedes an apparent external class.
576 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
577 unescaped square brackets may also appear as part of class names. For
578 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
579 more like Perl. (But see 8.20/1 above.)
581 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
582 was because it thought it was \N{name}, which is not supported).
584 39. Add minix to OS list not supporting the -S option in pcretest.
586 40. PCRE tries to detect cases of infinite recursion at compile time, but it
587 cannot analyze patterns in sufficient detail to catch mutual recursions
588 such as ((?1))((?2)). There is now a runtime test that gives an error if a
589 subgroup is called recursively as a subpattern for a second time at the
590 same position in the subject string. In previous releases this might have
591 been caught by the recursion limit, or it might have run out of stack.
593 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
594 happen only once. PCRE was, however incorrectly giving a compile time error
595 "recursive call could loop indefinitely" because it cannot analyze the
596 pattern in sufficient detail. The compile time test no longer happens when
597 PCRE is compiling a conditional subpattern, but actual runaway loops are
598 now caught at runtime (see 40 above).
600 42. It seems that Perl allows any characters other than a closing parenthesis
601 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
602 has been changed to be the same.
604 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
605 as not to get warnings when autogen.sh is called. Also changed
606 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
608 44. To help people who use pcregrep to scan files containing exceedingly long
609 lines, the following changes have been made:
611 (a) The default value of the buffer size parameter has been increased from
612 8K to 20K. (The actual buffer used is three times this size.)
614 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
615 PCRE is built.
617 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
618 to be set at run time.
620 (d) Numerical values in pcregrep options can be followed by K or M, for
621 example --buffer-size=50K.
623 (e) If a line being scanned overflows pcregrep's buffer, an error is now
624 given and the return code is set to 2.
626 45. Add a pointer to the latest mark to the callout data block.
628 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
629 partial match of an empty string instead of no match. This was specific to
630 the use of ".".
632 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
633 complete match instead of a partial match. This bug was dependent on both
634 the PCRE_UTF8 and PCRE_DOTALL options being set.
636 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
637 starting byte set, because \b was not being ignored.
640 Version 8.12 15-Jan-2011
641 ------------------------
643 1. Fixed some typos in the markup of the man pages, and wrote a script that
644 checks for such things as part of the documentation building process.
646 2. On a big-endian 64-bit system, pcregrep did not correctly process the
647 --match-limit and --recursion-limit options (added for 8.11). In
648 particular, this made one of the standard tests fail. (The integer value
649 went into the wrong half of a long int.)
651 3. If the --colour option was given to pcregrep with -v (invert match), it
652 did strange things, either producing crazy output, or crashing. It should,
653 of course, ignore a request for colour when reporting lines that do not
654 match.
656 4. Another pcregrep bug caused similar problems if --colour was specified with
657 -M (multiline) and the pattern match finished with a line ending.
659 5. In pcregrep, when a pattern that ended with a literal newline sequence was
660 matched in multiline mode, the following line was shown as part of the
661 match. This seems wrong, so I have changed it.
663 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
664 the check for further matches in the same line (so they could be coloured)
665 to overrun the end of the current line. If another match was found, it was
666 incorrectly shown (and then shown again when found in the next line).
668 7. If pcregrep was compiled under Windows, there was a reference to the
669 function pcregrep_exit() before it was defined. I am assuming this was
670 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
671 reported by a user. I've moved the definition above the reference.
674 Version 8.11 10-Dec-2010
675 ------------------------
677 1. (*THEN) was not working properly if there were untried alternatives prior
678 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
679 backtracked to try for "b" instead of moving to the next alternative branch
680 at the same level (in this case, to look for "c"). The Perl documentation
681 is clear that when (*THEN) is backtracked onto, it goes to the "next
682 alternative in the innermost enclosing group".
684 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
685 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
686 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
687 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
688 (*THEN).
690 3. If \s appeared in a character class, it removed the VT character from
691 the class, even if it had been included by some previous item, for example
692 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
693 of \s, but is part of the POSIX "space" class.)
695 4. A partial match never returns an empty string (because you can always
696 match an empty string at the end of the subject); however the checking for
697 an empty string was starting at the "start of match" point. This has been
698 changed to the "earliest inspected character" point, because the returned
699 data for a partial match starts at this character. This means that, for
700 example, /(?<=abc)def/ gives a partial match for the subject "abc"
701 (previously it gave "no match").
703 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
704 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
705 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
706 has an implication that the given string is incomplete (because a partial
707 match is preferred over a full match). For this reason, these items now
708 give a partial match in this situation. [Aside: previously, the one case
709 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
710 match rather than a full match, which was wrong by the old rules, but is
711 now correct.]
713 6. There was a bug in the handling of #-introduced comments, recognized when
714 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
715 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
716 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
717 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
718 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
719 places in pcre_compile().
721 7. Related to (6) above, when pcre_compile() was skipping #-introduced
722 comments when looking ahead for named forward references to subpatterns,
723 the only newline sequence it recognized was NL. It now handles newlines
724 according to the set newline convention.
726 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
727 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
728 cater for a lack of strerror(). These oversights have been fixed.
730 9. Added --match-limit and --recursion-limit to pcregrep.
732 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
734 11. When the -o option was used, pcregrep was setting a return code of 1, even
735 when matches were found, and --line-buffered was not being honoured.
737 12. Added an optional parentheses number to the -o and --only-matching options
738 of pcregrep.
740 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
741 can match an empty string. The code to do it in pcretest and pcredemo
742 needed fixing:
744 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
745 only one byte after an empty string match just before CRLF (this case
746 just got forgotten; "any" and "anycrlf" were OK).
748 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
749 mode when an empty string match preceded an ASCII character followed by
750 a non-ASCII character. (The code for advancing by one character rather
751 than one byte was nonsense.)
753 (c) The pcredemo.c sample program did not have any code at all to handle
754 the cases when CRLF is a valid newline sequence.
756 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
757 as a starting offset was within the subject string. There is now a new
758 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
759 negative or greater than the length of the string. In order to test this,
760 pcretest is extended to allow the setting of negative starting offsets.
762 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
763 starting offset points to the beginning of a UTF-8 character was
764 unnecessarily clumsy. I tidied it up.
766 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
767 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
769 17. Nobody had reported that the --include_dir option, which was added in
770 release 7.7 should have been called --include-dir (hyphen, not underscore)
771 for compatibility with GNU grep. I have changed it to --include-dir, but
772 left --include_dir as an undocumented synonym, and the same for
773 --exclude-dir, though that is not available in GNU grep, at least as of
774 release 2.5.4.
776 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
777 characters from a string of bytes) have been redefined so as not to use
778 loops, in order to improve performance in some environments. At the same
779 time, I abstracted some of the common code into auxiliary macros to save
780 repetition (this should not affect the compiled code).
782 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
783 compile-time error is now given if \c is not followed by an ASCII
784 character, that is, a byte less than 128. (In EBCDIC mode, the code is
785 different, and any byte value is allowed.)
787 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
788 START_OPTIMIZE option, which is now allowed at compile time - but just
789 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
790 to pcregrep and other applications that have no direct access to PCRE
791 options. The new /Y option in pcretest sets this option when calling
792 pcre_compile().
794 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
795 back references. Groups containing recursive back references were forced to
796 be atomic by that change, but in the case of named groups, the amount of
797 memory required was incorrectly computed, leading to "Failed: internal
798 error: code overflow". This has been fixed.
800 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
801 pcretest.c, to avoid build problems in some Borland environments.
804 Version 8.10 25-Jun-2010
805 ------------------------
807 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
808 THEN.
810 2. (*ACCEPT) was not working when inside an atomic group.
812 3. Inside a character class, \B is treated as a literal by default, but
813 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
814 causes the error). The code is unchanged, but I tidied the documentation.
816 4. Inside a character class, PCRE always treated \R and \X as literals,
817 whereas Perl faults them if its -w option is set. I have changed PCRE so
818 that it faults them when PCRE_EXTRA is set.
820 5. Added support for \N, which always matches any character other than
821 newline. (It is the same as "." when PCRE_DOTALL is not set.)
823 6. When compiling pcregrep with newer versions of gcc which may have
824 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
825 declared with attribute warn_unused_result" were given. Just casting the
826 result to (void) does not stop the warnings; a more elaborate fudge is
827 needed. I've used a macro to implement this.
829 7. Minor change to pcretest.c to avoid a compiler warning.
831 8. Added four artifical Unicode properties to help with an option to make
832 \s etc use properties (see next item). The new properties are: Xan
833 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
835 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
836 use Unicode properties. (*UCP) at the start of a pattern can be used to set
837 this option. Modified pcretest to add /W to test this facility. Added
838 REG_UCP to make it available via the POSIX interface.
840 10. Added --line-buffered to pcregrep.
842 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
843 studied, and the match started with a letter with a code point greater than
844 127 whose first byte was different to the first byte of the other case of
845 the letter, the other case of this starting letter was not recognized
846 (#976).
848 12. If a pattern that was studied started with a repeated Unicode property
849 test, for example, \p{Nd}+, there was the theoretical possibility of
850 setting up an incorrect bitmap of starting bytes, but fortunately it could
851 not have actually happened in practice until change 8 above was made (it
852 added property types that matched character-matching opcodes).
854 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
855 possible starting bytes for non-anchored patterns.
857 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
858 \R, and also a number of cases that involve Unicode properties, both
859 explicit and implicit when PCRE_UCP is set.
861 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
862 input, it could crash or give wrong results if characters with values
863 greater than 0xc0 were present in the subject string. (Detail: it assumed
864 UTF-8 input when processing these items.)
866 16. Added a lot of (int) casts to avoid compiler warnings in systems where
867 size_t is 64-bit (#991).
869 17. Added a check for running out of memory when PCRE is compiled with
870 --disable-stack-for-recursion (#990).
872 18. If the last data line in a file for pcretest does not have a newline on
873 the end, a newline was missing in the output.
875 19. The default pcre_chartables.c file recognizes only ASCII characters (values
876 less than 128) in its various bitmaps. However, there is a facility for
877 generating tables according to the current locale when PCRE is compiled. It
878 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
879 characters, are recognized by isspace() and therefore were getting set in
880 these tables, and indeed these tables seem to approximate to ISO 8859. This
881 caused a problem in UTF-8 mode when pcre_study() was used to create a list
882 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
883 which of course cannot start UTF-8 characters. I have changed the code so
884 that only real ASCII characters (less than 128) and the correct starting
885 bytes for UTF-8 encodings are set for characters greater than 127 when in
886 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
887 altogether.)
889 20. Added the /T option to pcretest so as to be able to run tests with non-
890 standard character tables, thus making it possible to include the tests
891 used for 19 above in the standard set of tests.
893 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
894 reference to a subpattern the other side of a comment that contains an
895 opening parenthesis caused either an internal compiling error, or a
896 reference to the wrong subpattern.
899 Version 8.02 19-Mar-2010
900 ------------------------
902 1. The Unicode data tables have been updated to Unicode 5.2.0.
904 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
905 configured.
907 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
908 original author of that file, following a query about its status.
910 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
911 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
913 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
914 quantifier applied to a forward-referencing subroutine call, could compile
915 incorrect code or give the error "internal error: previously-checked
916 referenced subpattern not found".
918 6. Both MS Visual Studio and Symbian OS have problems with initializing
919 variables to point to external functions. For these systems, therefore,
920 pcre_malloc etc. are now initialized to local functions that call the
921 relevant global functions.
923 7. There were two entries missing in the vectors called coptable and poptable
924 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
925 I've fixed the data, and added a kludgy way of testing at compile time that
926 the lengths are correct (equal to the number of opcodes).
928 8. Following on from 7, I added a similar kludge to check the length of the
929 eint vector in pcreposix.c.
931 9. Error texts for pcre_compile() are held as one long string to avoid too
932 much relocation at load time. To find a text, the string is searched,
933 counting zeros. There was no check for running off the end of the string,
934 which could happen if a new error number was added without updating the
935 string.
937 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
939 11. \K was not working if it appeared in an atomic group or in a group that
940 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
941 \K is "not well defined" if used in an assertion. PCRE now accepts it if
942 the assertion is positive, but not if it is negative.
944 12. Change 11 fortuitously reduced the size of the stack frame used in the
945 "match()" function of pcre_exec.c by one pointer. Forthcoming
946 implementation of support for (*MARK) will need an extra pointer on the
947 stack; I have reserved it now, so that the stack frame size does not
948 decrease.
950 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
951 item in branch that calls a recursion is a subroutine call - as in the
952 second branch in the above example - was incorrectly given the compile-
953 time error "recursive call could loop indefinitely" because pcre_compile()
954 was not correctly checking the subroutine for matching a non-empty string.
956 14. The checks for overrunning compiling workspace could trigger after an
957 overrun had occurred. This is a "should never occur" error, but it can be
958 triggered by pathological patterns such as hundreds of nested parentheses.
959 The checks now trigger 100 bytes before the end of the workspace.
961 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
964 Version 8.01 19-Jan-2010
965 ------------------------
967 1. If a pattern contained a conditional subpattern with only one branch (in
968 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
969 computed the wrong minimum data length (which is of course zero for such
970 subpatterns). This could cause incorrect "no match" results.
972 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
973 the pattern is reset in the first branch, pcre_compile() failed with
974 "internal error: code overflow at offset...". This happened only when
975 the reset was to the original external option setting. (An optimization
976 abstracts leading options settings into an external setting, which was the
977 cause of this.)
979 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
980 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
981 assertion pattern did not match (meaning that the assertion was true), it
982 was incorrectly treated as false if the SKIP had been reached during the
983 matching. This also applied to assertions used as conditions.
985 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
986 assertion subpattern, including such a pattern used as a condition,
987 unpredictable results occurred, instead of the error return
990 5. The C++ GlobalReplace function was not working like Perl for the special
991 situation when an empty string is matched. It now does the fancy magic
992 stuff that is necessary.
994 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
995 removed. (These were left over from very, very early versions of PCRE.)
997 7. Some cosmetic changes to the code to make life easier when compiling it
998 as part of something else:
1000 (a) Change DEBUG to PCRE_DEBUG.
1002 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1003 called "current" as "current_branch", to prevent a collision with the
1004 Linux macro when compiled as a kernel module.
1006 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1007 prevent a collision with the Linux macro when compiled as a kernel
1008 module.
1010 8. In pcre_compile() there are some checks for integer overflows that used to
1011 cast potentially large values to (double). This has been changed to that
1012 when building, a check for int64_t is made, and if it is found, it is used
1013 instead, thus avoiding the use of floating point arithmetic. (There is no
1014 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1015 double.
1017 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1018 2005 (difference between two addresses compared to an unsigned value).
1020 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1021 custom one, because of the following reported problem in Windows:
1023 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1024 under Win32.
1025 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1026 therefore missing the function definition.
1027 - The compiler thus generates a "C" signature for the test function.
1028 - The linker fails to find the "C" function.
1029 - PCRE fails to configure if asked to do so against libbz2.
1031 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1032 messages were output:
1034 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1035 rerunning libtoolize, to keep the correct libtool macros in-tree.
1036 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1038 I have done both of these things.
1040 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1041 most of the time, it *can* run out if it is given a pattern that contains a
1042 runaway infinite recursion. I updated the discussion in the pcrestack man
1043 page.
1045 13. Now that we have gone to the x.xx style of version numbers, the minor
1046 version may start with zero. Using 08 or 09 is a bad idea because users
1047 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1048 interpreted as invalid octal numbers. I've updated the previous comment in
1049 configure.ac, and also added a check that gives an error if 08 or 09 are
1050 used.
1052 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1053 causing partial matching to fail when the end of the subject matched \W
1054 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1056 15. There were some discrepancies between the declarations in pcre_internal.h
1057 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1058 their definitions. The declarations used "const uschar *" and the
1059 definitions used USPTR. Even though USPTR is normally defined as "const
1060 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1061 reported that: "This difference in casting confuses some C++ compilers, for
1062 example, SunCC recognizes above declarations as different functions and
1063 generates broken code for hbpcre." I have changed the declarations to use
1064 USPTR.
1066 16. GNU libtool is named differently on some systems. The autogen.sh script now
1067 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1068 (FreeBSD).
1070 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1071 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1072 comment: "Figure out how to create a longlong from a string: strtoll and
1073 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1074 instance, but it only takes 2 args instead of 3!"
1076 18. A subtle bug concerned with back references has been fixed by a change of
1077 specification, with a corresponding code fix. A pattern such as
1078 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1079 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1080 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1081 same bug. Such groups have to be quantified to be useful, or contained
1082 inside another quantified group. (If there's no repetition, the reference
1083 can never match.) The problem arises because, having left the group and
1084 moved on to the rest of the pattern, a later failure that backtracks into
1085 the group uses the captured value from the final iteration of the group
1086 rather than the correct earlier one. I have fixed this in PCRE by forcing
1087 any group that contains a reference to itself to be an atomic group; that
1088 is, there cannot be any backtracking into it once it has completed. This is
1089 similar to recursive and subroutine calls.
1092 Version 8.00 19-Oct-09
1093 ----------------------
1095 1. The table for translating pcre_compile() error codes into POSIX error codes
1096 was out-of-date, and there was no check on the pcre_compile() error code
1097 being within the table. This could lead to an OK return being given in
1098 error.
1100 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1101 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1102 in a Windows environment.
1104 3. The pcregrep --count option prints the count for each file even when it is
1105 zero, as does GNU grep. However, pcregrep was also printing all files when
1106 --files-with-matches was added. Now, when both options are given, it prints
1107 counts only for those files that have at least one match. (GNU grep just
1108 prints the file name in this circumstance, but including the count seems
1109 more useful - otherwise, why use --count?) Also ensured that the
1110 combination -clh just lists non-zero counts, with no names.
1112 4. The long form of the pcregrep -F option was incorrectly implemented as
1113 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1114 but it seems right to fix it, and I didn't think it was worth preserving
1115 the old behaviour.
1117 5. The command line items --regex=pattern and --regexp=pattern were not
1118 recognized by pcregrep, which required --regex pattern or --regexp pattern
1119 (with a space rather than an '='). The man page documented the '=' forms,
1120 which are compatible with GNU grep; these now work.
1122 6. No libpcreposix.pc file was created for pkg-config; there was just
1123 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1125 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1126 when UCP support is not needed, by modifying the Python script that
1127 generates it from Unicode data files. This should not matter if the module
1128 is correctly used as a library, but I received one complaint about 50K of
1129 unwanted data. My guess is that the person linked everything into his
1130 program rather than using a library. Anyway, it does no harm.
1132 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1133 was a minimum greater than 1 for a wide character in a possessive
1134 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1135 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1136 character. Chaos in the form of incorrect output or a compiling loop could
1137 result.
1139 9. The restrictions on what a pattern can contain when partial matching is
1140 requested for pcre_exec() have been removed. All patterns can now be
1141 partially matched by this function. In addition, if there are at least two
1142 slots in the offset vector, the offset of the earliest inspected character
1143 for the match and the offset of the end of the subject are set in them when
1144 PCRE_ERROR_PARTIAL is returned.
1146 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1147 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1148 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1149 and may be more useful for multi-segment matching.
1151 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1152 used to be given if ever the end of the subject was reached; now it is
1153 given only if matching could not proceed because another character was
1154 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1155 string "Z", which now yields "no match" instead of "partial match". In the
1156 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1157 final character ended with (*FAIL).
1159 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1160 if the pattern had a "must contain" character that was already found in the
1161 earlier partial match, unless partial matching was again requested. For
1162 example, with the pattern /dog.(body)?/, the "must contain" character is
1163 "g". If the first part-match was for the string "dog", restarting with
1164 "sbody" failed. This bug has been fixed.
1166 13. The string returned by pcre_dfa_exec() after a partial match has been
1167 changed so that it starts at the first inspected character rather than the
1168 first character of the match. This makes a difference only if the pattern
1169 starts with a lookbehind assertion or \b or \B (\K is not supported by
1170 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1171 matching functions compatible, and I think it's the right thing to do.
1173 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1174 so that the demonstration program is easily available in environments where
1175 PCRE has not been installed from source.
1177 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1178 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1179 library.
1181 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1182 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1183 is not the first non-POSIX option to be added. Clearly some people find
1184 these options useful.
1186 17. If a caller to the POSIX matching function regexec() passes a non-zero
1187 value for nmatch with a NULL value for pmatch, the value of
1188 nmatch is forced to zero.
1190 18. RunGrepTest did not have a test for the availability of the -u option of
1191 the diff command, as RunTest does. It now checks in the same way as
1192 RunTest, and also checks for the -b option.
1194 19. If an odd number of negated classes containing just a single character
1195 interposed, within parentheses, between a forward reference to a named
1196 subpattern and the definition of the subpattern, compilation crashed with
1197 an internal error, complaining that it could not find the referenced
1198 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1199 [The bug was that it was starting one character too far in when skipping
1200 over the character class, thus treating the ] as data rather than
1201 terminating the class. This meant it could skip too much.]
1203 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1204 /g option in pcretest when the pattern contains \K, which makes it possible
1205 to have an empty string match not at the start, even when the pattern is
1206 anchored. Updated pcretest and pcredemo to use this option.
1208 21. If the maximum number of capturing subpatterns in a recursion was greater
1209 than the maximum at the outer level, the higher number was returned, but
1210 with unset values at the outer level. The correct (outer level) value is
1211 now given.
1213 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1214 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1215 make it do so. The string so far is captured, making this feature
1216 compatible with Perl.
1218 23. The tests have been re-organized, adding tests 11 and 12, to make it
1219 possible to check the Perl 5.10 features against Perl 5.10.
1221 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1222 pattern matches a fixed length string. PCRE did not allow this; now it
1223 does. Neither allows recursion.
1225 25. I finally figured out how to implement a request to provide the minimum
1226 length of subject string that was needed in order to match a given pattern.
1227 (It was back references and recursion that I had previously got hung up
1228 on.) This code has now been added to pcre_study(); it finds a lower bound
1229 to the length of subject needed. It is not necessarily the greatest lower
1230 bound, but using it to avoid searching strings that are too short does give
1231 some useful speed-ups. The value is available to calling programs via
1232 pcre_fullinfo().
1234 26. While implementing 25, I discovered to my embarrassment that pcretest had
1235 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1236 study optimizations had never been tested with that matching function.
1237 Oops. What is worse, even when it was passed study data, there was a bug in
1238 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1239 were also very few tests of studied patterns with pcre_dfa_exec().
1241 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1242 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1243 on the other side of the coin, they are no longer allowed to have different
1244 names, because these cannot be distinguished in PCRE, and this has caused
1245 confusion. (This is a difference from Perl.)
1247 28. When duplicate subpattern names are present (necessarily with different
1248 numbers, as required by 27 above), and a test is made by name in a
1249 conditional pattern, either for a subpattern having been matched, or for
1250 recursion in such a pattern, all the associated numbered subpatterns are
1251 tested, and the overall condition is true if the condition is true for any
1252 one of them. This is the way Perl works, and is also more like the way
1253 testing by number works.
1256 Version 7.9 11-Apr-09
1257 ---------------------
1259 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1260 (pcretest), all targets were linked against these libraries. This included
1261 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1262 libraries. This caused unwanted dependencies to be created. This problem
1263 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1264 pcretest is linked with readline.
1266 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1267 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1268 moved outside it again, because FALSE and TRUE are already defined in AIX,
1269 but BOOL is not.
1271 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1272 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1274 4. The pcregrep documentation said spaces were inserted as well as colons (or
1275 hyphens) following file names and line numbers when outputting matching
1276 lines. This is not true; no spaces are inserted. I have also clarified the
1277 wording for the --colour (or --color) option.
1279 5. In pcregrep, when --colour was used with -o, the list of matching strings
1280 was not coloured; this is different to GNU grep, so I have changed it to be
1281 the same.
1283 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1284 each matching line was coloured. Now it goes on to look for further matches
1285 of any of the test patterns, which is the same behaviour as GNU grep.
1287 7. A pattern that could match an empty string could cause pcregrep to loop; it
1288 doesn't make sense to accept an empty string match in pcregrep, so I have
1289 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1290 seems to be how GNU grep behaves.
1292 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1293 start or after a newline", because the conditional assertion was not being
1294 correctly handled. The rule now is that both the assertion and what follows
1295 in the first alternative must satisfy the test.
1297 9. If auto-callout was enabled in a pattern with a conditional group whose
1298 condition was an assertion, PCRE could crash during matching, both with
1299 pcre_exec() and pcre_dfa_exec().
1301 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1302 used for matching.
1304 11. Unicode property support in character classes was not working for
1305 characters (bytes) greater than 127 when not in UTF-8 mode.
1307 12. Added the -M command line option to pcretest.
1309 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1311 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1313 16. Added comments and documentation about mis-use of no_arg in the C++
1314 wrapper.
1316 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1317 from Martin Jerabek that uses macro names for all relevant character and
1318 string constants.
1320 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1321 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1322 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1323 these, but not everybody uses configure.
1325 19. A conditional group that had only one branch was not being correctly
1326 recognized as an item that could match an empty string. This meant that an
1327 enclosing group might also not be so recognized, causing infinite looping
1328 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1329 with the subject "ab", where knowledge that the repeated group can match
1330 nothing is needed in order to break the loop.
1332 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1333 exec(), but without supplying a callout function, matching went wrong.
1335 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1336 leak if the size of the offset vector was greater than 30. When the vector
1337 is smaller, the saved offsets during recursion go onto a local stack
1338 vector, but for larger vectors malloc() is used. It was failing to free
1339 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1340 error, in fact).
1342 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1343 heapframe that is used only when UTF-8 support is enabled. This caused no
1344 problem, but was untidy.
1346 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1347 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1348 included within another project.
1350 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1351 slightly modified by me:
1353 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1354 not building pcregrep.
1356 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1357 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1359 25. Forward references, both numeric and by name, in patterns that made use of
1360 duplicate group numbers, could behave incorrectly or give incorrect errors,
1361 because when scanning forward to find the reference group, PCRE was not
1362 taking into account the duplicate group numbers. A pattern such as
1363 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1365 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1366 the feature of a custom pointer more persuasive (as requested by a user).
1368 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1369 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1370 pre-defined.
1372 28. Added support for (*UTF8) at the start of a pattern.
1374 29. Arrange for flags added by the "release type" setting in CMake to be shown
1375 in the configuration summary.
1378 Version 7.8 05-Sep-08
1379 ---------------------
1381 1. Replaced UCP searching code with optimized version as implemented for Ad
1382 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1383 stage table and inline lookup instead of a function, giving speed ups of 2
1384 to 5 times on some simple patterns that I tested. Permission was given to
1385 distribute the MultiStage2.py script that generates the tables (it's not in
1386 the tarball, but is in the Subversion repository).
1388 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1389 scripts.
1391 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1392 a group with a zero qualifier. The result of the study could be incorrect,
1393 or the function might crash, depending on the pattern.
1395 4. Caseless matching was not working for non-ASCII characters in back
1396 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1397 It now works when Unicode Property Support is available.
1399 5. In pcretest, an escape such as \x{de} in the data was always generating
1400 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1401 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1402 truncation.
1404 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1406 7. Added two (int) casts to pcregrep when printing the difference of two
1407 pointers, in case they are 64-bit values.
1409 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1410 test 2 if it fails.
1412 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1413 and a #define of that name to empty if it is not externally set. This is to
1414 allow users of MSVC to set it if necessary.
1416 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1417 the convenience functions in the pcre_get.c source file.
1419 11. An option change at the start of a pattern that had top-level alternatives
1420 could cause overwriting and/or a crash. This command provoked a crash in
1421 some environments:
1423 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1425 This potential security problem was recorded as CVE-2008-2371.
1427 12. For a pattern where the match had to start at the beginning or immediately
1428 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1429 pcre_dfa_exec() could read past the end of the passed subject if there was
1430 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1431 pcretest so that it places the subject at the end of its malloc-ed buffer.
1433 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1434 exec() might read past the end of the data buffer in UTF-8 mode.
1436 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1437 the data contained the byte 0x85 as part of a UTF-8 character within its
1438 first line. This applied both to normal and DFA matching.
1440 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1441 /^[^d]*?$/8 failed to match "abc".
1443 16. Added a missing copyright notice to pcrecpp_internal.h.
1445 17. Make it more clear in the documentation that values returned from
1446 pcre_exec() in ovector are byte offsets, not character counts.
1448 18. Tidied a few places to stop certain compilers from issuing warnings.
1450 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1451 supplied by Stefan Weber. I made a further small update for 7.8 because
1452 there is a change of source arrangements: the pcre_searchfuncs.c module is
1453 replaced by pcre_ucd.c.
1456 Version 7.7 07-May-08
1457 ---------------------
1459 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1460 a string to a long long, pretend we don't even have a long long." This is
1461 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1463 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1464 pre-7.6 versions, which defined a global no_arg variable instead of putting
1465 it in the RE class. (See also #8 below.)
1467 3. Remove a line of dead code, identified by coverity and reported by Nuno
1468 Lopes.
1470 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1472 (1) The include/exclude patterns were being applied to the whole pathnames
1473 of files, instead of just to the final components.
1475 (2) If there was more than one level of directory, the subdirectories were
1476 skipped unless they satisfied the include/exclude conditions. This is
1477 inconsistent with GNU grep (and could even be seen as contrary to the
1478 pcregrep specification - which I improved to make it absolutely clear).
1479 The action now is always to scan all levels of directory, and just
1480 apply the include/exclude patterns to regular files.
1482 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1483 --exclude_dir in the tests to avoid scanning .svn directories.
1485 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1486 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1487 doesn't support NULs in patterns.
1489 7. Added some missing "const"s to declarations of static tables in
1490 pcre_compile.c and pcre_dfa_exec.c.
1492 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1493 caused by fix #2 above. (Subsequently also a second patch to fix the
1494 first patch. And a third patch - this was a messy problem.)
1496 9. Applied Craig's patch to remove the use of push_back().
1498 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1499 matching function regexec().
1501 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1502 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1503 references. PCRE supports relative numbers with this syntax (I don't think
1504 Oniguruma does).
1506 12. Previously, a group with a zero repeat such as (...){0} was completely
1507 omitted from the compiled regex. However, this means that if the group
1508 was called as a subroutine from elsewhere in the pattern, things went wrong
1509 (an internal error was given). Such groups are now left in the compiled
1510 pattern, with a new opcode that causes them to be skipped at execution
1511 time.
1513 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1514 to the way PCRE behaves:
1516 (a) A lone ] character is dis-allowed (Perl treats it as data).
1518 (b) A back reference to an unmatched subpattern matches an empty string
1519 (Perl fails the current match path).
1521 (c) A data ] in a character class must be notated as \] because if the
1522 first data character in a class is ], it defines an empty class. (In
1523 Perl it is not possible to have an empty class.) The empty class []
1524 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1525 The negative empty class [^] matches any one character, independently
1526 of the DOTALL setting.
1528 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1529 non-existent subpattern following a character class starting with ']' and
1530 containing () gave an internal compiling error instead of "reference to
1531 non-existent subpattern". Fortunately, when the pattern did exist, the
1532 compiled code was correct. (When scanning forwards to check for the
1533 existencd of the subpattern, it was treating the data ']' as terminating
1534 the class, so got the count wrong. When actually compiling, the reference
1535 was subsequently set up correctly.)
1537 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1538 it was being rejected as not supported by pcre_dfa_exec(), even though
1539 other assertions are supported. I have made pcre_dfa_exec() support
1540 (*FAIL).
1542 16. The implementation of 13c above involved the invention of a new opcode,
1543 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1544 cannot be changed at match time, I realized I could make a small
1545 improvement to matching performance by compiling OP_ALLANY instead of
1546 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1547 on the OP_ANY path.
1549 17. Compiling pcretest on Windows with readline support failed without the
1550 following two fixes: (1) Make the unistd.h include conditional on
1551 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1553 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1554 ncurses library to be included for pcretest when ReadLine support is
1555 requested, but also to allow for it to be overridden. This patch came from
1556 Daniel Bergström.
1558 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1559 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1560 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1561 spotting this.
1564 Version 7.6 28-Jan-08
1565 ---------------------
1567 1. A character class containing a very large number of characters with
1568 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1569 overflow.
1571 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1572 HAVE_LONG_LONG is not defined.
1574 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1575 bring them up to date and include new features. This patch includes:
1577 - Fixed PH's badly added libz and libbz2 support.
1578 - Fixed a problem with static linking.
1579 - Added pcredemo. [But later removed - see 7 below.]
1580 - Fixed dftables problem and added an option.
1581 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1583 - Added readline support for pcretest.
1584 - Added an listing of the option settings after cmake has run.
1586 4. A user submitted a patch to Makefile that makes it easy to create
1587 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1588 Makefile.am that cause it to include this special target, without
1589 affecting anything else. Note that the same mingw target plus all
1590 the other distribution libraries and programs are now supported
1591 when configuring with CMake (see 6 below) instead of with
1592 Configure/Make.
1594 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1595 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1596 exported in the Windows port". It has not yet been confirmed that the patch
1597 solves the problem, but it does no harm.
1599 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1600 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1601 with CMake, and also correct the comment about stack recursion.
1603 7. Remove the automatic building of pcredemo from the ./configure system and
1604 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1605 of a program that users should build themselves after PCRE is installed, so
1606 building it automatically is not really right. What is more, it gave
1607 trouble in some build environments.
1609 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1612 Version 7.5 10-Jan-08
1613 ---------------------
1615 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1616 values in parens when parsing an RE using the C++ wrapper."
1618 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1619 Characters greater than 255 were excluded from the class instead of being
1620 included.
1622 3. The same bug as (2) above applied to negated POSIX classes such as
1623 [:^space:].
1625 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1626 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1627 I have changed it.
1629 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1630 first named subpattern) and a construct such as (?&a) would reference the
1631 first named subpattern whose name started with "a" (in other words, the
1632 length check was missing). Both these problems are fixed. "Subpattern name
1633 expected" is now given for (?&) (a zero-length name), and this patch also
1634 makes it give the same error for \k'' (previously it complained that that
1635 was a reference to a non-existent subpattern).
1637 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1638 this is right because (?- can be followed by option settings as well as by
1639 digits. I have, however, made the messages clearer.
1641 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1642 than the number used in the conditional) now cause a compile-time error.
1643 This is actually not compatible with Perl, which accepts such patterns, but
1644 treats the conditional as always being FALSE (as PCRE used to), but it
1645 seems to me that giving a diagnostic is better.
1647 8. Change "alphameric" to the more common word "alphanumeric" in comments
1648 and messages.
1650 9. Fix two occurrences of "backslash" in comments that should have been
1651 "backspace".
1653 10. Remove two redundant lines of code that can never be obeyed (their function
1654 was moved elsewhere).
1656 11. The program that makes PCRE's Unicode character property table had a bug
1657 which caused it to generate incorrect table entries for sequences of
1658 characters that have the same character type, but are in different scripts.
1659 It amalgamated them into a single range, with the script of the first of
1660 them. In other words, some characters were in the wrong script. There were
1661 thirteen such cases, affecting characters in the following ranges:
1663 U+002b0 - U+002c1
1664 U+0060c - U+0060d
1665 U+0061e - U+00612
1666 U+0064b - U+0065e
1667 U+0074d - U+0076d
1668 U+01800 - U+01805
1669 U+01d00 - U+01d77
1670 U+01d9b - U+01dbf
1671 U+0200b - U+0200f
1672 U+030fc - U+030fe
1673 U+03260 - U+0327f
1674 U+0fb46 - U+0fbb1
1675 U+10450 - U+1049d
1677 12. The -o option (show only the matching part of a line) for pcregrep was not
1678 compatible with GNU grep in that, if there was more than one match in a
1679 line, it showed only the first of them. It now behaves in the same way as
1680 GNU grep.
1682 13. If the -o and -v options were combined for pcregrep, it printed a blank
1683 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1684 does the same. The return code can be used to tell if there were any
1685 non-matching lines.
1687 14. Added --file-offsets and --line-offsets to pcregrep.
1689 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1690 infinitely looping recursion. The bug was that positive lookaheads were not
1691 being skipped when checking for a possible empty match (negative lookaheads
1692 and both kinds of lookbehind were skipped).
1694 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1695 inclusion of <windows.h> to before rather than after the definition of
1696 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1698 17. Specifying a possessive quantifier with a specific limit for a Unicode
1699 character property caused pcre_compile() to compile bad code, which led at
1700 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1701 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1702 caused the error; without that there was no problem.
1704 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1706 19. Added --enable-pcretest-libreadline.
1708 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1709 RE::GlobalReplace(). As a result, the number of replacements returned was
1710 double what it should be. I removed one of the increments, but Craig sent a
1711 later patch that removed the other one (the right fix) and added unit tests
1712 that check the return values (which was not done before).
1714 21. Several CMake things:
1716 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1717 the names libpcre and libpcreposix, not just pcre and pcreposix.
1719 (2) The above change means that pcretest and pcregrep are now correctly
1720 linked with the newly-built libraries, not previously installed ones.
1724 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1725 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1726 UTF-8 newline character). The key issue is that the pattern starts .*;
1727 this means that the match must be either at the beginning, or after a
1728 newline. The bug was in the code for advancing after a failed match and
1729 checking that the new position followed a newline. It was not taking
1730 account of UTF-8 characters correctly.
1732 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1733 character classes. PCRE was not treating the sequence [:...:] as a
1734 character class unless the ... were all letters. Perl, however, seems to
1735 allow any characters between [: and :], though of course it rejects as
1736 unknown any "names" that contain non-letters, because all the known class
1737 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1738 for example, whereas PCRE did not - it did not recognize a POSIX character
1739 class. This seemed a bit dangerous, so the code has been changed to be
1740 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1741 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1742 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1743 Perl does, and where it didn't before.
1745 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1746 Windows environments %n is disabled by default.
1749 Version 7.4 21-Sep-07
1750 ---------------------
1752 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1753 means that a class such as [\s] counted as "explicit reference to CR or
1754 LF". That isn't really right - the whole point of the change was to try to
1755 help when there was an actual mention of one of the two characters. So now
1756 the change happens only if \r or \n (or a literal CR or LF) character is
1757 encountered.
1759 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1760 of both had grown to the point where there were only 3 bits left.
1761 Fortunately, there was spare space in the data structure, and so I have
1762 moved the internal flags into a new 16-bit field to free up more option
1763 bits.
1765 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1766 but did not set the internal JCHANGED flag - either of these is enough to
1767 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1768 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1769 start sets both bits.
1771 4. Added options (at build time, compile time, exec time) to change \R from
1772 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1774 5. doc/pcresyntax.html was missing from the distribution.
1776 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1777 compatibility, even though it is no longer used.
1779 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1780 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1781 windows.h file is present (where different names are used). [This was
1782 reversed later after testing - see 16 below.]
1784 8. Changed all #include <config.h> to #include "config.h". There were also
1785 some further <pcre.h> cases that I changed to "pcre.h".
1787 9. When pcregrep was used with the --colour option, it missed the line ending
1788 sequence off the lines that it output.
1790 10. It was pointed out to me that arrays of string pointers cause lots of
1791 relocations when a shared library is dynamically loaded. A technique of
1792 using a single long string with a table of offsets can drastically reduce
1793 these. I have refactored PCRE in four places to do this. The result is
1794 dramatic:
1796 Originally: 290
1797 After changing UCP table: 187
1798 After changing error message table: 43
1799 After changing table of "verbs" 36
1800 After changing table of Posix names 22
1802 Thanks to the folks working on Gregex for glib for this insight.
1804 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1805 unicode-properties was also set.
1807 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1809 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1810 checked only for CRLF.
1812 14. Added casts to pcretest.c to avoid compiler warnings.
1814 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1816 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1817 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1818 entirely. This removes changes made in 7 above.
1820 17. The CMake files have been updated, and there is now more information about
1821 building with CMake in the NON-UNIX-USE document.
1824 Version 7.3 28-Aug-07
1825 ---------------------
1827 1. In the rejigging of the build system that eventually resulted in 7.1, the
1828 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1829 brackets there is not right, since it causes compilers to look for an
1830 installed pcre.h, not the version that is in the source that is being
1831 compiled (which of course may be different). I have changed it back to:
1833 #include "pcre.h"
1835 I have a vague recollection that the change was concerned with compiling in
1836 different directories, but in the new build system, that is taken care of
1837 by the VPATH setting the Makefile.
1839 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1840 when the subject happened to end in the byte 0x85 (e.g. if the last
1841 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1842 characters but of course it shouldn't be taken as a newline when it is part
1843 of another character. The bug was that, for an unlimited repeat of . in
1844 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1845 characters when looking for a newline.
1847 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1849 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1850 in debug output.
1852 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1853 long printing in the pcrecpp unittest when running under MinGW.
1855 6. ESC_K was left out of the EBCDIC table.
1857 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1858 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1859 limit also applies to "virtual nesting" when a pattern is recursive, and in
1860 this case 1000 isn't so big. I have been able to remove this limit at the
1861 expense of backing off one optimization in certain circumstances. Normally,
1862 when pcre_exec() would call its internal match() function recursively and
1863 immediately return the result unconditionally, it uses a "tail recursion"
1864 feature to save stack. However, when a subpattern that can match an empty
1865 string has an unlimited repetition quantifier, it no longer makes this
1866 optimization. That gives it a stack frame in which to save the data for
1867 checking that an empty string has been matched. Previously this was taken
1868 from the 1000-entry workspace that had been reserved. So now there is no
1869 explicit limit, but more stack is used.
1871 8. Applied Daniel's patches to solve problems with the import/export magic
1872 syntax that is required for Windows, and which was going wrong for the
1873 pcreposix and pcrecpp parts of the library. These were overlooked when this
1874 problem was solved for the main library.
1876 9. There were some crude static tests to avoid integer overflow when computing
1877 the size of patterns that contain repeated groups with explicit upper
1878 limits. As the maximum quantifier is 65535, the maximum group length was
1879 set at 30,000 so that the product of these two numbers did not overflow a
1880 32-bit integer. However, it turns out that people want to use groups that
1881 are longer than 30,000 bytes (though not repeat them that many times).
1882 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1883 made it possible to implement the integer overflow checks in a much more
1884 dynamic way, which I have now done. The artificial limitation on group
1885 length has been removed - we now have only the limit on the total length of
1886 the compiled pattern, which depends on the LINK_SIZE setting.
1888 10. Fixed a bug in the documentation for get/copy named substring when
1889 duplicate names are permitted. If none of the named substrings are set, the
1890 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1891 empty string.
1893 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1894 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1895 because the ] is interpreted as the first data character and the
1896 terminating ] is not found. PCRE has been made compatible with Perl in this
1897 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1898 cause memory overwriting.
1900 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1901 string has been matched (to stop an infinite loop). It was not recognizing
1902 a conditional subpattern that could match an empty string if that
1903 subpattern was within another subpattern. For example, it looped when
1904 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1905 condition was not nested. This bug has been fixed.
1907 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1908 past the start of the subject in the presence of bytes with the top bit
1909 set, for example "\x8aBCD".
1911 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1912 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1914 14. Optimized (?!) to (*FAIL).
1916 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1917 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1918 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1919 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1920 does: it's just the validity check that is more restrictive.
1922 16. Inserted checks for integer overflows during escape sequence (backslash)
1923 processing, and also fixed erroneous offset values for syntax errors during
1924 backslash processing.
1926 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1927 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1929 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1930 caused an overrun.
1932 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1933 something other than just ASCII characters) inside a group that had an
1934 unlimited repeat caused a loop at compile time (while checking to see
1935 whether the group could match an empty string).
1937 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1938 [\P{Any}] did so. (Error in the code for printing property names.)
1940 21. An orphan \E inside a character class could cause a crash.
1942 22. A repeated capturing bracket such as (A)? could cause a wild memory
1943 reference during compilation.
1945 23. There are several functions in pcre_compile() that scan along a compiled
1946 expression for various reasons (e.g. to see if it's fixed length for look
1947 behind). There were bugs in these functions when a repeated \p or \P was
1948 present in the pattern. These operators have additional parameters compared
1949 with \d, etc, and these were not being taken into account when moving along
1950 the compiled data. Specifically:
1952 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1953 length.
1955 (b) An item such as \pL+ within a repeated group could cause crashes or
1956 loops.
1958 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1959 "reference to non-existent subpattern" error.
1961 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1963 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1964 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1966 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1968 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1969 character were causing crashes (broken optimization).
1971 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1972 \p or \P) caused a compile-time loop.
1974 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1975 break. For example, the unstudied pattern [\r\n]A does not match the string
1976 "\r\nA" because change 7.0/46 below moves the current point on by two
1977 characters after failing to match at the start. However, the pattern \nA
1978 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1979 the same is true. There doesn't seem any very clean way out of this, but
1980 what I have chosen to do makes the common cases work: PCRE now takes note
1981 of whether there can be an explicit match for \r or \n anywhere in the
1982 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1983 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1984 pattern has explicit CR or LF references.
1986 29. Added (*CR) etc for changing newline setting at start of pattern.
1989 Version 7.2 19-Jun-07
1990 ---------------------
1992 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1993 which is apparently normally available under Windows.
1995 2. Re-jig the pcregrep tests with different newline settings in an attempt
1996 to make them independent of the local environment's newline setting.
1998 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2000 4. Some of the "internals" tests were previously cut out when the link size
2001 was not 2, because the output contained actual offsets. The recent new
2002 "Z" feature of pcretest means that these can be cut out, making the tests
2003 usable with all link sizes.
2005 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2006 stack recursion. This gives a massive performance boost under BSD, but just
2007 a small improvement under Linux. However, it saves one field in the frame
2008 in all cases.
2010 6. Added more features from the forthcoming Perl 5.10:
2012 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2013 recursion call. It refers to the nth most recently opened parentheses.
2015 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2016 to be opened parentheses.
2018 (c) Conditions that refer to capturing parentheses can be specified
2019 relatively, for example, (?(-2)... or (?(+3)...
2021 (d) \K resets the start of the current match so that everything before
2022 is not part of it.
2024 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2026 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2027 reference syntax.
2029 (g) (?| introduces a group in which the numbering of parentheses in each
2030 alternative starts with the same number.
2032 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2034 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2037 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2038 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2039 for detecting groups that can match an empty string.
2041 9. A pattern with a very large number of alternatives (more than several
2042 hundred) was running out of internal workspace during the pre-compile
2043 phase, where pcre_compile() figures out how much memory will be needed. A
2044 bit of new cunning has reduced the workspace needed for groups with
2045 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2046 workspace instead of running out of the 4096 that are available.
2048 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2050 11. Applied patch from Google to remove an optimization that didn't quite work.
2051 The report of the bug said:
2053 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2054 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2055 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2057 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2058 it matched the wrong number of bytes.
2061 Version 7.1 24-Apr-07
2062 ---------------------
2064 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2065 that is more "standard", making use of automake and other Autotools. There
2066 is some re-arrangement of the files and adjustment of comments consequent
2067 on this.
2069 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2070 for recursive directory scanning broke on some systems because the files
2071 are not scanned in any specific order and on different systems the order
2072 was different. A call to "sort" has been inserted into RunGrepTest for the
2073 approprate test as a short-term fix. In the longer term there may be an
2074 alternative.
2076 3. I had an email from Eric Raymond about problems translating some of PCRE's
2077 man pages to HTML (despite the fact that I distribute HTML pages, some
2078 people do their own conversions for various reasons). The problems
2079 concerned the use of low-level troff macros .br and .in. I have therefore
2080 removed all such uses from the man pages (some were redundant, some could
2081 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2082 HTML has been updated to handle .nf/.fi and to complain if it encounters
2083 .br or .in.
2085 4. Updated comments in configure.ac that get placed in config.h.in and also
2086 arranged for config.h to be included in the distribution, with the name
2087 config.h.generic, for the benefit of those who have to compile without
2088 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2090 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2091 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2092 makevp.bat for the current PCRE, using the additional files
2093 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2095 6. A Windows user reported a minor discrepancy with test 2, which turned out
2096 to be caused by a trailing space on an input line that had got lost in his
2097 copy. The trailing space was an accident, so I've just removed it.
2099 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2100 that is needed.
2102 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2103 as "const" (a) because they are and (b) because it helps the PHP
2104 maintainers who have recently made a script to detect big data structures
2105 in the php code that should be moved to the .rodata section. I remembered
2106 to update Builducptable as well, so it won't revert if ucptable.h is ever
2107 re-created.
2109 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2110 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2112 support is not required. This saves 1.5-2K of code, which is important in
2113 some applications.
2115 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2116 so as not to refer to the tables, even though these functions will never be
2117 called when UTF-8 support is disabled. Otherwise there are problems with a
2118 shared library.
2120 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2122 (a) It was defining its arguments as char * instead of void *.
2124 (b) It was assuming that all moves were upwards in memory; this was true
2125 a long time ago when I wrote it, but is no longer the case.
2127 The emulated memove() is provided for those environments that have neither
2128 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2129 is clearly not the case, as these two bugs were recently reported.
2131 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2132 and Detrail to create the HTML documentation, the .txt form of the man
2133 pages, and it removes trailing spaces from listed files. It also creates
2134 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2135 case, it wraps all the #defines with #ifndefs. This script should be run
2136 before "make dist".
2138 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2139 with Unicode property support.
2141 (a) For a maximizing quantifier, if the two different cases of the
2142 character were of different lengths in their UTF-8 codings (there are
2143 some cases like this - I found 11), and the matching function had to
2144 back up over a mixture of the two cases, it incorrectly assumed they
2145 were both the same length.
2147 (b) When PCRE was configured to use the heap rather than the stack for
2148 recursion during matching, it was not correctly preserving the data for
2149 the other case of a UTF-8 character when checking ahead for a match
2150 while processing a minimizing repeat. If the check also involved
2151 matching a wide character, but failed, corruption could cause an
2152 erroneous result when trying to check for a repeat of the original
2153 character.
2155 13. Some tidying changes to the testing mechanism:
2157 (a) The RunTest script now detects the internal link size and whether there
2158 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2159 values substituted by "configure". (The RunGrepTest script already did
2160 this for UTF-8.) The configure.ac script no longer substitutes the
2161 relevant variables.
2163 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2164 with length and offset values. This means that the output is different
2165 for different internal link sizes. Test 2 is skipped for link sizes
2166 other than 2 because of this, bypassing the problem. Unfortunately,
2167 there was also a test in test 3 (the locale tests) that used /B and
2168 failed for link sizes other than 2. Rather than cut the whole test out,
2169 I have added a new /Z option to pcretest that replaces the length and
2170 offset values with spaces. This is now used to make test 3 independent
2171 of link size. (Test 2 will be tidied up later.)
2173 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2174 segmentation fault instead of returning the appropriate error message.
2176 15. In multiline mode when the newline sequence was set to "any", the pattern
2177 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2178 This doesn't seem right; it now treats the CRLF combination as the line
2179 ending, and so does not match in that case. It's only a pattern such as ^$
2180 that would hit this one: something like ^ABC$ would have failed after \r
2181 and then tried again after \r\n.
2183 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2184 in an attempt to make files that differ only in their line terminators
2185 compare equal. This works on Linux.
2187 17. Under certain error circumstances pcregrep might try to free random memory
2188 as it exited. This is now fixed, thanks to valgrind.
2190 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2191 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2192 was because its rules for how to advance for /g after matching an empty
2193 string at the end of a line did not allow for this case. They now check for
2194 it specially.
2196 20. pcretest is supposed to handle patterns and data of any length, by
2197 extending its buffers when necessary. It was getting this wrong when the
2198 buffer for a data line had to be extended.
2200 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2201 CRLF as a newline sequence.
2203 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2204 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2205 I have nevertheless tidied it up.
2207 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2209 24. Added a man page for pcre-config.
2212 Version 7.0 19-Dec-06
2213 ---------------------
2215 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2216 moving to gcc 4.1.1.
2218 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2219 sys/time.h, which is documented as needed for this function. It doesn't
2220 seem to matter on Linux, but it showed up on some releases of OS X.
2222 3. It seems that there are systems where bytes whose values are greater than
2223 127 match isprint() in the "C" locale. The "C" locale should be the
2224 default when a C program starts up. In most systems, only ASCII printing
2225 characters match isprint(). This difference caused the output from pcretest
2226 to vary, making some of the tests fail. I have changed pcretest so that:
2228 (a) When it is outputting text in the compiled version of a pattern, bytes
2229 other than 32-126 are always shown as hex escapes.
2231 (b) When it is outputting text that is a matched part of a subject string,
2232 it does the same, unless a different locale has been set for the match
2233 (using the /L modifier). In this case, it uses isprint() to decide.
2235 4. Fixed a major bug that caused incorrect computation of the amount of memory
2236 required for a compiled pattern when options that changed within the
2237 pattern affected the logic of the preliminary scan that determines the
2238 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2239 that the computed length was too small. The symptoms of this bug were
2240 either the PCRE error "internal error: code overflow" from pcre_compile(),
2241 or a glibc crash with a message such as "pcretest: free(): invalid next
2242 size (fast)". Examples of patterns that provoked this bug (shown in
2243 pcretest format) are:
2245 /(?-x: )/x
2246 /(?x)(?-x: \s*#\s*)/
2247 /((?i)[\x{c0}])/8
2248 /(?i:[\x{c0}])/8
2250 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2251 is now done differently.
2253 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2254 wrapper classes; (b) implement a new function in the C++ scanner that is
2255 more efficient than the old way of doing things because it avoids levels of
2256 recursion in the regex matching; (c) add a paragraph to the documentation
2257 for the FullMatch() function.
2259 6. The escape sequence \n was being treated as whatever was defined as
2260 "newline". Not only was this contrary to the documentation, which states
2261 that \n is character 10 (hex 0A), but it also went horribly wrong when
2262 "newline" was defined as CRLF. This has been fixed.
2264 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2265 was being set to -1 for the "end of line" case (supposedly a value that no
2266 character can have). Though this value is never used (the check for end of
2267 line is "zero bytes in current character"), it caused compiler complaints.
2268 I've changed it to 0xffffffff.
2270 8. In pcre_version.c, the version string was being built by a sequence of
2271 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2272 string (as it is for production releases) called a macro with an empty
2273 argument. The C standard says the result of this is undefined. The gcc
2274 compiler treats it as an empty string (which was what was wanted) but it is
2275 reported that Visual C gives an error. The source has been hacked around to
2276 avoid this problem.
2278 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2279 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2280 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2281 of them did).
2283 10. Originally, pcretest opened its input and output without "b"; then I was
2284 told that "b" was needed in some environments, so it was added for release
2285 5.0 to both the input and output. (It makes no difference on Unix-like
2286 systems.) Later I was told that it is wrong for the input on Windows. I've
2287 now abstracted the modes into two macros, to make it easier to fiddle with
2288 them, and removed "b" from the input mode under Windows.
2290 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2292 12. Added -help and --help to pcretest as an official way of being reminded
2293 of the options.
2295 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2296 and pcrecpp.cc because they annoy compilers at high warning levels.
2298 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2300 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2301 scripts are not C programs :-) and which was not noticed because it works
2302 on Linux.
2304 16. pcretest is supposed to handle any length of pattern and data line (as one
2305 line or as a continued sequence of lines) by extending its input buffer if
2306 necessary. This feature was broken for very long pattern lines, leading to
2307 a string of junk being passed to pcre_compile() if the pattern was longer
2308 than about 50K.
2310 17. I have done a major re-factoring of the way pcre_compile() computes the
2311 amount of memory needed for a compiled pattern. Previously, there was code
2312 that made a preliminary scan of the pattern in order to do this. That was
2313 OK when PCRE was new, but as the facilities have expanded, it has become
2314 harder and harder to keep it in step with the real compile phase, and there
2315 have been a number of bugs (see for example, 4 above). I have now found a
2316 cunning way of running the real compile function in a "fake" mode that
2317 enables it to compute how much memory it would need, while actually only
2318 ever using a few hundred bytes of working memory and without too many
2319 tests of the mode. This should make future maintenance and development
2320 easier. A side effect of this work is that the limit of 200 on the nesting
2321 depth of parentheses has been removed (though this was never a serious
2322 limitation, I suspect). However, there is a downside: pcre_compile() now
2323 runs more slowly than before (30% or more, depending on the pattern). I
2324 hope this isn't a big issue. There is no effect on runtime performance.
2326 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2327 newline (only possible for the last line of a file) and it was a
2328 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2330 19. Added additional timing features to pcretest. (1) The -tm option now times
2331 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2332 separate command line item, by a number that specifies the number of
2333 repeats to use when timing. The default is 50000; this gives better
2334 precision, but takes uncomfortably long for very large patterns.
2336 20. Extended pcre_study() to be more clever in cases where a branch of a
2337 subpattern has no definite first character. For example, (a*|b*)[cd] would
2338 previously give no result from pcre_study(). Now it recognizes that the
2339 first character must be a, b, c, or d.
2341 21. There was an incorrect error "recursive call could loop indefinitely" if
2342 a subpattern (or the entire pattern) that was being tested for matching an
2343 empty string contained only one non-empty item after a nested subpattern.
2344 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2345 incorrectly, because the \d was being skipped in the check.
2347 22. The pcretest program now has a new pattern option /B and a command line
2348 option -b, which is equivalent to adding /B to every pattern. This causes
2349 it to show the compiled bytecode, without the additional information that
2350 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2351 is the same as /B/I).
2353 23. A new optimization is now able automatically to treat some sequences such
2354 as a*b as a*+b. More specifically, if something simple (such as a character
2355 or a simple class like \d) has an unlimited quantifier, and is followed by
2356 something that cannot possibly match the quantified thing, the quantifier
2357 is automatically "possessified".
2359 24. A recursive reference to a subpattern whose number was greater than 39
2360 went wrong under certain circumstances in UTF-8 mode. This bug could also
2361 have affected the operation of pcre_study().
2363 25. Realized that a little bit of performance could be had by replacing
2364 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2366 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2368 27. Possessive quantifiers such as a++ were previously implemented by turning
2369 them into atomic groups such as ($>a+). Now they have their own opcodes,
2370 which improves performance. This includes the automatically created ones
2371 from 23 above.
2373 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2374 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2375 the first matched character to be a colon. This applied both to named and
2376 numbered groups.
2378 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2380 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2381 building PCRE on Windows easier, so I have included it in the distribution.
2383 31. There is now a check in pcretest against a ridiculously large number being
2384 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2385 loop, the loop is abandoned.
2387 32. Forward references to subpatterns in conditions such as (?(2)...) where
2388 subpattern 2 is defined later cause pcre_compile() to search forwards in
2389 the pattern for the relevant set of parentheses. This search went wrong
2390 when there were unescaped parentheses in a character class, parentheses
2391 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2393 33. "Subroutine" calls and backreferences were previously restricted to
2394 referencing subpatterns earlier in the regex. This restriction has now
2395 been removed.
2397 34. Added a number of extra features that are going to be in Perl 5.10. On the
2398 whole, these are just syntactic alternatives for features that PCRE had
2399 previously implemented using the Python syntax or my own invention. The
2400 other formats are all retained for compatibility.
2402 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2403 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2404 also .NET compatible.
2406 (b) A recursion or subroutine call to a named group can now be defined as
2407 (?&name) as well as (?P>name).
2409 (c) A backreference to a named group can now be defined as \k<name> or
2410 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2411 5.10, are also .NET compatible.
2413 (d) A conditional reference to a named group can now use the syntax
2414 (?(<name>) or (?('name') as well as (?(name).
2416 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2417 groups (named and numbered) that are never evaluated inline, but can be
2418 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2419 is always false. There may be only one alternative in such a group.
2421 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2422 as the simple (?(R). The condition is true only if the most recent
2423 recursion is that of the given number or name. It does not search out
2424 through the entire recursion stack.
2426 (g) The escape \gN or \g{N} has been added, where N is a positive or
2427 negative number, specifying an absolute or relative reference.
2429 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2430 some "unreachable code" warnings.
2432 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2433 things, this adds five new scripts.
2435 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2436 There were also incompatibilities regarding the handling of \Q..\E inside
2437 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2438 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2440 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2441 matches an empty string, and forcibly breaks the loop. There were bugs in
2442 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2443 against aaaa the result was just "a" rather than "aaaa", for example. Two
2444 separate and independent bugs (that affected different cases) have been
2445 fixed.
2447 39. Refactored the code to abolish the use of different opcodes for small
2448 capturing bracket numbers. This is a tidy that I avoided doing when I
2449 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2450 The new approach is not only tidier, it makes it possible to reduce the
2451 memory needed to fix the previous bug (38).
2453 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2454 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2455 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2456 mode.
2458 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2459 report.
2461 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2462 copy construction and assignment in the C++ wrapper.
2464 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2465 couple of bugs that could have given wrong results in the "--newline=crlf"
2466 case.
2468 44. Added a number of casts and did some reorganization of signed/unsigned int
2469 variables following suggestions from Dair Grant. Also renamed the variable
2470 "this" as "item" because it is a C++ keyword.
2472 45. Arranged for dftables to add
2474 #include "pcre_internal.h"
2476 to pcre_chartables.c because without it, gcc 4.x may remove the array
2477 definition from the final binary if PCRE is built into a static library and
2478 dead code stripping is activated.
2480 46. For an unanchored pattern, if a match attempt fails at the start of a
2481 newline sequence, and the newline setting is CRLF or ANY, and the next two
2482 characters are CRLF, advance by two characters instead of one.
2485 Version 6.7 04-Jul-06
2486 ---------------------
2488 1. In order to handle tests when input lines are enormously long, pcretest has
2489 been re-factored so that it automatically extends its buffers when
2490 necessary. The code is crude, but this _is_ just a test program. The
2491 default size has been increased from 32K to 50K.
2493 2. The code in pcre_study() was using the value of the re argument before
2494 testing it for NULL. (Of course, in any sensible call of the function, it
2495 won't be NULL.)
2497 3. The memmove() emulation function in pcre_internal.h, which is used on
2498 systems that lack both memmove() and bcopy() - that is, hardly ever -
2499 was missing a "static" storage class specifier.
2501 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2502 containing an extended class (one that cannot be represented by a bitmap
2503 because it contains high-valued characters or Unicode property items, e.g.
2504 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2505 pattern, but PCRE should not loop if you do not (it no longer does).
2506 [Detail: two cases were found: (a) a repeated subpattern containing an
2507 extended class; (b) a recursive reference to a subpattern that followed a
2508 previous extended class. It wasn't skipping over the extended class
2509 correctly when UTF-8 mode was not set.]
2511 5. A negated single-character class was not being recognized as fixed-length
2512 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2513 compile error "lookbehind assertion is not fixed length".
2515 6. The RunPerlTest auxiliary script was showing an unexpected difference
2516 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2517 write a Perl script that can interpret lines of an input file either as
2518 byte characters or as UTF-8, which is what "perltest" was being required to
2519 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2520 can't do is switch easily at run time between having the "use utf8;" pragma
2521 or not. In the end, I fudged it by using the RunPerlTest script to insert
2522 "use utf8;" explicitly for the UTF-8 tests.
2524 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2525 the end of the subject string, contrary to the documentation and to what
2526 Perl does. This was true of both matching functions. Now it matches only at
2527 the start of the subject and immediately after *internal* newlines.
2529 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2530 a pointer to an int instead of a pointer to an unsigned long int. This
2531 caused problems on 64-bit systems.
2533 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2534 instance of the 'standard' template library not being so standard".
2536 10. There was no check on the number of named subpatterns nor the maximum
2537 length of a subpattern name. The product of these values is used to compute
2538 the size of the memory block for a compiled pattern. By supplying a very
2539 long subpattern name and a large number of named subpatterns, the size
2540 computation could be caused to overflow. This is now prevented by limiting
2541 the length of names to 32 characters, and the number of named subpatterns
2542 to 10,000.
2544 11. Subpatterns that are repeated with specific counts have to be replicated in
2545 the compiled pattern. The size of memory for this was computed from the
2546 length of the subpattern and the repeat count. The latter is limited to
2547 65535, but there was no limit on the former, meaning that integer overflow
2548 could in principle occur. The compiled length of a repeated subpattern is
2549 now limited to 30,000 bytes in order to prevent this.
2551 12. Added the optional facility to have named substrings with the same name.
2553 13. Added the ability to use a named substring as a condition, using the
2554 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2555 are numbers (not recommended). Forward references are permitted.
2557 14. Added forward references in named backreferences (if you see what I mean).
2559 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2560 pattern could run off the end of the subject. For example, the pattern
2561 "(?s)(.{1,5})"8 did this with the subject "ab".
2563 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2564 PCRE_CASELESS was set when matching characters that were quantified with ?
2565 or *.
2567 17. A character class other than a single negated character that had a minimum
2568 but no maximum quantifier - for example [ab]{6,} - was not handled
2569 correctly by pce_dfa_exec(). It would match only one character.
2571 18. A valid (though odd) pattern that looked like a POSIX character
2572 class but used an invalid character after [ (for example [[,abc,]]) caused
2573 pcre_compile() to give the error "Failed: internal error: code overflow" or
2574 in some cases to crash with a glibc free() error. This could even happen if
2575 the pattern terminated after [[ but there just happened to be a sequence of
2576 letters, a binary zero, and a closing ] in the memory that followed.
2578 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2579 over the years. Originally (before any Unicode support), just the bottom 8
2580 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2581 output from "man perlunicode" includes this:
2583 The regular expression compiler produces polymorphic opcodes. That
2584 is, the pattern adapts to the data and automatically switches to
2585 the Unicode character scheme when presented with Unicode data--or
2586 instead uses a traditional byte scheme when presented with byte
2587 data.
2589 Sadly, a wide octal escape does not cause a switch, and in a string with
2590 no other multibyte characters, these octal escapes are treated as before.
2591 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2592 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2593 Unicode string.
2595 I have not perpetrated such confusion in PCRE. Up till now, it took just
2596 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2597 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2598 translate to the appropriate multibyte character.
2600 29. Applied some refactoring to reduce the number of warnings from Microsoft
2601 and Borland compilers. This has included removing the fudge introduced
2602 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2603 a warning about an unused variable.
2605 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2606 characters since release 4.0, because Perl (from release 5.004) does not.
2607 [Or at least, is documented not to: some releases seem to be in conflict
2608 with the documentation.] However, when a pattern was studied with
2609 pcre_study() and all its branches started with \s, PCRE still included VT
2610 as a possible starting character. Of course, this did no harm; it just
2611 caused an unnecessary match attempt.
2613 22. Removed a now-redundant internal flag bit that recorded the fact that case
2614 dependency changed within the pattern. This was once needed for "required
2615 byte" processing, but is no longer used. This recovers a now-scarce options
2616 bit. Also moved the least significant internal flag bit to the most-
2617 significant bit of the word, which was not previously used (hangover from
2618 the days when it was an int rather than a uint) to free up another bit for
2619 the future.
2621 23. Added support for CRLF line endings as well as CR and LF. As well as the
2622 default being selectable at build time, it can now be changed at runtime
2623 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2624 specify that it is scanning data with non-default line endings.
2626 24. Changed the definition of CXXLINK to make it agree with the definition of
2627 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2629 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2630 recursions. This makes a big different to stack usage for some patterns.
2632 26. If a subpattern containing a named recursion or subroutine reference such
2633 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2634 the space required for the compiled pattern went wrong and gave too small a
2635 value. Depending on the environment, this could lead to "Failed: internal
2636 error: code overflow at offset 49" or "glibc detected double free or
2637 corruption" errors.
2639 27. Applied patches from Google (a) to support the new newline modes and (b) to
2640 advance over multibyte UTF-8 characters in GlobalReplace.
2642 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2643 difference for some implementation of PCRE in some Windows version.
2645 29. Added some extra testing facilities to pcretest:
2647 \q<number> in a data line sets the "match limit" value
2648 \Q<number> in a data line sets the "match recursion limt" value
2649 -S <number> sets the stack size, where <number> is in megabytes
2651 The -S option isn't available for Windows.
2654 Version 6.6 06-Feb-06
2655 ---------------------
2657 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2658 in pcreposix.h. I have copied the definition from pcre.h.
2660 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2661 because pcre.h is no longer a built file.
2663 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2664 not normally included in the compiled code.
2667 Version 6.5 01-Feb-06
2668 ---------------------
2670 1. When using the partial match feature with pcre_dfa_exec(), it was not
2671 anchoring the second and subsequent partial matches at the new starting
2672 point. This could lead to incorrect results. For example, with the pattern
2673 /1234/, partially matching against "123" and then "a4" gave a match.
2675 2. Changes to pcregrep:
2677 (a) All non-match returns from pcre_exec() were being treated as failures
2678 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2679 error message is output. Some extra information is given for the
2681 probably the only errors that are likely to be caused by users (by
2682 specifying a regex that has nested indefinite repeats, for instance).
2683 If there are more than 20 of these errors, pcregrep is abandoned.
2685 (b) A binary zero was treated as data while matching, but terminated the
2686 output line if it was written out. This has been fixed: binary zeroes
2687 are now no different to any other data bytes.
2689 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2690 used to set a locale for matching. The --locale=xxxx long option has
2691 been added (no short equivalent) to specify a locale explicitly on the
2692 pcregrep command, overriding the environment variables.
2694 (d) When -B was used with -n, some line numbers in the output were one less
2695 than they should have been.
2697 (e) Added the -o (--only-matching) option.
2699 (f) If -A or -C was used with -c (count only), some lines of context were
2700 accidentally printed for the final match.
2702 (g) Added the -H (--with-filename) option.
2704 (h) The combination of options -rh failed to suppress file names for files
2705 that were found from directory arguments.
2707 (i) Added the -D (--devices) and -d (--directories) options.
2709 (j) Added the -F (--fixed-strings) option.
2711 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2713 (l) Added the --colo(u)r option.
2715 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2716 is not present by default.
2718 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2719 items such as (?R) or (?1), when the recursion could match a number of
2720 alternatives. If it matched one of the alternatives, but subsequently,
2721 outside the recursion, there was a failure, the code tried to back up into
2722 the recursion. However, because of the way PCRE is implemented, this is not
2723 possible, and the result was an incorrect result from the match.
2725 In order to prevent this happening, the specification of recursion has
2726 been changed so that all such subpatterns are automatically treated as
2727 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2729 4. I had overlooked the fact that, in some locales, there are characters for
2730 which isalpha() is true but neither isupper() nor islower() are true. In
2731 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2732 and ordfeminine) are like this. This affected the treatment of \w and \W
2733 when they appeared in character classes, but not when they appeared outside
2734 a character class. The bit map for "word" characters is now created
2735 separately from the results of isalnum() instead of just taking it from the
2736 upper, lower, and digit maps. (Plus the underscore character, of course.)
2738 5. The above bug also affected the handling of POSIX character classes such as
2739 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2740 permanent tables. Instead, the bit maps for such a class were previously
2741 created as the appropriate unions of the upper, lower, and digit bitmaps.
2742 Now they are created by subtraction from the [[:word:]] class, which has
2743 its own bitmap.
2745 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2746 It is created by subtracting the vertical space characters (\x09, \x0a,
2747 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2748 subtraction was done in the overall bitmap for a character class, meaning
2749 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2750 be recognized. This bug has been fixed.
2752 7. Patches from the folks at Google:
2754 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2755 real life, but is still worth protecting against".
2757 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2758 regular expressions".
2760 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2761 have it.
2763 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2764 "configure" and the latter not, in order to fix a problem somebody had
2765 with compiling the Arg class on HP-UX.
2767 (e) Improve the error-handling of the C++ wrapper a little bit.
2769 (f) New tests for checking recursion limiting.
2771 8. The pcre_memmove() function, which is used only if the environment does not
2772 have a standard memmove() function (and is therefore rarely compiled),
2773 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2774 returning a result (though PCRE never actually uses the result).
2776 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2777 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2778 returned instead of calling malloc() with an overflowing number that would
2779 most likely cause subsequent chaos.
2781 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2783 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2784 with this option is matched, the nmatch and pmatch options of regexec() are
2785 ignored.
2787 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2788 provided in case anyone wants to the the POSIX interface with UTF-8
2789 strings.
2791 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2792 C++ linking (needed for some HP-UX environments).
2794 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2795 (unused parameter) and in the pcre_printint() function (omitted "default"
2796 switch label when the default is to do nothing).
2798 15. Added some code to make it possible, when PCRE is compiled as a C++
2799 library, to replace subject pointers for pcre_exec() with a smart pointer
2800 class, thus making it possible to process discontinuous strings.
2802 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2803 much the same function. They were added by different people who were trying
2804 to make PCRE easy to compile on non-Unix systems. It has been suggested
2805 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2806 for compiling on Windows systems. I have therefore replaced it with
2807 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2808 defaults to "extern" for C or "extern C" for C++, which works fine on
2809 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2810 SCOPE with something explicit in config.h. In addition:
2812 (a) pcreposix.h still had just "extern" instead of either of these macros;
2813 I have replaced it with PCRE_DATA_SCOPE.
2815 (b) Functions such as _pcre_xclass(), which are internal to the library,
2816 but external in the C sense, all had PCRE_EXPORT in their definitions.
2817 This is apparently wrong for the Windows case, so I have removed it.
2818 (It makes no difference on Unix-like systems.)
2820 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2821 of recursive calls to match(). This is different to MATCH_LIMIT because
2822 that limits the total number of calls to match(), not all of which increase
2823 the depth of recursion. Limiting the recursion depth limits the amount of
2824 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2825 when PCRE is compiled, and changed at run time. A patch from Google adds
2826 this functionality to the C++ interface.
2828 18. Changes to the handling of Unicode character properties:
2830 (a) Updated the table to Unicode 4.1.0.
2832 (b) Recognize characters that are not in the table as "Cn" (undefined).
2834 (c) I revised the way the table is implemented to a much improved format
2835 which includes recognition of ranges. It now supports the ranges that
2836 are defined in UnicodeData.txt, and it also amalgamates other
2837 characters into ranges. This has reduced the number of entries in the
2838 table from around 16,000 to around 3,000, thus reducing its size
2839 considerably. I realized I did not need to use a tree structure after
2840 all - a binary chop search is just as efficient. Having reduced the
2841 number of entries, I extended their size from 6 bytes to 8 bytes to
2842 allow for more data.
2844 (d) Added support for Unicode script names via properties such as \p{Han}.
2846 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2847 matching that character.
2849 20. When matching a repeated Unicode property with a minimum greater than zero,
2850 (for example \pL{2,}), PCRE could look past the end of the subject if it
2851 reached it while seeking the minimum number of characters. This could
2852 happen only if some of the characters were more than one byte long, because
2853 there is a check for at least the minimum number of bytes.
2855 21. Refactored the implementation of \p and \P so as to be more general, to
2856 allow for more different types of property in future. This has changed the
2857 compiled form incompatibly. Anybody with saved compiled patterns that use
2858 \p or \P will have to recompile them.
2860 22. Added "Any" and "L&" to the supported property types.
2862 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2863 but give a compile time error if the value is greater than 0xff.
2865 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2866 accidentally not being installed or uninstalled.
2868 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2869 made were to insert the current release number. This seemed silly, because
2870 it made things harder for people building PCRE on systems that don't run
2871 "configure". I have turned pcre.h into a distributed file, no longer built
2872 by "configure", with the version identification directly included. There is
2873 no longer a pcre.h.in file.
2875 However, this change necessitated a change to the pcre-config script as
2876 well. It is built from pcre-config.in, and one of the substitutions was the
2877 release number. I have updated configure.ac so that ./configure now finds
2878 the release number by grepping pcre.h.
2880 26. Added the ability to run the tests under valgrind.
2883 Version 6.4 05-Sep-05
2884 ---------------------
2886 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2887 "--" to be printed when multiple files were scanned, even when none of the
2888 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2889 consider it to be a bug, and have restored the previous behaviour.
2891 2. A couple of code tidies to get rid of compiler warnings.
2893 3. The pcretest program used to cheat by referring to symbols in the library
2894 whose names begin with _pcre_. These are internal symbols that are not
2895 really supposed to be visible externally, and in some environments it is
2896 possible to suppress them. The cheating is now confined to including
2897 certain files from the library's source, which is a bit cleaner.
2899 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2900 file's purpose clearer.
2902 5. Reorganized pcre_ucp_findchar().
2905 Version 6.3 15-Aug-05
2906 ---------------------
2908 1. The file libpcre.pc.in did not have general read permission in the tarball.
2910 2. There were some problems when building without C++ support:
2912 (a) If C++ support was not built, "make install" and "make test" still
2913 tried to test it.
2915 (b) There were problems when the value of CXX was explicitly set. Some
2916 changes have been made to try to fix these, and ...
2918 (c) --disable-cpp can now be used to explicitly disable C++ support.
2920 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2921 backslash in a target when C++ was disabled. This confuses some
2922 versions of "make", apparently. Using an intermediate variable solves
2923 this. (Same for CPP_LOBJ.)
2925 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2926 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2927 necessary on certain architectures.
2929 4. Added a setting of -export-symbols-regex to the link command to remove
2930 those symbols that are exported in the C sense, but actually are local
2931 within the library, and not documented. Their names all begin with
2932 "_pcre_". This is not a perfect job, because (a) we have to except some
2933 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2934 available (and never for static libraries). I have made a note to try to
2935 find a way round (a) in the future.
2938 Version 6.2 01-Aug-05
2939 ---------------------
2941 1. There was no test for integer overflow of quantifier values. A construction
2942 such as {1111111111111111} would give undefined results. What is worse, if
2943 a minimum quantifier for a parenthesized subpattern overflowed and became
2944 negative, the calculation of the memory size went wrong. This could have
2945 led to memory overwriting.
2947 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2949 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2950 operating environments where this matters.
2952 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2953 PCRE options from within the C++ wrapper.
2955 5. Named capturing subpatterns were not being correctly counted when a pattern
2956 was compiled. This caused two problems: (a) If there were more than 100
2957 such subpatterns, the calculation of the memory needed for the whole
2958 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2959 back references of the form \12, where the number was greater than 9, were
2960 not recognized as back references, even though there were sufficient
2961 previous subpatterns.
2963 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2964 versions of gcc, e.g. 2.95.4.
2967 Version 6.1 21-Jun-05
2968 ---------------------
2970 1. There was one reference to the variable "posix" in pcretest.c that was not
2971 surrounded by "#if !defined NOPOSIX".
2973 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2974 the cross-check on the old pcre_info() function, for the benefit of the
2975 cut-down version of PCRE that is currently imported into Exim.
2977 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2978 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2979 patterns that start (?i) but I don't think that matters. The use of (?i) is
2980 just an example; this all applies to the other options as well.
2982 4. Since libtool seems to echo the compile commands it is issuing, the output
2983 from "make" can be reduced a bit by putting "@" in front of each libtool
2984 compile command.
2986 5. Patch from the folks at Google for configure.in to be a bit more thorough
2987 in checking for a suitable C++ installation before trying to compile the
2988 C++ stuff. This should fix a reported problem when a compiler was present,
2989 but no suitable headers.
2991 6. The man pages all had just "PCRE" as their title. I have changed them to
2992 be the relevant file name. I have also arranged that these names are
2993 retained in the file doc/pcre.txt, which is a concatenation in text format
2994 of all the man pages except the little individual ones for each function.
2996 7. The NON-UNIX-USE file had not been updated for the different set of source
2997 files that come with release 6. I also added a few comments about the C++
2998 wrapper.
3001 Version 6.0 07-Jun-05
3002 ---------------------
3004 1. Some minor internal re-organization to help with my DFA experiments.
3006 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3007 didn't matter for the library itself when fully configured, but did matter
3008 when compiling without UCP support, or within Exim, where the ucp files are
3009 not imported.
3011 3. Refactoring of the library code to split up the various functions into
3012 different source modules. The addition of the new DFA matching code (see
3013 below) to a single monolithic source would have made it really too
3014 unwieldy, quite apart from causing all the code to be include in a
3015 statically linked application, when only some functions are used. This is
3016 relevant even without the DFA addition now that patterns can be compiled in
3017 one application and matched in another.
3019 The downside of splitting up is that there have to be some external
3020 functions and data tables that are used internally in different modules of
3021 the library but which are not part of the API. These have all had their
3022 names changed to start with "_pcre_" so that they are unlikely to clash
3023 with other external names.
3025 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3026 a different (DFA) algorithm. Although it is slower than the original
3027 function, it does have some advantages for certain types of matching
3028 problem.
3030 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3031 including restarting after a partial match.
3033 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3034 defined when compiling for Windows was sent to me. I have put it into the
3035 code, though I have no means of testing or verifying it.
3037 7. Added the pcre_refcount() auxiliary function.
3039 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3040 match before or at the first newline in the subject string. In pcretest,
3041 the /f option on a pattern can be used to set this.
3043 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3044 would behave wrongly. This has been present in PCRE since release 4.0.
3046 10. A number of changes to the pcregrep command:
3048 (a) Refactored how -x works; insert ^(...)$ instead of setting
3049 PCRE_ANCHORED and checking the length, in preparation for adding
3050 something similar for -w.
3052 (b) Added the -w (match as a word) option.
3054 (c) Refactored the way lines are read and buffered so as to have more
3055 than one at a time available.
3057 (d) Implemented a pcregrep test script.
3059 (e) Added the -M (multiline match) option. This allows patterns to match
3060 over several lines of the subject. The buffering ensures that at least
3061 8K, or the rest of the document (whichever is the shorter) is available
3062 for matching (and similarly the previous 8K for lookbehind assertions).
3064 (f) Changed the --help output so that it now says
3066 -w, --word-regex(p)
3068 instead of two lines, one with "regex" and the other with "regexp"
3069 because that confused at least one person since the short forms are the
3070 same. (This required a bit of code, as the output is generated
3071 automatically from a table. It wasn't just a text change.)
3073 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3074 option but starts with a hyphen. Could be a pattern or a path name
3075 starting with a hyphen, for instance.
3077 (h) "-" can be given as a file name to represent stdin.
3079 (i) When file names are being printed, "(standard input)" is used for
3080 the standard input, for compatibility with GNU grep. Previously
3081 "<stdin>" was used.
3083 (j) The option --label=xxx can be used to supply a name to be used for
3084 stdin when file names are being printed. There is no short form.
3086 (k) Re-factored the options decoding logic because we are going to add
3087 two more options that take data. Such options can now be given in four
3088 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3090 (l) Added the -A, -B, and -C options for requesting that lines of context
3091 around matches be printed.
3093 (m) Added the -L option to print the names of files that do not contain
3094 any matching lines, that is, the complement of -l.
3096 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3097 continue to scan other files.
3099 (o) The -s option was incorrectly implemented. For compatibility with other
3100 greps, it now suppresses the error message for a non-existent or non-
3101 accessible file (but not the return code). There is a new option called
3102 -q that suppresses the output of matching lines, which was what -s was
3103 previously doing.
3105 (p) Added --include and --exclude options to specify files for inclusion
3106 and exclusion when recursing.
3108 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3109 Hopefully, it now does.
3111 12. Missing cast in pcre_study().
3113 13. Added an "uninstall" target to the makefile.
3115 14. Replaced "extern" in the function prototypes in Makefile.in with
3116 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3117 world, but is set differently for Windows.
3119 15. Added a second compiling function called pcre_compile2(). The only
3120 difference is that it has an extra argument, which is a pointer to an
3121 integer error code. When there is a compile-time failure, this is set
3122 non-zero, in addition to the error test pointer being set to point to an
3123 error message. The new argument may be NULL if no error number is required
3124 (but then you may as well call pcre_compile(), which is now just a
3125 wrapper). This facility is provided because some applications need a
3126 numeric error indication, but it has also enabled me to tidy up the way
3127 compile-time errors are handled in the POSIX wrapper.
3129 16. Added VPATH=.libs to the makefile; this should help when building with one
3130 prefix path and installing with another. (Or so I'm told by someone who
3131 knows more about this stuff than I do.)
3133 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3134 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3135 match everything, including newlines. This is not POSIX-compatible, but
3136 somebody wanted the feature. From pcretest it can be activated by using
3137 both the P and the s flags.
3139 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3141 19. libpcre.pc was being incorrectly installed as executable.
3143 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3144 it now also looks for '\r' so that it will work unmodified on Windows.
3146 21. Added Google's contributed C++ wrapper to the distribution.
3148 22. Added some untidy missing memory free() calls in pcretest, to keep
3149 Electric Fence happy when testing.
3153 Version 5.0 13-Sep-04
3154 ---------------------
3156 1. Internal change: literal characters are no longer packed up into items
3157 containing multiple characters in a single byte-string. Each character
3158 is now matched using a separate opcode. However, there may be more than one
3159 byte in the character in UTF-8 mode.
3161 2. The pcre_callout_block structure has two new fields: pattern_position and
3162 next_item_length. These contain the offset in the pattern to the next match
3163 item, and its length, respectively.
3165 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3166 insertion of callouts before each pattern item. Added the /C option to
3167 pcretest to make use of this.
3169 4. On the advice of a Windows user, the lines
3171 #if defined(_WIN32) || defined(WIN32)
3172 _setmode( _fileno( stdout ), 0x8000 );
3173 #endif /* defined(_WIN32) || defined(WIN32) */
3175 have been added to the source of pcretest. This apparently does useful
3176 magic in relation to line terminators.
3178 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3179 for the benefit of those environments where the "b" makes a difference.
3181 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3182 to know about it. I have put a hack into configure.in that adds in code
3183 to set GCC=yes if CC=icc. This seems to end up at a point in the
3184 generated configure script that is early enough to affect the setting of
3185 compiler options, which is what is needed, but I have no means of testing
3186 whether it really works. (The user who reported this had patched the
3187 generated configure script, which of course I cannot do.)
3189 LATER: After change 22 below (new libtool files), the configure script
3190 seems to know about icc (and also ecc). Therefore, I have commented out
3191 this hack in configure.in.
3193 7. Added support for pkg-config (2 patches were sent in).
3195 8. Negated POSIX character classes that used a combination of internal tables
3196 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3197 [[:^ascii]]. Typically, they would match almost any characters. The other
3198 POSIX classes were not broken in this way.
3200 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3201 to find the match, as PCRE was deluded into thinking that the match had to
3202 start at the start point or following a newline. The same bug applied to
3203 patterns with negative forward assertions or any backward assertions
3204 preceding ".*" at the start, unless the pattern required a fixed first
3205 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3207 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3208 starting at the last subject character, bytes beyond the end of the subject
3209 string were read.
3211 11. Renamed the variable "class" as "classbits" to make life easier for C++
3212 users. (Previously there was a macro definition, but it apparently wasn't
3213 enough.)
3215 12. Added the new field "tables" to the extra data so that tables can be passed
3216 in at exec time, or the internal tables can be re-selected. This allows
3217 a compiled regex to be saved and re-used at a later time by a different
3218 program that might have everything at different addresses.
3220 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3221 -R library as well as a -L library.
3223 14. The debugging options of pcretest (-d on the command line or D on a
3224 pattern) showed incorrect output for anything following an extended class
3225 that contained multibyte characters and which was followed by a quantifier.
3227 15. Added optional support for general category Unicode character properties
3228 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3229 support. It adds about 90K to the size of the library. The meanings of the
3230 inbuilt class escapes such as \d and \s have NOT been changed.
3232 16. Updated pcredemo.c to include calls to free() to release the memory for the
3233 compiled pattern.
3235 17. The generated file chartables.c was being created in the source directory
3236 instead of in the building directory. This caused the build to fail if the
3237 source directory was different from the building directory, and was
3238 read-only.
3240 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3241 file. No doubt somebody will tell me if they don't make sense... Also added
3242 Dan Mooney's comments about building on OpenVMS.
3244 19. Added support for partial matching via the PCRE_PARTIAL option for
3245 pcre_exec() and the \P data escape in pcretest.
3247 20. Extended pcretest with 3 new pattern features:
3249 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3250 write the compiled pattern to the file whose name is "rest-of-line".
3251 This is a straight binary dump of the data, with the saved pointer to
3252 the character tables forced to be NULL. The study data, if any, is
3253 written too. After writing, pcretest reads a new pattern.
3255 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3256 compiled pattern from the given file. There must not be any
3257 occurrences of "<" in the file name (pretty unlikely); if there are,
3258 pcretest will instead treat the initial "<" as a pattern delimiter.
3259 After reading in the pattern, pcretest goes on to read data lines as
3260 usual.
3262 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3263 and 16-bit fields in a compiled pattern, to simulate a pattern that
3264 was compiled on a host of opposite endianness.
3266 21. The pcre-exec() function can now cope with patterns that were compiled on
3267 hosts of opposite endianness, with this restriction:
3269 As for any compiled expression that is saved and used later, the tables
3270 pointer field cannot be preserved; the extra_data field in the arguments
3271 to pcre_exec() should be used to pass in a tables address if a value
3272 other than the default internal tables were used at compile time.
3274 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3275 now diagnosed as an error. Previously, most of the time, a negative number
3276 would have been treated as zero, but if in addition "ovector" was passed as
3277 NULL, a crash could occur.
3279 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3280 new versions from the libtool 1.5 distribution (the last one is a copy of
3281 a file called libtool.m4). This seems to have fixed the need to patch
3282 "configure" to support Darwin 1.3 (which I used to do). However, I still
3283 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3284 workstation).
3286 24. Changed the PCRE licence to be the more standard "BSD" licence.
3289 Version 4.5 01-Dec-03
3290 ---------------------
3292 1. There has been some re-arrangement of the code for the match() function so
3293 that it can be compiled in a version that does not call itself recursively.
3294 Instead, it keeps those local variables that need separate instances for
3295 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3296 needs to "recurse". Keeping track of where control must go is done by means
3297 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3298 hide most of the details from the main code, and operates only if
3299 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3300 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3301 operating.
3303 To make it easier for callers to provide specially tailored get/free
3304 functions for this usage, two new functions, pcre_stack_malloc, and
3305 pcre_stack_free, are used. They are always called in strict stacking order,
3306 and the size of block requested is always the same.
3308 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3309 PCRE has been compiled to use the stack or the heap for recursion. The
3310 -C option of pcretest uses this to show which version is compiled.
3312 A new data escape \S, is added to pcretest; it causes the amounts of store
3313 obtained and freed by both kinds of malloc/free at match time to be added
3314 to the output.
3316 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3317 what's available on my current Linux desktop machine.
3319 3. When matching a UTF-8 string, the test for a valid string at the start has
3320 been extended. If start_offset is not zero, PCRE now checks that it points
3321 to a byte that is the start of a UTF-8 character. If not, it returns
3322 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3323 this is necessary because there may be backward assertions in the pattern.
3324 When matching the same subject several times, it may save resources to use
3325 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3327 4. The code for checking the validity of UTF-8 strings has been tightened so
3328 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3329 containing "overlong sequences".
3331 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3332 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3333 should have been "&", but it just so happened that all the cases this let
3334 through by mistake were picked up later in the function.
3336 6. I had used a variable called "isblank" - this is a C99 function, causing
3337 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3339 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3340 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3341 so the version is shown; (c) stop "make test" echoing "./RunTest".
3343 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3345 9. The prototype for memmove() for systems that don't have it was using
3346 size_t, but the inclusion of the header that defines size_t was later. I've
3347 moved the #includes for the C headers earlier to avoid this.
3349 10. Added some adjustments to the code to make it easier to compiler on certain
3350 special systems:
3352 (a) Some "const" qualifiers were missing.
3353 (b) Added the macro EXPORT before all exported functions; by default this
3354 is defined to be empty.
3355 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3356 that it reads its output file name as an argument instead of writing
3357 to the standard output and assuming this can be redirected.
3359 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3360 class containing characters with values greater than 255, PCRE compilation
3361 went into a loop.
3363 12. A recursive reference to a subpattern that was within another subpattern
3364 that had a minimum quantifier of zero caused PCRE to crash. For example,
3365 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3366 recursion. If the recursively-called subpattern itself had a zero repeat,
3367 that was OK.
3369 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3370 buffer into which it was copied (for escape processing) was still set at
3371 1024, so long lines caused crashes.
3373 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3374 "internal error: code overflow...". This applied to any character class
3375 that was followed by a possessive quantifier.
3377 15. Modified the Makefile to add libpcre.la as a prerequisite for
3378 libpcreposix.la because I was told this is needed for a parallel build to
3379 work.
3381 16. If a pattern that contained .* following optional items at the start was
3382 studied, the wrong optimizing data was generated, leading to matching
3383 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3384 matching string must start with a or b or c. The correct conclusion for
3385 this pattern is that a match can start with any character.
3388 Version 4.4 13-Aug-03
3389 ---------------------
3391 1. In UTF-8 mode, a character class containing characters with values between
3392 127 and 255 was not handled correctly if the compiled pattern was studied.
3393 In fixing this, I have also improved the studying algorithm for such
3394 classes (slightly).
3396 2. Three internal functions had redundant arguments passed to them. Removal
3397 might give a very teeny performance improvement.
3399 3. Documentation bug: the value of the capture_top field in a callout is *one
3400 more than* the number of the hightest numbered captured substring.
3402 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3403 in incorrectly linking with a previously installed version. They now link
3404 explicitly with libpcre.la.
3406 5. configure.in no longer needs to recognize Cygwin specially.
3408 6. A problem in pcre.in for Windows platforms is fixed.
3410 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3411 pcretest, it used to include the size of the study block as part of its
3412 output. Unfortunately, the structure contains a field that has a different
3413 size on different hardware architectures. This meant that the tests that
3414 showed this size failed. As the block is currently always of a fixed size,
3415 this information isn't actually particularly useful in pcretest output, so
3416 I have just removed it.
3418 8. Three pre-processor statements accidentally did not start in column 1.
3419 Sadly, there are *still* compilers around that complain, even though
3420 standard C has not required this for well over a decade. Sigh.
3422 9. In pcretest, the code for checking callouts passed small integers in the
3423 callout_data field, which is a void * field. However, some picky compilers
3424 complained about the casts involved for this on 64-bit systems. Now
3425 pcretest passes the address of the small integer instead, which should get
3426 rid of the warnings.
3428 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3429 both compile and run time, and gives an error if an invalid UTF-8 sequence
3430 is found. There is a option for disabling this check in cases where the
3431 string is known to be correct and/or the maximum performance is wanted.
3433 11. In response to a bug report, I changed one line in Makefile.in from
3435 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3436 to
3437 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3439 to look similar to other lines, but I have no way of telling whether this
3440 is the right thing to do, as I do not use Windows. No doubt I'll get told
3441 if it's wrong...
3444 Version 4.3 21-May-03
3445 ---------------------
3447 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3448 Makefile.
3450 2. Some refactoring to improve the quality of the code:
3452 (i) The utf8_table... variables are now declared "const".
3454 (ii) The code for \cx, which used the "case flipping" table to upper case
3455 lower case letters, now just substracts 32. This is ASCII-specific,
3456 but the whole concept of \cx is ASCII-specific, so it seems
3457 reasonable.
3459 (iii) PCRE was using its character types table to recognize decimal and
3460 hexadecimal digits in the pattern. This is silly, because it handles
3461 only 0-9, a-f, and A-F, but the character types table is locale-
3462 specific, which means strange things might happen. A private
3463 table is now used for this - though it costs 256 bytes, a table is
3464 much faster than multiple explicit tests. Of course, the standard
3465 character types table is still used for matching digits in subject
3466 strings against \d.
3468 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3469 ending in _t are). So I've renamed it as ESC_tee.
3471 3. The first argument for regexec() in the POSIX wrapper should have been
3472 defined as "const".
3474 4. Changed pcretest to use malloc() for its buffers so that they can be
3475 Electric Fenced for debugging.
3477 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3478 to read one or more bytes before the start of the subject string. Often this
3479 had no effect on PCRE's behaviour, but in some circumstances it could
3480 provoke a segmentation fault.
3482 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3483 to try to read one or more bytes before the start of the subject string.
3485 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3486 UTF-8 support could misbehave in various ways if the subject string
3487 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3488 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3489 back over UTF-8 characters.)
3492 Version 4.2 14-Apr-03
3493 ---------------------
3495 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3497 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3498 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3499 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3500 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3501 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3503 Note: automatic setting of the BUILD variables is not yet working
3504 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3505 compile-time but not at link-time
3506 [LINK]: use for linking executables only
3507 make different versions for Windows and non-Windows
3508 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3509 libraries
3510 [LINK_FOR_BUILD]: new variable
3511 [OBJEXT]: use throughout
3512 [EXEEXT]: use throughout
3513 <winshared>: new target
3514 <wininstall>: new target
3515 <dftables.o>: use native compiler
3516 <dftables>: use native linker
3517 <install>: handle Windows platform correctly
3518 <clean>: ditto
3519 <check>: ditto
3520 copy DLL to top builddir before testing
3522 As part of these changes, -no-undefined was removed again. This was reported
3523 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3524 in any case.
3526 3. Some tidies to get rid of compiler warnings:
3528 . In the match_data structure, match_limit was an unsigned long int, whereas
3529 match_call_count was an int. I've made them both unsigned long ints.
3531 . In pcretest the fact that a const uschar * doesn't automatically cast to
3532 a void * provoked a warning.
3534 . Turning on some more compiler warnings threw up some "shadow" variables
3535 and a few more missing casts.
3537 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3538 option, a class that contained a single character with a value between 128
3539 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3541 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3542 option, a class that contained several characters, but with at least one
3543 whose value was between 128 and 255 caused PCRE to crash.
3546 Version 4.1 12-Mar-03
3547 ---------------------
3549 1. Compiling with gcc -pedantic found a couple of places where casts were
3550 needed, and a string in dftables.c that was longer than standard compilers are
3551 required to support.
3553 2. Compiling with Sun's compiler found a few more places where the code could
3554 be tidied up in order to avoid warnings.
3556 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3557 first of these names is deprecated in the latest Autoconf in favour of the name
3558 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3559 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3560 analogy I have changed it to CFLAGS_FOR_BUILD.
3562 4. Added -no-undefined to the linking command in the Makefile, because this is
3563 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3564 linking step for the pcreposix library.
3566 5. PCRE was failing to diagnose the case of two named groups with the same
3567 name.
3569 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3570 literal character that is needed in the subject for a match, and scans along to
3571 ensure that it is present before embarking on the full matching process. This
3572 saves time in cases of nested unlimited repeats that are never going to match.
3573 Problem: the scan can take a lot of time if the subject is very long (e.g.
3574 megabytes), thus penalizing straightforward matches. It is now done only if the
3575 amount of subject to be scanned is less than 1000 bytes.
3577 7. A lesser problem with the same optimization is that it was recording the
3578 first character of an anchored pattern as "needed", thus provoking a search
3579 right along the subject, even when the first match of the pattern was going to
3580 fail. The "needed" character is now not set for anchored patterns, unless it
3581 follows something in the pattern that is of non-fixed length. Thus, it still
3582 fulfils its original purpose of finding quick non-matches in cases of nested
3583 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3586 Version 4.0 17-Feb-03
3587 ---------------------
3589 1. If a comment in an extended regex that started immediately after a meta-item
3590 extended to the end of string, PCRE compiled incorrect data. This could lead to
3591 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3593 2. Moved to autoconf 2.53 and libtool 1.4.2.
3595 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3596 the special perltest8 script is no longer needed - all the tests can be run
3597 from a single perltest script.
3599 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3600 by \s. It has now been removed in PCRE. This means it isn't recognized as
3601 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3602 class [:space:] *does* include VT, thereby creating a mess.
3604 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3605 space and tab.
3607 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3608 its new features into the main test script, reducing the number of scripts.
3610 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3611 were backward compatible, and made the (?i) apply to the whole pattern, as if
3612 /i were given. Now it behaves more logically, and applies the option setting
3613 only to what follows. PCRE has been changed to follow suit. However, if it
3614 finds options settings right at the start of the pattern, it extracts them into
3615 the global options, as before. Thus, they show up in the info data.
3617 8. Added support for the \Q...\E escape sequence. Characters in between are
3618 treated as literals. This is slightly different from Perl in that $ and @ are
3619 also handled as literals inside the quotes. In Perl, they will cause variable
3620 interpolation. Note the following examples:
3622 Pattern PCRE matches Perl matches
3624 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3625 \Qabc\$xyz\E abc\$xyz abc\$xyz
3626 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3628 For compatibility with Perl, \Q...\E sequences are recognized inside character
3629 classes as well as outside them.
3631 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3632 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3633 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3634 signed/unsigned warnings.
3636 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3637 option for pcretest, so I've replaced it by a simple function that does just
3638 that job.
3640 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3641 "pcregrep -".
3643 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3644 Java package. This provides some syntactic sugar for simple cases of what my
3645 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3646 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3647 item, you can use this simplified notation. Note that only makes sense with
3648 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3649 greediness, whatever the setting of the PCRE_UNGREEDY option.
3651 13. A change of greediness default within a pattern was not taking effect at
3652 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3653 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3654 was abstracted outside.
3656 14. PCRE now supports the \G assertion. It is true when the current matching
3657 position is at the start point of the match. This differs from \A when the
3658 starting offset is non-zero. Used with the /g option of pcretest (or similar
3659 code), it works in the same way as it does for Perl's /g option. If all
3660 alternatives of a regex begin with \G, the expression is anchored to the start
3661 match position, and the "anchored" flag is set in the compiled expression.
3663 15. Some bugs concerning the handling of certain option changes within patterns
3664 have been fixed. These applied to options other than (?ims). For example,
3665 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3666 been the other way round. Some of this was related to change 7 above.
3668 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3669 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3670 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3671 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3673 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3674 mode. Unlike ".", it always matches newline, whatever the setting of
3675 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3676 assertions. Perl allows it, but it doesn't (in general) work because it can't
3677 calculate the length of the lookbehind. At least, that's the case for Perl
3678 5.8.0 - I've been told they are going to document that it doesn't work in
3679 future.
3681 18. Added an error diagnosis for escapes that PCRE does not support: these are
3682 \L, \l, \N, \P, \p, \U, \u, and \X.
3684 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3685 reading past the end of the pattern in cases such as /[abcd/.
3687 20. PCRE was getting more memory than necessary for patterns with classes that
3688 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3690 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3691 compiling PCRE for use with Virtual Pascal.
3693 22. Small fix to the Makefile to make it work properly if the build is done
3694 outside the source tree.
3696 23. Added a new extension: a condition to go with recursion. If a conditional
3697 subpattern starts with (?(R) the "true" branch is used if recursion has
3698 happened, whereas the "false" branch is used only at the top level.
3700 24. When there was a very long string of literal characters (over 255 bytes
3701 without UTF support, over 250 bytes with UTF support), the computation of how
3702 much memory was required could be incorrect, leading to segfaults or other
3703 strange effects.
3705 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3706 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3707 there was a subsequent back reference to those brackets. This meant that, for
3708 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3709 possible to check for precisely this case. All we can do is abandon the
3710 optimization if .* occurs inside capturing brackets when there are any back
3711 references whatsoever. (See below for a better fix that came later.)
3713 26. The handling of the optimization for finding the first character of a
3714 non-anchored pattern, and for finding a character that is required later in the
3715 match were failing in some cases. This didn't break the matching; it just
3716 failed to optimize when it could. The way this is done has been re-implemented.
3718 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3720 28. Added a new feature that provides some of the functionality that Perl
3721 provides with (?{...}). The facility is termed a "callout". The way it is done
3722 in PCRE is for the caller to provide an optional function, by setting
3723 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3724 global variable. By default it is unset, which disables all calling out. To get
3725 the function called, the regex must include (?C) at appropriate points. This
3726 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3727 This provides a means of identifying different callout points. When PCRE
3728 reaches such a point in the regex, if pcre_callout has been set, the external
3729 function is called. It is provided with data in a structure called
3730 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3731 matching continues; if it returns a non-zero value, the match at the current
3732 point fails. However, backtracking will occur if possible. [This was changed
3733 later and other features added - see item 49 below.]
3735 29. pcretest is upgraded to test the callout functionality. It provides a
3736 callout function that displays information. By default, it shows the start of
3737 the match and the current position in the text. There are some new data escapes
3738 to vary what happens:
3740 \C+ in addition, show current contents of captured substrings
3741 \C- do not supply a callout function
3742 \C!n return 1 when callout number n is reached
3743 \C!n!m return 1 when callout number n is reached for the mth time
3745 30. If pcregrep was called with the -l option and just a single file name, it
3746 output "<stdin>" if a match was found, instead of the file name.
3748 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3749 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3750 pcre_exec(). This saves a malloc/free per call. The default value of
3751 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3752 when configuring.
3754 32. The default maximum size of a compiled pattern is 64K. There have been a
3755 few cases of people hitting this limit. The code now uses macros to handle the
3756 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3757 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3758 configuring. Tests 2 and 5 work only with 2-byte links because they output
3759 debugging information about compiled patterns.
3761 33. Internal code re-arrangements:
3763 (a) Moved the debugging function for printing out a compiled regex into
3764 its own source file (printint.c) and used #include to pull it into
3765 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3766 separate copies.
3768 (b) Defined the list of op-code names for debugging as a macro in
3769 internal.h so that it is next to the definition of the opcodes.
3771 (c) Defined a table of op-code lengths for simpler skipping along compiled
3772 code. This is again a macro in internal.h so that it is next to the
3773 definition of the opcodes.
3775 34. Added support for recursive calls to individual subpatterns, along the
3776 lines of Robin Houston's patch (but implemented somewhat differently).
3778 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3779 allow it to read and process whole directories in Win32. This code was
3780 contributed by Lionel Fourquaux; it has not been tested by me.
3782 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3783 used to name a group. Names consist of alphanumerics and underscores, and must
3784 be unique. Back references use the syntax (?P=name) and recursive calls use
3785 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3786 numbers. The function pcre_fullinfo() can be used after compilation to extract
3787 a name/number map. There are three relevant calls:
3789 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3790 PCRE_INFO_NAMECOUNT yields the number of entries
3791 PCRE_INFO_NAMETABLE yields a pointer to the map.
3793 The map is a vector of fixed-size entries. The size of each entry depends on
3794 the length of the longest name used. The first two bytes of each entry are the
3795 group number, most significant byte first. There follows the corresponding
3796 name, zero terminated. The names are in alphabetical order.
3798 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3799 case instead of 255. Making it the same both with and without UTF-8 support
3800 means that the same test output works with both.
3802 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3803 calling malloc() with a zero argument.
3805 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3806 optimization. I've improved things by keeping a bitmap of backreferences with
3807 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3808 fact referenced, the optimization can be applied. It is unlikely that a
3809 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3810 the match to follow \n) will appear inside brackets with a number greater than
3811 31, but if it does, any back reference > 31 suppresses the optimization.
3813 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3814 of disabling numbered capturing parentheses. Any opening parenthesis that is
3815 not followed by ? behaves as if it were followed by ?: but named parentheses
3816 can still be used for capturing (and they will acquire numbers in the usual
3817 way).
3819 41. Redesigned the return codes from the match() function into yes/no/error so
3820 that errors can be passed back from deep inside the nested calls. A malloc
3821 failure while inside a recursive subpattern call now causes the
3822 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3824 42. It is now possible to set a limit on the number of times the match()
3825 function is called in a call to pcre_exec(). This facility makes it possible to
3826 limit the amount of recursion and backtracking, though not in a directly
3827 obvious way, because the match() function is used in a number of different
3828 circumstances. The count starts from zero for each position in the subject
3829 string (for non-anchored patterns). The default limit is, for compatibility, a
3830 large number, namely 10 000 000. You can change this in two ways:
3832 (a) When configuring PCRE before making, you can use --with-match-limit=n
3833 to set a default value for the compiled library.
3835 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3836 a different value is set. See 45 below.
3838 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3840 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3841 of things that can be changed at compile time. The first argument specifies
3842 what is wanted and the second points to where the information is to be placed.
3843 The current list of available information is:
3847 The output is an integer that is set to one if UTF-8 support is available;
3848 otherwise it is set to zero.
3852 The output is an integer that it set to the value of the code that is used for
3853 newline. It is either LF (10) or CR (13).
3857 The output is an integer that contains the number of bytes used for internal
3858 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3862 The output is an integer that contains the threshold above which the POSIX
3863 interface uses malloc() for output vectors. See item 31 above.
3867 The output is an unsigned integer that contains the default limit of the number
3868 of match() calls in a pcre_exec() execution. See 42 above.
3870 44. pcretest has been upgraded by the addition of the -C option. This causes it
3871 to extract all the available output from the new pcre_config() function, and to
3872 output it. The program then exits immediately.
3874 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3875 order to support additional features. One way would have been to define
3876 pcre_exec2() (for example) with extra arguments, but this would not have been
3877 extensible, and would also have required all calls to the original function to
3878 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3879 is used for passing in "extra" data from pcre_study().
3881 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3882 contains the following fields:
3884 flags a bitmap indicating which of the following fields are set
3885 study_data opaque data from pcre_study()
3886 match_limit a way of specifying a limit on match() calls for a specific
3887 call to pcre_exec()
3888 callout_data data for callouts (see 49 below)
3890 The flag bits are also defined in pcre.h, and are
3896 The pcre_study() function now returns one of these new pcre_extra blocks, with
3897 the actual study data pointed to by the study_data field, and the
3898 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3899 before. That is, this change is entirely upwards-compatible and requires no
3900 change to existing code.
3902 If you want to pass in additional data to pcre_exec(), you can either place it
3903 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3904 block.
3906 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3907 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3908 times with different match limits, until it finds the minimum value needed for
3909 pcre_exec() to complete. The value is then output. This can be instructive; for
3910 most simple matches the number is quite small, but for pathological cases it
3911 gets very large very quickly.
3913 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3914 returns the size of the data block pointed to by the study_data field in a
3915 pcre_extra block, that is, the value that was passed as the argument to
3916 pcre_malloc() when PCRE was getting memory in which to place the information
3917 created by pcre_study(). The fourth argument should point to a size_t variable.
3918 pcretest has been extended so that this information is shown after a successful
3919 pcre_study() call when information about the compiled regex is being displayed.
3921 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3922 because what follows is always an absolute path. (Later: it turns out that this
3923 is more than cosmetic for MinGW, because it doesn't like empty path
3924 components.)
3926 49. Some changes have been made to the callout feature (see 28 above):
3928 (i) A callout function now has three choices for what it returns:
3930 0 => success, carry on matching
3931 > 0 => failure at this point, but backtrack if possible
3932 < 0 => serious error, return this value from pcre_exec()
3934 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3935 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3936 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3937 use by callout functions. It will never be used by PCRE itself.
3939 (ii) The pcre_extra structure (see 45 above) has a void * field called
3940 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3941 pcre_callout_block structure has a field of the same name. The contents of
3942 the field passed in the pcre_extra structure are passed to the callout
3943 function in the corresponding field in the callout block. This makes it
3944 easier to use the same callout-containing regex from multiple threads. For
3945 testing, the pcretest program has a new data escape
3947 \C*n pass the number n (may be negative) as callout_data
3949 If the callout function in pcretest receives a non-zero value as
3950 callout_data, it returns that value.
3952 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3953 there were some redundant $(CFLAGS) in commands that are now specified as
3954 $(LINK), which already includes $(CFLAGS).
3956 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3957 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3958 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3959 one-byte characters throughout. Note that case-insensitive matching applies
3960 only to characters whose values are less than 256. PCRE doesn't support the
3961 notion of cases for higher-valued characters.
3963 (i) A character class whose characters are all within 0-255 is handled as
3964 a bit map, and the map is inverted for negative classes. Previously, a
3965 character > 255 always failed to match such a class; however it should
3966 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3968 (ii) A negated character class with a single character < 255 is coded as
3969 "not this character" (OP_NOT). This wasn't working properly when the test
3970 character was multibyte, either singly or repeated.
3972 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3973 mode, for example: \x{100}{2,3}.
3975 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3976 singly or repeated) now correctly test multibyte characters. However,
3977 PCRE doesn't recognize any characters with values greater than 255 as
3978 digits, spaces, or word characters. Such characters always match \D, \S,
3979 and \W, and never match \d, \s, or \w.
3981 (v) Classes may now contain characters and character ranges with values
3982 greater than 255. For example: [ab\x{100}-\x{400}].
3984 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3985 PCRE in UTF-8 mode.
3987 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3988 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3989 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3990 value.)
3992 53. The single man page has become too large. I have therefore split it up into
3993 a number of separate man pages. These also give rise to individual HTML pages;
3994 these are now put in a separate directory, and there is an index.html page that
3995 lists them all. Some hyperlinking between the pages has been installed.
3997 54. Added convenience functions for handling named capturing parentheses.
3999 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4000 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4001 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4002 are faulted.
4004 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4005 calling configure. These values are used when compiling the dftables.c program
4006 which is run to generate the source of the default character tables. They
4007 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4008 you will need to set these values.
4010 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4013 Version 3.9 02-Jan-02
4014 ---------------------
4016 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4018 2. If --disable-static was given, the building process failed when trying to
4019 build pcretest and pcregrep. (For some reason it was using libtool to compile
4020 them, which is not right, as they aren't part of the library.)
4023 Version 3.8 18-Dec-01
4024 ---------------------
4026 1. The experimental UTF-8 code was completely screwed up. It was packing the
4027 bytes in the wrong order. How dumb can you get?
4030 Version 3.7 29-Oct-01
4031 ---------------------
4033 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4034 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4035 this didn't happen under Solaris 8, where I normally test things.
4037 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4038 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4039 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4040 this fix an hour or so after the initial 3.7 release.)
4043 Version 3.6 23-Oct-01
4044 ---------------------
4046 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4047 offsets passed as NULL with zero offset count.
4049 2. The config.guess and config.sub files had not been updated when I moved to
4050 the latest autoconf.
4053 Version 3.5 15-Aug-01
4054 ---------------------
4056 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4057 had been forgotten.
4059 2. By using declared but undefined structures, we can avoid using "void"
4060 definitions in pcre.h while keeping the internal definitions of the structures
4061 private.
4063 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4064 user point of view, this means that both static and shared libraries are built
4065 by default, but this can be individually controlled. More of the work of
4066 handling this static/shared cases is now inside libtool instead of PCRE's make
4067 file.
4069 4. The pcretest utility is now installed along with pcregrep because it is
4070 useful for users (to test regexs) and by doing this, it automatically gets
4071 relinked by libtool. The documentation has been turned into a man page, so
4072 there are now .1, .txt, and .html versions in /doc.
4074 5. Upgrades to pcregrep:
4075 (i) Added long-form option names like gnu grep.
4076 (ii) Added --help to list all options with an explanatory phrase.
4077 (iii) Added -r, --recursive to recurse into sub-directories.
4078 (iv) Added -f, --file to read patterns from a file.
4080 6. pcre_exec() was referring to its "code" argument before testing that
4081 argument for NULL (and giving an error if it was NULL).
4083 7. Upgraded Makefile.in to allow for compiling in a different directory from
4084 the source directory.
4086 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4087 options bits, the pointer it was passed was to an int instead of to an unsigned
4088 long int. This mattered only on 64-bit systems.
4090 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4091 generated) instead of pcre.in, which it its source. Also made the same change
4092 in several of the .c files.
4094 10. A new release of gcc defines printf() as a macro, which broke pcretest
4095 because it had an ifdef in the middle of a string argument for printf(). Fixed
4096 by using separate calls to printf().
4098 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4099 script, to force use of CR or LF instead of \n in the source. On non-Unix
4100 systems, the value can be set in config.h.
4102 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4103 absolute limit. Changed the text of the error message to make this clear, and
4104 likewise updated the man page.
4106 13. The limit of 99 on the number of capturing subpatterns has been removed.
4107 The new limit is 65535, which I hope will not be a "real" limit.
4110 Version 3.4 22-Aug-00
4111 ---------------------
4113 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4115 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4118 Version 3.3 01-Aug-00
4119 ---------------------
4121 1. If an octal character was given, but the value was greater than \377, it
4122 was not getting masked to the least significant bits, as documented. This could
4123 lead to crashes in some systems.
4125 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4126 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4128 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4129 These just pass their arguments on to (pcre_free)(), but they are provided
4130 because some uses of PCRE bind it to non-C systems that can call its functions,
4131 but cannot call free() or pcre_free() directly.
4133 4. Add "make test" as a synonym for "make check". Corrected some comments in
4134 the Makefile.
4136 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4137 Makefile.
4139 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4140 command called pgrep for grepping around the active processes.
4142 7. Added the beginnings of support for UTF-8 character strings.
4144 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4145 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4146 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4147 out for the ar command.)
4150 Version 3.2 12-May-00
4151 ---------------------
4153 This is purely a bug fixing release.
4155 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4156 of ZA. This was just one example of several cases that could provoke this bug,
4157 which was introduced by change 9 of version 2.00. The code for breaking
4158 infinite loops after an iteration that matches an empty string was't working
4159 correctly.
4161 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4162 when matched against abbab (for example). After matching an empty string, it
4163 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4164 caused it to match further down the string than it should.
4166 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4167 was there because it doesn't seem to be needed, and it causes trouble on some
4168 systems, as it is not a Standard C header. It has been removed.
4170 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4171 were reported on the Macintosh. The changes were from
4173 while ((c = *(++ptr)) != 0 && c != '\n');
4174 to
4175 while ((c = *(++ptr)) != 0 && c != '\n') ;
4177 Totally extraordinary, but if that's what it takes...
4179 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4180 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4181 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4182 assumes the way PCRE uses memmove() (always moving upwards).
4184 6. PCRE is being used in one environment where strchr() is not available. There
4185 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4186 faster code anyway.
4189 Version 3.1 09-Feb-00
4190 ---------------------
4192 The only change in this release is the fixing of some bugs in Makefile.in for
4193 the "install" target:
4195 (1) It was failing to install pcreposix.h.
4197 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4200 Version 3.0 01-Feb-00
4201 ---------------------
4203 1. Add support for the /+ modifier to perltest (to output $` like it does in
4204 pcretest).
4206 2. Add support for the /g modifier to perltest.
4208 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4209 matches null strings.
4211 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4212 pattern. Perl treats empty patterns specially - it reuses the most recent
4213 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4214 effect.
4216 5. The POSIX interface was broken in that it was just handing over the POSIX
4217 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4218 required a bigger vector, with some working space on the end. This means that
4219 the POSIX wrapper now has to get and free some memory, and copy the results.
4221 6. Added some simple autoconf support, placing the test data and the
4222 documentation in separate directories, re-organizing some of the
4223 information files, and making it build pcre-config (a GNU standard). Also added
4224 libtool support for building PCRE as a shared library, which is now the
4225 default.
4227 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4228 09 are not valid octal constants. Single digits will be used for minor values
4229 less than 10.
4231 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4232 existing programs that set these in the POSIX interface can use PCRE without
4233 modification.
4235 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4236 return all that pcre_info() returns, plus additional data. The pcre_info()
4237 function is retained for compatibility, but is considered to be obsolete.
4239 10. Added experimental recursion feature (?R) to handle one common case that
4240 Perl 5.6 will be able to do with (?p{...}).
4242 11. Added support for POSIX character classes like [:alpha:], which Perl is
4243 adopting.
4246 Version 2.08 31-Aug-99
4247 ----------------------
4249 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4250 trying to match at the startoffset position, but instead was moving forward to
4251 the next newline as if a previous match had failed.
4253 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4254 and could get into a loop if a null string was matched other than at the start
4255 of the subject.
4257 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4258 be distinguished at compile time, and for completeness also added PCRE_DATE.
4260 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4261 in GnuWin32 environments.
4264 Version 2.07 29-Jul-99
4265 ----------------------
4267 1. The documentation is now supplied in plain text form and HTML as well as in
4268 the form of man page sources.
4270 2. C++ compilers don't like assigning (void *) values to other pointer types.
4271 In particular this affects malloc(). Although there is no problem in Standard
4272 C, I've put in casts to keep C++ compilers happy.
4274 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4275 should be (const char *).
4277 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4278 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4279 However, I haven't made this a standard facility. The documentation doesn't
4280 mention it, and the Makefile doesn't support it.
4282 5. The Makefile now contains an "install" target, with editable destinations at
4283 the top of the file. The pcretest program is not installed.
4285 6. pgrep -V now gives the PCRE version number and date.
4287 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4288 causing the entire string to be ignored, instead of just the last character.
4290 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4291 non-matching string, it can take a very, very long time, even for strings of
4292 quite modest length, because of the nested recursion. PCRE now does better in
4293 some of these cases. It does this by remembering the last required literal
4294 character in the pattern, and pre-searching the subject to ensure it is present
4295 before running the real match. In other words, it applies a heuristic to detect
4296 some types of certain failure quickly, and in the above example, if presented
4297 with a string that has no trailing " it gives "no match" very quickly.
4299 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4300 other alternatives are tried instead.
4303 Version 2.06 09-Jun-99
4304 ----------------------
4306 1. Change pcretest's output for amount of store used to show just the code
4307 space, because the remainder (the data block) varies in size between 32-bit and
4308 64-bit systems.
4310 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4311 start matching at. This allows lookbehinds to work when searching for multiple
4312 occurrences in a string.
4314 3. Added additional options to pcretest for testing multiple occurrences:
4316 /+ outputs the rest of the string that follows a match
4317 /g loops for multiple occurrences, using the new startoffset argument
4318 /G loops for multiple occurrences by passing an incremented pointer
4320 4. PCRE wasn't doing the "first character" optimization for patterns starting
4321 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4322 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4323 the letter 'x'. On long subject strings, this gives a significant speed-up.
4326 Version 2.05 21-Apr-99
4327 ----------------------
4329 1. Changed the type of magic_number from int to long int so that it works
4330 properly on 16-bit systems.
4332 2. Fixed a bug which caused patterns starting with .* not to work correctly
4333 when the subject string contained newline characters. PCRE was assuming
4334 anchoring for such patterns in all cases, which is not correct because .* will
4335 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4336 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4337 must be retried after every newline in the subject.
4340 Version 2.04 18-Feb-99
4341 ----------------------
4343 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4344 computation of the store needed to hold the pattern was incorrect (too large).
4345 If such patterns were nested a few deep, this could multiply and become a real
4346 problem.
4348 2. Added /M option to pcretest to show the memory requirement of a specific
4349 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4351 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4352 compiled in such a way that the backtracking after subsequent failure was
4353 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4354 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4357 Version 2.03 02-Feb-99
4358 ----------------------
4360 1. Fixed typo and small mistake in man page.
4362 2. Added 4th condition (GPL supersedes if conflict) and created separate
4363 LICENCE file containing the conditions.
4365 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4366 Perl, that is the internal \ allows the delimiter to be included in the
4367 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4368 the final delimiter, add \ to the end of the pattern (to test the error).
4370 4. Added the convenience functions for extracting substrings after a successful
4371 match. Updated pcretest to make it able to test these functions.
4374 Version 2.02 14-Jan-99
4375 ----------------------
4377 1. Initialized the working variables associated with each extraction so that
4378 their saving and restoring doesn't refer to uninitialized store.
4380 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4381 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4382 fix the problem.
4384 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4385 calls, and wasn't printing the correct value for compiling calls. Increased the
4386 default value of LOOPREPEAT, and the number of significant figures in the
4387 times.
4389 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4391 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4392 a building problem on Windows NT with a FAT file system.
4395 Version 2.01 21-Oct-98
4396 ----------------------
4398 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4399 to character tables built by pcre_maketables() in the current locale. If NULL
4400 is passed, the default tables are used.
4403 Version 2.00 24-Sep-98
4404 ----------------------
4406 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4407 it any more.
4409 2. Allow quantification of (?>) groups, and make it work correctly.
4411 3. The first character computation wasn't working for (?>) groups.
4413 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4414 end of the subject) and add 5.005's \z, which really does match only at the
4415 very end of the subject.
4417 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4419 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4420 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4421 localized options. All options to pcre_study() were also removed.
4423 7. Add other new features from 5.005:
4425 $(?<= positive lookbehind
4426 $(?<! negative lookbehind
4427 (?imsx-imsx) added the unsetting capability
4428 such a setting is global if at outer level; local otherwise
4429 (?imsx-imsx:) non-capturing groups with option setting
4430 (?(cond)re|re) conditional pattern matching
4432 A backreference to itself in a repeated group matches the previous
4433 captured string.
4435 8. General tidying up of studying (both automatic and via "study")
4436 consequential on the addition of new assertions.
4438 9. As in 5.005, unlimited repeated groups that could match an empty substring
4439 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4440 runtime if any iteration does actually match an empty substring.
4442 10. Include the RunTest script in the distribution.
4444 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4445 discrepancies, some of which were old and were also with respect to 5.004. They
4446 have now been fixed.
4449 Version 1.09 28-Apr-98
4450 ----------------------
4452 1. A negated single character class followed by a quantifier with a minimum
4453 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4454 program crashes, or just wrong answers. This did not apply to negated classes
4455 containing more than one character, or to minima other than one.
4458 Version 1.08 27-Mar-98
4459 ----------------------
4461 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4463 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4464 latter must appear before anything that relies on it in the pattern.
4467 Version 1.07 16-Feb-98
4468 ----------------------
4470 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4471 repeat of a potentially empty string).
4474 Version 1.06 23-Jan-98
4475 ----------------------
4477 1. Added Markus Oberhumer's little patches for C++.
4479 2. Literal strings longer than 255 characters were broken.
4482 Version 1.05 23-Dec-97
4483 ----------------------
4485 1. Negated character classes containing more than one character were failing if
4486 PCRE_CASELESS was set at run time.
4489 Version 1.04 19-Dec-97
4490 ----------------------
4492 1. Corrected the man page, where some "const" qualifiers had been omitted.
4494 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4495 input syntax.
4497 3. Fixed memory leak which occurred when a regex with back references was
4498 matched with an offsets vector that wasn't big enough. The temporary memory
4499 that is used in this case wasn't being freed if the match failed.
4501 4. Tidied pcretest to ensure it frees memory that it gets.
4503 5. Temporary memory was being obtained in the case where the passed offsets
4504 vector was exactly big enough.
4506 6. Corrected definition of offsetof() from change 5 below.
4508 7. I had screwed up change 6 below and broken the rules for the use of
4509 setjmp(). Now fixed.
4512 Version 1.03 18-Dec-97
4513 ----------------------
4515 1. A erroneous regex with a missing opening parenthesis was correctly
4516 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4517 on some systems.
4519 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4520 it was reported that one broken compiler failed on the former because "code" is
4521 also an independent variable.
4523 3. The erroneous regex a[]b caused an array overrun reference.
4525 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4526 fail on data ending with that character. (It was going on too far, and checking
4527 the next character, typically a binary zero.) This was specific to the
4528 optimized code for single-character negative classes.
4530 5. Added a contributed patch from the TIN world which does the following:
4532 + Add an undef for memmove, in case the the system defines a macro for it.
4534 + Add a definition of offsetof(), in case there isn't one. (I don't know
4535 the reason behind this - offsetof() is part of the ANSI standard - but
4536 it does no harm).
4538 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4539 most of the places where whitespace preceded '#'. I have given up and
4540 allowed the remaining 2 cases to be at the margin.
4542 + Rename some variables in pcre to eliminate shadowing. This seems very
4543 pedantic, but does no harm, of course.
4545 6. Moved the call to setjmp() into its own function, to get rid of warnings
4546 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4548 7. Constructs such as \d{8,} were compiling into the equivalent of
4549 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4550 outcome, but in this particular case used more store than had been allocated,
4551 which caused the bug to be discovered because it threw up an internal error.
4553 8. The debugging code in both pcre and pcretest for outputting the compiled
4554 form of a regex was going wrong in the case of back references followed by
4555 curly-bracketed repeats.
4558 Version 1.02 12-Dec-97
4559 ----------------------
4561 1. Typos in pcre.3 and comments in the source fixed.
4563 2. Applied a contributed patch to get rid of places where it used to remove
4564 'const' from variables, and fixed some signed/unsigned and uninitialized
4565 variable warnings.
4567 3. Added the "runtest" target to Makefile.
4569 4. Set default compiler flag to -O2 rather than just -O.
4572 Version 1.01 19-Nov-97
4573 ----------------------
4575 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4576 like /([ab]*)*/, that is, for classes with more than one character in them.
4578 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4579 as /((?>a*))*/ (a PCRE_EXTRA facility).
4582 Version 1.00 18-Nov-97
4583 ----------------------
4585 1. Added compile-time macros to support systems such as SunOS4 which don't have
4586 memmove() or strerror() but have other things that can be used instead.
4588 2. Arranged that "make clean" removes the executables.
4591 Version 0.99 27-Oct-97
4592 ----------------------
4594 1. Fixed bug in code for optimizing classes with only one character. It was
4595 initializing a 32-byte map regardless, which could cause it to run off the end
4596 of the memory it had got.
4598 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4601 Version 0.98 22-Oct-97
4602 ----------------------
4604 1. Fixed bug in code for handling temporary memory usage when there are more
4605 back references than supplied space in the ovector. This could cause segfaults.
4608 Version 0.97 21-Oct-97
4609 ----------------------
4611 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4613 2. Optimized negated single characters not to use a bit map.
4615 3. Brought error texts together as macro definitions; clarified some of them;
4616 fixed one that was wrong - it said "range out of order" when it meant "invalid
4617 escape sequence".
4619 4. Changed some char * arguments to const char *.
4621 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4623 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4624 pcretest.
4627 Version 0.96 16-Oct-97
4628 ----------------------
4630 1. Added a simple "pgrep" utility to the distribution.
4632 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4633 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4634 where "ddd" means "one or more decimal digits".
4636 3. Fixed serious bug. If a pattern had a back reference, but the call to
4637 pcre_exec() didn't supply a large enough ovector to record the related
4638 identifying subpattern, the match always failed. PCRE now remembers the number
4639 of the largest back reference, and gets some temporary memory in which to save
4640 the offsets during matching if necessary, in order to ensure that
4641 backreferences always work.
4643 4. Increased the compatibility with Perl in a number of ways:
4645 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4646 to request this handling. The option can be set at compile or exec time.
4648 (b) $ matches before a terminating newline by default; an option
4649 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4650 mode). The option can be set at compile or exec time.
4652 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4653 the same as Perl's. If the decimal number it represents is less than 10
4654 or there aren't that many previous left capturing parentheses, an octal
4655 escape is read. Inside a character class, it's always an octal escape,
4656 even if it is a single digit.
4658 (d) An escaped but undefined alphabetic character is taken as a literal,
4659 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4660 escapes.
4662 (e) {0} is now permitted. (The previous item is removed from the compiled
4663 pattern).
4665 5. Changed all the names of code files so that the basic parts are no longer
4666 than 10 characters, and abolished the teeny "globals.c" file.
4668 6. Changed the handling of character classes; they are now done with a 32-byte
4669 bit map always.
4671 7. Added the -d and /D options to pcretest to make it possible to look at the
4672 internals of compilation without having to recompile pcre.
4675 Version 0.95 23-Sep-97
4676 ----------------------
4678 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4679 \x20 at the start of a run of normal characters. These were being treated as
4680 real characters, instead of the source characters being re-checked.
4683 Version 0.94 18-Sep-97
4684 ----------------------
4686 1. The functions are now thread-safe, with the caveat that the global variables
4687 containing pointers to malloc() and free() or alternative functions are the
4688 same for all threads.
4690 2. Get pcre_study() to generate a bitmap of initial characters for non-
4691 anchored patterns when this is possible, and use it if passed to pcre_exec().
4694 Version 0.93 15-Sep-97
4695 ----------------------
4697 1. /(b)|(:+)/ was computing an incorrect first character.
4699 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4700 but not actually doing anything yet.
4702 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4703 as Perl does (e.g. [-az] or [az-]).
4705 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4706 all possible positions.
4708 5. Split up into different modules to avoid including unneeded functions in a
4709 compiled binary. However, compile and exec are still in one module. The "study"
4710 function is split off.
4712 6. The character tables are now in a separate module whose source is generated
4713 by an auxiliary program - but can then be edited by hand if required. There are
4714 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4715 toupper() in the code.
4717 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4718 make them global. Abolish the function for setting them, as the caller can now
4719 set them directly.
4722 Version 0.92 11-Sep-97
4723 ----------------------
4725 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4726 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4728 2. Caseless matching was not working in character classes if the characters in
4729 the pattern were in upper case.
4731 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4733 4. Make PCRE_ANCHORED public and accept as a compile option.
4735 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4736 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4737 pass them.
4739 6. Give an error if bad option bits passed at compile or run time.
4741 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4742 pcretest to cause it to pass that flag.
4744 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4745 options, and the first character, if set.
4747 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4750 Version 0.91 10-Sep-97
4751 ----------------------
4753 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4754 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4756 2. PCRE was looping on encountering an indefinitely repeated back reference to
4757 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4758 Perl does - treats the match as successful.
4760 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5