ViewVC logotype

Diff of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 77 by nigel, Sat Feb 24 21:40:45 2007 UTC revision 987 by zherczeg, Sat Jul 7 04:11:29 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.32
5    ------------
7    1.  Improved JIT compiler optimizations for first character search and single
8        character iterators.
10    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
11        Patch by Daniel Richard G.
14    Version 8.31 06-July-2012
15    -------------------------
17    1.  Fixing a wrong JIT test case and some compiler warnings.
19    2.  Removed a bashism from the RunTest script.
21    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
22        to unsigned type, result still unsigned" that was given by an MS compiler
23        on encountering the code "-sizeof(xxx)".
25    4.  Partial matching support is added to the JIT compiler.
27    5.  Fixed several bugs concerned with partial matching of items that consist
28        of more than one character:
30        (a) /^(..)\1/ did not partially match "aba" because checking references was
31            done on an "all or nothing" basis. This also applied to repeated
32            references.
34        (b) \R did not give a hard partial match if \r was found at the end of the
35            subject.
37        (c) \X did not give a hard partial match after matching one or more
38            characters at the end of the subject.
40        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
41            a partial match for the string "\r".
43        (e) When newline was set to CRLF, the metacharacter "." did not recognize
44            a partial match for a CR character at the end of the subject string.
46    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
47        running pcretest, the text "(JIT)" added to the output whenever JIT is
48        actually used to run the match.
50    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
51        or /S+[+] with a digit between 1 and 7.
53    8.  OP_NOT now supports any UTF character not just single-byte ones.
55    9.  (*MARK) control verb is now supported by the JIT compiler.
57    10. The command "./RunTest list" lists the available tests without actually
58        running any of them. (Because I keep forgetting what they all are.)
62    12. Applied a (slightly modified) user-supplied patch that improves performance
63        when the heap is used for recursion (compiled with --disable-stack-for-
64        recursion). Instead of malloc and free for each heap frame each time a
65        logical recursion happens, frames are retained on a chain and re-used where
66        possible. This sometimes gives as much as 30% improvement.
68    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
69        call.
71    14. As documented, (*COMMIT) is now confined to within a positive assertion.
73    15. It is now possible to link pcretest with libedit as an alternative to
74        libreadline.
76    16. (*COMMIT) control verb is now supported by the JIT compiler.
78    17. The Unicode data tables have been updated to Unicode 6.1.0.
80    18. Added --file-list option to pcregrep.
82    19. Added binary file support to pcregrep, including the -a, --binary-files,
83        -I, and --text options.
85    20. The madvise function is renamed for posix_madvise for QNX compatibility
86        reasons. Fixed by Giuseppe D'Angelo.
88    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
90    22. Changed the option for creating symbolic links for 16-bit man pages from
91        -s to -sf so that re-installing does not cause issues.
93    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
95    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
96        matches in certain environments (the workspace was not being correctly
97        retained). Also added to pcre_dfa_exec() a simple plausibility check on
98        some of the workspace data at the beginning of a restart.
100    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
101        was not doing so when it should - probably a typo introduced by SVN 528
102        (change 8.10/14).
104    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
105        \w+ when the character tables indicated that \x{c4} was a word character.
106        There were several related cases, all because the tests for doing a table
107        lookup were testing for characters less than 127 instead of 255.
109    27. If a pattern contains capturing parentheses that are not used in a match,
110        their slots in the ovector are set to -1. For those that are higher than
111        any matched groups, this happens at the end of processing. In the case when
112        there were back references that the ovector was too small to contain
113        (causing temporary malloc'd memory to be used during matching), and the
114        highest capturing number was not used, memory off the end of the ovector
115        was incorrectly being set to -1. (It was using the size of the temporary
116        memory instead of the true size.)
118    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
119        ovector size, it uses memory at the end of the block that it has got.
121    29. Check for an overlong MARK name and give an error at compile time. The
122        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
124    30. JIT compiler update.
126    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
127        Rill for the patch.
129    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
131    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
133    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
134        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
135        was enabled.
137    35. Improve JIT code generation for greedy plus quantifier.
139    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
140        1 to "aa" instead of to an empty string. The bug affected repeated groups
141        that could potentially match an empty string.
143    37. Optimizing single character iterators in JIT.
145    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
146        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
147        codepoints that are too big for the mode are faulted, and in a UTF mode,
148        disallowed codepoints are also faulted.
150    39. If PCRE was compiled with UTF support, in three places in the DFA
151        matcher there was code that should only have been obeyed in UTF mode, but
152        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
153        processing when bytes with values greater than 127 were present. In 16-bit
154        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
155        both cases the values are those that cannot be the first data item in a UTF
156        character. The three items that might have provoked this were recursions,
157        possessively repeated groups, and atomic groups.
159    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
160        and pcregrep, because some OS require shared objects to be explicitly
161        passed to ld, causing the link step to fail if they are not.
163    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
164        mode, patterns that started with \h* or \R* might be incorrectly matched.
167    Version 8.30 04-February-2012
168    -----------------------------
170    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
171        name is defined in ctype.h.
173    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
174        only in quite long subpatterns.
176    3.  Removed the function pcre_info(), which has been obsolete and deprecated
177        since it was replaced by pcre_fullinfo() in February 2000.
179    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
180        match a (*MARK), and the match failed at the start of the subject, a
181        reference to memory before the start of the subject could occur. This bug
182        was introduced by fix 17 of release 8.21.
184    5.  A reference to an unset group with zero minimum repetition was giving
185        totally wrong answers (in non-JavaScript-compatibility mode). For example,
186        /(another)?(\1?)test/ matched against "hello world test". This bug was
187        introduced in release 8.13.
189    6.  Add support for 16-bit character strings (a large amount of work involving
190        many changes and refactorings).
192    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
193        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
194        from a file.
196    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
197        rounding is not applied in this particular case).
199    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
200        if they appear, or are escaped, in patterns.
202    10. Get rid of a number of -Wunused-but-set-variable warnings.
204    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
205        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
206        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
207        also returns the mark "x". This bug applied to capturing parentheses,
208        non-capturing parentheses, and atomic parentheses. It also applied to some
209        assertions.
211    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
212        information out of configure.ac instead of relying on pcre.h.generic, which
213        is not stored in the repository.
215    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
216        -lreadline.
218    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
220    15. Applied Graycode's patch to put the top-level frame on the stack rather
221        than the heap when not using the stack for recursion. This gives a
222        performance improvement in many cases when recursion is not deep.
224    16. Experimental code added to "pcretest -C" to output the stack frame size.
227    Version 8.21 12-Dec-2011
228    ------------------------
230    1.  Updating the JIT compiler.
232    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
233        are added as well.
235    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
236        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
237        calling _pcre_jit_exec. Some extra comments are added.
239    4.  (*MARK) settings inside atomic groups that do not contain any capturing
240        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
241        was introduced by change 18 for 8.20.
243    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
244        ECMA-262 standard.
246    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
247        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
248        This bug was probably introduced by change 9 of 8.13.
250    7.  While fixing 6 above, I noticed that a number of other items were being
251        incorrectly rejected as "not fixed length". This arose partly because newer
252        opcodes had not been added to the fixed-length checking code. I have (a)
253        corrected the bug and added tests for these items, and (b) arranged for an
254        error to occur if an unknown opcode is encountered while checking for fixed
255        length instead of just assuming "not fixed length". The items that were
256        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
257        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
258        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
260    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
261        being incorrectly compiled and would have given unpredicatble results.
263    9.  A possessively repeated subpattern with minimum repeat count greater than
264        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
265        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
266        the first (A) could occur when it should not.
268    10. Add a cast and remove a redundant test from the code.
270    11. JIT should use pcre_malloc/pcre_free for allocation.
272    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
273        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
274        is anything other than /usr, -L is still shown).
276    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
278    14. Perl does not support \N without a following name in a [] class; PCRE now
279        also gives an error.
281    15. If a forward reference was repeated with an upper limit of around 2000,
282        it caused the error "internal error: overran compiling workspace". The
283        maximum number of forward references (including repeats) was limited by the
284        internal workspace, and dependent on the LINK_SIZE. The code has been
285        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
286        the default depends on LINK_SIZE. There is a new upper limit (for safety)
287        of around 200,000 forward references. While doing this, I also speeded up
288        the filling in of repeated forward references.
290    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
291        incorrectly expecting the subject to contain another "a" after the start.
293    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
294        in the match, the SKIP should be ignored. This was not happening; instead
295        the SKIP was being treated as NOMATCH. For patterns such as
296        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
297        tested.
299    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
300        now much more compatible with Perl, in particular in cases where the result
301        is a non-match for a non-anchored pattern. For example, if
302        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
303        "m", where previously it did not return a name. A side effect of this
304        change is that for partial matches, the last encountered mark name is
305        returned, as for non matches. A number of tests that were previously not
306        Perl-compatible have been moved into the Perl-compatible test files. The
307        refactoring has had the pleasing side effect of removing one argument from
308        the match() function, thus reducing its stack requirements.
310    19. If the /S+ option was used in pcretest to study a pattern using JIT,
311        subsequent uses of /S (without +) incorrectly behaved like /S+.
313    21. Retrieve executable code size support for the JIT compiler and fixing
314        some warnings.
316    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
317        not work when the shorter character appeared right at the end of the
318        subject string.
320    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
321        systems.
323    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
324        output it when the /M option is used in pcretest.
326    25. The CheckMan script was not being included in the distribution. Also, added
327        an explicit "perl" to run Perl scripts from the PrepareRelease script
328        because this is reportedly needed in Windows.
330    26. If study data was being save in a file and studying had not found a set of
331        "starts with" bytes for the pattern, the data written to the file (though
332        never used) was taken from uninitialized memory and so caused valgrind to
333        complain.
335    27. Updated RunTest.bat as provided by Sheri Pierce.
337    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
339    29. Computation of memory usage for the table of capturing group names was
340        giving an unnecessarily large value.
343    Version 8.20 21-Oct-2011
344    ------------------------
346    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
347        a POSIX class. After further experiments with Perl, which convinced me that
348        Perl has bugs and confusions, a closing square bracket is no longer allowed
349        in a POSIX name. This bug also affected patterns with classes that started
350        with full stops.
352    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
353        captured substring, but while checking the failing first alternative,
354        substring 1 is temporarily captured. If the output vector supplied to
355        pcre_exec() was not big enough for this capture, the yield of the function
356        was still zero ("insufficient space for captured substrings"). This cannot
357        be totally fixed without adding another stack variable, which seems a lot
358        of expense for a edge case. However, I have improved the situation in cases
359        such as /(a)(b)x|abc/ matched against "abc", where the return code
360        indicates that fewer than the maximum number of slots in the ovector have
361        been set.
363    3.  Related to (2) above: when there are more back references in a pattern than
364        slots in the output vector, pcre_exec() uses temporary memory during
365        matching, and copies in the captures as far as possible afterwards. It was
366        using the entire output vector, but this conflicts with the specification
367        that only 2/3 is used for passing back captured substrings. Now it uses
368        only the first 2/3, for compatibility. This is, of course, another edge
369        case.
371    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
372        main code base, and can be used by building with --enable-jit. When this is
373        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
374        runtime --no-jit option is given.
376    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
377        ovector, the return from the function was zero, implying that there were
378        other matches that did not fit. The correct "exactly full" value is now
379        returned.
381    6.  If a subpattern that was called recursively or as a subroutine contained
382        (*PRUNE) or any other control that caused it to give a non-standard return,
383        invalid errors such as "Error -26 (nested recursion at the same subject
384        position)" or even infinite loops could occur.
386    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
387        computing the minimum length on reaching *ACCEPT, and so ended up with the
388        wrong value of 1 rather than 0. Further investigation indicates that
389        computing a minimum subject length in the presence of *ACCEPT is difficult
390        (think back references, subroutine calls), and so I have changed the code
391        so that no minimum is registered for a pattern that contains *ACCEPT.
393    8.  If (*THEN) was present in the first (true) branch of a conditional group,
394        it was not handled as intended. [But see 16 below.]
396    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
397        Sheri Pierce.
399    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
400        the first byte in a match must be "a".
402    11. Change 17 for 8.13 increased the recursion depth for patterns like
403        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
404        pattern contains any instances of (*THEN). If it does not, the old
405        optimizations are restored. It would be nice to do this on a per-group
406        basis, but at the moment that is not feasible.
408    12. In some environments, the output of pcretest -C is CRLF terminated. This
409        broke RunTest's code that checks for the link size. A single white space
410        character after the value is now allowed for.
412    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
413        For "fr", it uses the Windows-specific input and output files.
415    14. If (*THEN) appeared in a group that was called recursively or as a
416        subroutine, it did not work as intended. [But see next item.]
418    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
419        pattern fragments (but not containing any | characters). If A and B are
420        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
421        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
422        D. In other words, Perl considers parentheses that do not contain any |
423        characters to be part of a surrounding alternative, whereas PCRE was
424        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
425        differently. PCRE now behaves in the same way as Perl, except in the case
426        of subroutine/recursion calls such as (?1) which have in any case always
427        been different (but PCRE had them first :-).
429    16. Related to 15 above: Perl does not treat the | in a conditional group as
430        creating alternatives. Such a group is treated in the same way as an
431        ordinary group without any | characters when processing (*THEN). PCRE has
432        been changed to match Perl's behaviour.
434    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
435        RunGrepTest script failed.
437    18. Change 22 for version 13 caused atomic groups to use more stack. This is
438        inevitable for groups that contain captures, but it can lead to a lot of
439        stack use in large patterns. The old behaviour has been restored for atomic
440        groups that do not contain any capturing parentheses.
442    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
443        suppress the check for a minimum subject length at run time. (If it was
444        given to pcre_exec() or pcre_dfa_exec() it did work.)
446    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
447        fail to work when decoding hex characters in data strings in EBCDIC
448        environments.
450    21. It appears that in at least one Mac OS environment, the isxdigit() function
451        is implemented as a macro that evaluates to its argument more than once,
452        contravening the C 90 Standard (I haven't checked a later standard). There
453        was an instance in pcretest which caused it to go wrong when processing
454        \x{...} escapes in subject strings. The has been rewritten to avoid using
455        things like p++ in the argument of isxdigit().
458    Version 8.13 16-Aug-2011
459    ------------------------
461    1.  The Unicode data tables have been updated to Unicode 6.0.0.
463    2.  Two minor typos in pcre_internal.h have been fixed.
465    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
466        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
467        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
469    4.  There were a number of related bugs in the code for matching backrefences
470        caselessly in UTF-8 mode when codes for the characters concerned were
471        different numbers of bytes. For example, U+023A and U+2C65 are an upper
472        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
473        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
474        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
475        2-byte code at the end of the subject (it thought there wasn't enough data
476        left).
478    5.  Comprehensive information about what went wrong is now returned by
479        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
480        as the output vector has at least 2 elements. The offset of the start of
481        the failing character and a reason code are placed in the vector.
483    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
484        now returned is for the first byte of the failing character, instead of the
485        last byte inspected. This is an incompatible change, but I hope it is small
486        enough not to be a problem. It makes the returned offset consistent with
487        pcre_exec() and pcre_dfa_exec().
489    7.  pcretest now gives a text phrase as well as the error number when
490        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
491        failure, the offset and reason code are output.
493    8.  When \R was used with a maximizing quantifier it failed to skip backwards
494        over a \r\n pair if the subsequent match failed. Instead, it just skipped
495        back over a single character (\n). This seems wrong (because it treated the
496        two characters as a single entity when going forwards), conflicts with the
497        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
498        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
499        has been changed.
501    9.  Some internal refactoring has changed the processing so that the handling
502        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
503        time (the PCRE_DOTALL option was changed this way some time ago: version
504        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
505        which was always a bit of a fudge. It also means that there is one less
506        argument for the match() function, which reduces its stack requirements
507        slightly. This change also fixes an incompatibility with Perl: the pattern
508        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
510    10. More internal refactoring has drastically reduced the number of recursive
511        calls to match() for possessively repeated groups such as (abc)++ when
512        using pcre_exec().
514    11. While implementing 10, a number of bugs in the handling of groups were
515        discovered and fixed:
517        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
518        (a|)*(?1) gave a compile-time internal error.
519        ((a|)+)+  did not notice that the outer group could match an empty string.
520        (^a|^)+   was not marked as anchored.
521        (.*a|.*)+ was not marked as matching at start or after a newline.
523    12. Yet more internal refactoring has removed another argument from the match()
524        function. Special calls to this function are now indicated by setting a
525        value in a variable in the "match data" data block.
527    13. Be more explicit in pcre_study() instead of relying on "default" for
528        opcodes that mean there is no starting character; this means that when new
529        ones are added and accidentally left out of pcre_study(), testing should
530        pick them up.
532    14. The -s option of pcretest has been documented for ages as being an old
533        synonym of -m (show memory usage). I have changed it to mean "force study
534        for every regex", that is, assume /S for every regex. This is similar to -i
535        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
536        using it. It makes it easier to run collections of tests with and without
537        study enabled, and thereby test pcre_study() more easily. All the standard
538        tests are now run with and without -s (but some patterns can be marked as
539        "never study" - see 20 below).
541    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
542        restoration of the capturing data to the outer values was not happening
543        correctly.
545    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
546        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
547        pattern had matched an empty string, and so incorrectly returned a no
548        match.
550    17. There was optimizing code for the last branch of non-capturing parentheses,
551        and also for the obeyed branch of a conditional subexpression, which used
552        tail recursion to cut down on stack usage. Unfortunately, now that there is
553        the possibility of (*THEN) occurring in these branches, tail recursion is
554        no longer possible because the return has to be checked for (*THEN). These
555        two optimizations have therefore been removed. [But see 8.20/11 above.]
557    18. If a pattern containing \R was studied, it was assumed that \R always
558        matched two bytes, thus causing the minimum subject length to be
559        incorrectly computed because \R can also match just one byte.
561    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
562        was incorrectly computed.
564    20. If /S is present twice on a test pattern in pcretest input, it now
565        *disables* studying, thereby overriding the use of -s on the command line
566        (see 14 above). This is necessary for one or two tests to keep the output
567        identical in both cases.
569    21. When (*ACCEPT) was used in an assertion that matched an empty string and
570        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
572    22. When an atomic group that contained a capturing parenthesis was
573        successfully matched, but the branch in which it appeared failed, the
574        capturing was not being forgotten if a higher numbered group was later
575        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
576        group 1 to "a", when in fact it should be unset. This applied to multi-
577        branched capturing and non-capturing groups, repeated or not, and also to
578        positive assertions (capturing in negative assertions does not happen
579        in PCRE) and also to nested atomic groups.
581    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
582        subject after a captured substring, to make it easier to tell which of a
583        number of identical substrings has been captured.
585    24. The way atomic groups are processed by pcre_exec() has been changed so that
586        if they are repeated, backtracking one repetition now resets captured
587        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
588        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
589        "aaa". Previously, it would have been "a". As part of this code
590        refactoring, the way recursive calls are handled has also been changed.
592    25. If an assertion condition captured any substrings, they were not passed
593        back unless some other capturing happened later. For example, if
594        (?(?=(a))a) was matched against "a", no capturing was returned.
596    26. When studying a pattern that contained subroutine calls or assertions,
597        the code for finding the minimum length of a possible match was handling
598        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
599        group 1 called group 2 while simultaneously a separate group 2 called group
600        1). A stack overflow occurred in this case. I have fixed this by limiting
601        the recursion depth to 10.
603    27. Updated RunTest.bat in the distribution to the version supplied by Tom
604        Fortmann. This supports explicit test numbers on the command line, and has
605        argument validation and error reporting.
607    28. An instance of \X with an unlimited repeat could fail if at any point the
608        first character it looked at was a mark character.
610    29. Some minor code refactoring concerning Unicode properties and scripts
611        should reduce the stack requirement of match() slightly.
613    30. Added the '=' option to pcretest to check the setting of unused capturing
614        slots at the end of the pattern, which are documented as being -1, but are
615        not included in the return count.
617    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
618        compiled something random. Now it gives a compile-time error (as does
619        Perl).
621    32. A *MARK encountered during the processing of a positive assertion is now
622        recorded and passed back (compatible with Perl).
624    33. If --only-matching or --colour was set on a pcregrep call whose pattern
625        had alternative anchored branches, the search for a second match in a line
626        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
627        matched the line "0102" twice. The same bug affected patterns that started
628        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
629        twice.
631    34. Previously, PCRE did not allow quantification of assertions. However, Perl
632        does, and because of capturing effects, quantifying parenthesized
633        assertions may at times be useful. Quantifiers are now allowed for
634        parenthesized assertions.
636    35. A minor code tidy in pcre_compile() when checking options for \R usage.
638    36. \g was being checked for fancy things in a character class, when it should
639        just be a literal "g".
641    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
642        appearance of a nested POSIX class supersedes an apparent external class.
643        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
644        unescaped square brackets may also appear as part of class names. For
645        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
646        more like Perl. (But see 8.20/1 above.)
648    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
649        was because it thought it was \N{name}, which is not supported).
651    39. Add minix to OS list not supporting the -S option in pcretest.
653    40. PCRE tries to detect cases of infinite recursion at compile time, but it
654        cannot analyze patterns in sufficient detail to catch mutual recursions
655        such as ((?1))((?2)). There is now a runtime test that gives an error if a
656        subgroup is called recursively as a subpattern for a second time at the
657        same position in the subject string. In previous releases this might have
658        been caught by the recursion limit, or it might have run out of stack.
660    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
661        happen only once. PCRE was, however incorrectly giving a compile time error
662        "recursive call could loop indefinitely" because it cannot analyze the
663        pattern in sufficient detail. The compile time test no longer happens when
664        PCRE is compiling a conditional subpattern, but actual runaway loops are
665        now caught at runtime (see 40 above).
667    42. It seems that Perl allows any characters other than a closing parenthesis
668        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
669        has been changed to be the same.
671    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
672        as not to get warnings when autogen.sh is called. Also changed
673        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
675    44. To help people who use pcregrep to scan files containing exceedingly long
676        lines, the following changes have been made:
678        (a) The default value of the buffer size parameter has been increased from
679            8K to 20K. (The actual buffer used is three times this size.)
681        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
682            PCRE is built.
684        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
685            to be set at run time.
687        (d) Numerical values in pcregrep options can be followed by K or M, for
688            example --buffer-size=50K.
690        (e) If a line being scanned overflows pcregrep's buffer, an error is now
691            given and the return code is set to 2.
693    45. Add a pointer to the latest mark to the callout data block.
695    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
696        partial match of an empty string instead of no match. This was specific to
697        the use of ".".
699    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
700        complete match instead of a partial match. This bug was dependent on both
701        the PCRE_UTF8 and PCRE_DOTALL options being set.
703    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
704        starting byte set, because \b was not being ignored.
707    Version 8.12 15-Jan-2011
708    ------------------------
710    1.  Fixed some typos in the markup of the man pages, and wrote a script that
711        checks for such things as part of the documentation building process.
713    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
714        --match-limit and --recursion-limit options (added for 8.11). In
715        particular, this made one of the standard tests fail. (The integer value
716        went into the wrong half of a long int.)
718    3.  If the --colour option was given to pcregrep with -v (invert match), it
719        did strange things, either producing crazy output, or crashing. It should,
720        of course, ignore a request for colour when reporting lines that do not
721        match.
723    4.  Another pcregrep bug caused similar problems if --colour was specified with
724        -M (multiline) and the pattern match finished with a line ending.
726    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
727        matched in multiline mode, the following line was shown as part of the
728        match. This seems wrong, so I have changed it.
730    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
731        the check for further matches in the same line (so they could be coloured)
732        to overrun the end of the current line. If another match was found, it was
733        incorrectly shown (and then shown again when found in the next line).
735    7.  If pcregrep was compiled under Windows, there was a reference to the
736        function pcregrep_exit() before it was defined. I am assuming this was
737        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
738        reported by a user. I've moved the definition above the reference.
741    Version 8.11 10-Dec-2010
742    ------------------------
744    1.  (*THEN) was not working properly if there were untried alternatives prior
745        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
746        backtracked to try for "b" instead of moving to the next alternative branch
747        at the same level (in this case, to look for "c"). The Perl documentation
748        is clear that when (*THEN) is backtracked onto, it goes to the "next
749        alternative in the innermost enclosing group".
751    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
752        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
753        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
754        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
755        (*THEN).
757    3.  If \s appeared in a character class, it removed the VT character from
758        the class, even if it had been included by some previous item, for example
759        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
760        of \s, but is part of the POSIX "space" class.)
762    4.  A partial match never returns an empty string (because you can always
763        match an empty string at the end of the subject); however the checking for
764        an empty string was starting at the "start of match" point. This has been
765        changed to the "earliest inspected character" point, because the returned
766        data for a partial match starts at this character. This means that, for
767        example, /(?<=abc)def/ gives a partial match for the subject "abc"
768        (previously it gave "no match").
770    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
771        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
772        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
773        has an implication that the given string is incomplete (because a partial
774        match is preferred over a full match). For this reason, these items now
775        give a partial match in this situation. [Aside: previously, the one case
776        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
777        match rather than a full match, which was wrong by the old rules, but is
778        now correct.]
780    6.  There was a bug in the handling of #-introduced comments, recognized when
781        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
782        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
783        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
784        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
785        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
786        places in pcre_compile().
788    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
789        comments when looking ahead for named forward references to subpatterns,
790        the only newline sequence it recognized was NL. It now handles newlines
791        according to the set newline convention.
793    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
794        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
795        cater for a lack of strerror(). These oversights have been fixed.
797    9.  Added --match-limit and --recursion-limit to pcregrep.
799    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
801    11. When the -o option was used, pcregrep was setting a return code of 1, even
802        when matches were found, and --line-buffered was not being honoured.
804    12. Added an optional parentheses number to the -o and --only-matching options
805        of pcregrep.
807    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
808        can match an empty string. The code to do it in pcretest and pcredemo
809        needed fixing:
811        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
812            only one byte after an empty string match just before CRLF (this case
813            just got forgotten; "any" and "anycrlf" were OK).
815        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
816            mode when an empty string match preceded an ASCII character followed by
817            a non-ASCII character. (The code for advancing by one character rather
818            than one byte was nonsense.)
820        (c) The pcredemo.c sample program did not have any code at all to handle
821            the cases when CRLF is a valid newline sequence.
823    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
824        as a starting offset was within the subject string. There is now a new
825        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
826        negative or greater than the length of the string. In order to test this,
827        pcretest is extended to allow the setting of negative starting offsets.
829    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
830        starting offset points to the beginning of a UTF-8 character was
831        unnecessarily clumsy. I tidied it up.
833    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
834        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
836    17. Nobody had reported that the --include_dir option, which was added in
837        release 7.7 should have been called --include-dir (hyphen, not underscore)
838        for compatibility with GNU grep. I have changed it to --include-dir, but
839        left --include_dir as an undocumented synonym, and the same for
840        --exclude-dir, though that is not available in GNU grep, at least as of
841        release 2.5.4.
843    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
844        characters from a string of bytes) have been redefined so as not to use
845        loops, in order to improve performance in some environments. At the same
846        time, I abstracted some of the common code into auxiliary macros to save
847        repetition (this should not affect the compiled code).
849    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
850        compile-time error is now given if \c is not followed by an ASCII
851        character, that is, a byte less than 128. (In EBCDIC mode, the code is
852        different, and any byte value is allowed.)
854    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
855        START_OPTIMIZE option, which is now allowed at compile time - but just
856        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
857        to pcregrep and other applications that have no direct access to PCRE
858        options. The new /Y option in pcretest sets this option when calling
859        pcre_compile().
861    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
862        back references. Groups containing recursive back references were forced to
863        be atomic by that change, but in the case of named groups, the amount of
864        memory required was incorrectly computed, leading to "Failed: internal
865        error: code overflow". This has been fixed.
867    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
868        pcretest.c, to avoid build problems in some Borland environments.
871    Version 8.10 25-Jun-2010
872    ------------------------
874    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
875        THEN.
877    2.  (*ACCEPT) was not working when inside an atomic group.
879    3.  Inside a character class, \B is treated as a literal by default, but
880        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
881        causes the error). The code is unchanged, but I tidied the documentation.
883    4.  Inside a character class, PCRE always treated \R and \X as literals,
884        whereas Perl faults them if its -w option is set. I have changed PCRE so
885        that it faults them when PCRE_EXTRA is set.
887    5.  Added support for \N, which always matches any character other than
888        newline. (It is the same as "." when PCRE_DOTALL is not set.)
890    6.  When compiling pcregrep with newer versions of gcc which may have
891        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
892        declared with attribute warn_unused_result" were given. Just casting the
893        result to (void) does not stop the warnings; a more elaborate fudge is
894        needed. I've used a macro to implement this.
896    7.  Minor change to pcretest.c to avoid a compiler warning.
898    8.  Added four artifical Unicode properties to help with an option to make
899        \s etc use properties (see next item). The new properties are: Xan
900        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
902    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
903        use Unicode properties. (*UCP) at the start of a pattern can be used to set
904        this option. Modified pcretest to add /W to test this facility. Added
905        REG_UCP to make it available via the POSIX interface.
907    10. Added --line-buffered to pcregrep.
909    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
910        studied, and the match started with a letter with a code point greater than
911        127 whose first byte was different to the first byte of the other case of
912        the letter, the other case of this starting letter was not recognized
913        (#976).
915    12. If a pattern that was studied started with a repeated Unicode property
916        test, for example, \p{Nd}+, there was the theoretical possibility of
917        setting up an incorrect bitmap of starting bytes, but fortunately it could
918        not have actually happened in practice until change 8 above was made (it
919        added property types that matched character-matching opcodes).
921    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
922        possible starting bytes for non-anchored patterns.
924    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
925        \R, and also a number of cases that involve Unicode properties, both
926        explicit and implicit when PCRE_UCP is set.
928    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
929        input, it could crash or give wrong results if characters with values
930        greater than 0xc0 were present in the subject string. (Detail: it assumed
931        UTF-8 input when processing these items.)
933    16. Added a lot of (int) casts to avoid compiler warnings in systems where
934        size_t is 64-bit (#991).
936    17. Added a check for running out of memory when PCRE is compiled with
937        --disable-stack-for-recursion (#990).
939    18. If the last data line in a file for pcretest does not have a newline on
940        the end, a newline was missing in the output.
942    19. The default pcre_chartables.c file recognizes only ASCII characters (values
943        less than 128) in its various bitmaps. However, there is a facility for
944        generating tables according to the current locale when PCRE is compiled. It
945        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
946        characters, are recognized by isspace() and therefore were getting set in
947        these tables, and indeed these tables seem to approximate to ISO 8859. This
948        caused a problem in UTF-8 mode when pcre_study() was used to create a list
949        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
950        which of course cannot start UTF-8 characters. I have changed the code so
951        that only real ASCII characters (less than 128) and the correct starting
952        bytes for UTF-8 encodings are set for characters greater than 127 when in
953        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
954        altogether.)
956    20. Added the /T option to pcretest so as to be able to run tests with non-
957        standard character tables, thus making it possible to include the tests
958        used for 19 above in the standard set of tests.
960    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
961        reference to a subpattern the other side of a comment that contains an
962        opening parenthesis caused either an internal compiling error, or a
963        reference to the wrong subpattern.
966    Version 8.02 19-Mar-2010
967    ------------------------
969    1.  The Unicode data tables have been updated to Unicode 5.2.0.
971    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
972        configured.
974    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
975        original author of that file, following a query about its status.
977    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
978        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
980    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
981        quantifier applied to a forward-referencing subroutine call, could compile
982        incorrect code or give the error "internal error: previously-checked
983        referenced subpattern not found".
985    6.  Both MS Visual Studio and Symbian OS have problems with initializing
986        variables to point to external functions. For these systems, therefore,
987        pcre_malloc etc. are now initialized to local functions that call the
988        relevant global functions.
990    7.  There were two entries missing in the vectors called coptable and poptable
991        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
992        I've fixed the data, and added a kludgy way of testing at compile time that
993        the lengths are correct (equal to the number of opcodes).
995    8.  Following on from 7, I added a similar kludge to check the length of the
996        eint vector in pcreposix.c.
998    9.  Error texts for pcre_compile() are held as one long string to avoid too
999        much relocation at load time. To find a text, the string is searched,
1000        counting zeros. There was no check for running off the end of the string,
1001        which could happen if a new error number was added without updating the
1002        string.
1004    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1006    11. \K was not working if it appeared in an atomic group or in a group that
1007        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1008        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1009        the assertion is positive, but not if it is negative.
1011    12. Change 11 fortuitously reduced the size of the stack frame used in the
1012        "match()" function of pcre_exec.c by one pointer. Forthcoming
1013        implementation of support for (*MARK) will need an extra pointer on the
1014        stack; I have reserved it now, so that the stack frame size does not
1015        decrease.
1017    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1018        item in branch that calls a recursion is a subroutine call - as in the
1019        second branch in the above example - was incorrectly given the compile-
1020        time error "recursive call could loop indefinitely" because pcre_compile()
1021        was not correctly checking the subroutine for matching a non-empty string.
1023    14. The checks for overrunning compiling workspace could trigger after an
1024        overrun had occurred. This is a "should never occur" error, but it can be
1025        triggered by pathological patterns such as hundreds of nested parentheses.
1026        The checks now trigger 100 bytes before the end of the workspace.
1028    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1031    Version 8.01 19-Jan-2010
1032    ------------------------
1034    1.  If a pattern contained a conditional subpattern with only one branch (in
1035        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1036        computed the wrong minimum data length (which is of course zero for such
1037        subpatterns). This could cause incorrect "no match" results.
1039    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1040        the pattern is reset in the first branch, pcre_compile() failed with
1041        "internal error: code overflow at offset...". This happened only when
1042        the reset was to the original external option setting. (An optimization
1043        abstracts leading options settings into an external setting, which was the
1044        cause of this.)
1046    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1047        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1048        assertion pattern did not match (meaning that the assertion was true), it
1049        was incorrectly treated as false if the SKIP had been reached during the
1050        matching. This also applied to assertions used as conditions.
1052    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1053        assertion subpattern, including such a pattern used as a condition,
1054        unpredictable results occurred, instead of the error return
1057    5.  The C++ GlobalReplace function was not working like Perl for the special
1058        situation when an empty string is matched. It now does the fancy magic
1059        stuff that is necessary.
1061    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1062        removed. (These were left over from very, very early versions of PCRE.)
1064    7.  Some cosmetic changes to the code to make life easier when compiling it
1065        as part of something else:
1067        (a) Change DEBUG to PCRE_DEBUG.
1069        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1070            called "current" as "current_branch", to prevent a collision with the
1071            Linux macro when compiled as a kernel module.
1073        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1074            prevent a collision with the Linux macro when compiled as a kernel
1075            module.
1077    8.  In pcre_compile() there are some checks for integer overflows that used to
1078        cast potentially large values to (double). This has been changed to that
1079        when building, a check for int64_t is made, and if it is found, it is used
1080        instead, thus avoiding the use of floating point arithmetic. (There is no
1081        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1082        double.
1084    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1085        2005 (difference between two addresses compared to an unsigned value).
1087    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1088        custom one, because of the following reported problem in Windows:
1090          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1091              under Win32.
1092          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1093              therefore missing the function definition.
1094          - The compiler thus generates a "C" signature for the test function.
1095          - The linker fails to find the "C" function.
1096          - PCRE fails to configure if asked to do so against libbz2.
1098    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1099        messages were output:
1101          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1102          rerunning libtoolize, to keep the correct libtool macros in-tree.
1103          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1105        I have done both of these things.
1107    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1108        most of the time, it *can* run out if it is given a pattern that contains a
1109        runaway infinite recursion. I updated the discussion in the pcrestack man
1110        page.
1112    13. Now that we have gone to the x.xx style of version numbers, the minor
1113        version may start with zero. Using 08 or 09 is a bad idea because users
1114        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1115        interpreted as invalid octal numbers. I've updated the previous comment in
1116        configure.ac, and also added a check that gives an error if 08 or 09 are
1117        used.
1119    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1120        causing partial matching to fail when the end of the subject matched \W
1121        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1123    15. There were some discrepancies between the declarations in pcre_internal.h
1124        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1125        their definitions. The declarations used "const uschar *" and the
1126        definitions used USPTR. Even though USPTR is normally defined as "const
1127        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1128        reported that: "This difference in casting confuses some C++ compilers, for
1129        example, SunCC recognizes above declarations as different functions and
1130        generates broken code for hbpcre." I have changed the declarations to use
1131        USPTR.
1133    16. GNU libtool is named differently on some systems. The autogen.sh script now
1134        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1135        (FreeBSD).
1137    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1138        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1139        comment: "Figure out how to create a longlong from a string: strtoll and
1140        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1141        instance, but it only takes 2 args instead of 3!"
1143    18. A subtle bug concerned with back references has been fixed by a change of
1144        specification, with a corresponding code fix. A pattern such as
1145        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1146        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1147        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1148        same bug. Such groups have to be quantified to be useful, or contained
1149        inside another quantified group. (If there's no repetition, the reference
1150        can never match.) The problem arises because, having left the group and
1151        moved on to the rest of the pattern, a later failure that backtracks into
1152        the group uses the captured value from the final iteration of the group
1153        rather than the correct earlier one. I have fixed this in PCRE by forcing
1154        any group that contains a reference to itself to be an atomic group; that
1155        is, there cannot be any backtracking into it once it has completed. This is
1156        similar to recursive and subroutine calls.
1159    Version 8.00 19-Oct-09
1160    ----------------------
1162    1.  The table for translating pcre_compile() error codes into POSIX error codes
1163        was out-of-date, and there was no check on the pcre_compile() error code
1164        being within the table. This could lead to an OK return being given in
1165        error.
1167    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1168        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1169        in a Windows environment.
1171    3.  The pcregrep --count option prints the count for each file even when it is
1172        zero, as does GNU grep. However, pcregrep was also printing all files when
1173        --files-with-matches was added. Now, when both options are given, it prints
1174        counts only for those files that have at least one match. (GNU grep just
1175        prints the file name in this circumstance, but including the count seems
1176        more useful - otherwise, why use --count?) Also ensured that the
1177        combination -clh just lists non-zero counts, with no names.
1179    4.  The long form of the pcregrep -F option was incorrectly implemented as
1180        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1181        but it seems right to fix it, and I didn't think it was worth preserving
1182        the old behaviour.
1184    5.  The command line items --regex=pattern and --regexp=pattern were not
1185        recognized by pcregrep, which required --regex pattern or --regexp pattern
1186        (with a space rather than an '='). The man page documented the '=' forms,
1187        which are compatible with GNU grep; these now work.
1189    6.  No libpcreposix.pc file was created for pkg-config; there was just
1190        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1192    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1193        when UCP support is not needed, by modifying the Python script that
1194        generates it from Unicode data files. This should not matter if the module
1195        is correctly used as a library, but I received one complaint about 50K of
1196        unwanted data. My guess is that the person linked everything into his
1197        program rather than using a library. Anyway, it does no harm.
1199    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1200        was a minimum greater than 1 for a wide character in a possessive
1201        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1202        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1203        character. Chaos in the form of incorrect output or a compiling loop could
1204        result.
1206    9.  The restrictions on what a pattern can contain when partial matching is
1207        requested for pcre_exec() have been removed. All patterns can now be
1208        partially matched by this function. In addition, if there are at least two
1209        slots in the offset vector, the offset of the earliest inspected character
1210        for the match and the offset of the end of the subject are set in them when
1211        PCRE_ERROR_PARTIAL is returned.
1213    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1214        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1215        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1216        and may be more useful for multi-segment matching.
1218    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1219        used to be given if ever the end of the subject was reached; now it is
1220        given only if matching could not proceed because another character was
1221        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1222        string "Z", which now yields "no match" instead of "partial match". In the
1223        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1224        final character ended with (*FAIL).
1226    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1227        if the pattern had a "must contain" character that was already found in the
1228        earlier partial match, unless partial matching was again requested. For
1229        example, with the pattern /dog.(body)?/, the "must contain" character is
1230        "g". If the first part-match was for the string "dog", restarting with
1231        "sbody" failed. This bug has been fixed.
1233    13. The string returned by pcre_dfa_exec() after a partial match has been
1234        changed so that it starts at the first inspected character rather than the
1235        first character of the match. This makes a difference only if the pattern
1236        starts with a lookbehind assertion or \b or \B (\K is not supported by
1237        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1238        matching functions compatible, and I think it's the right thing to do.
1240    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1241        so that the demonstration program is easily available in environments where
1242        PCRE has not been installed from source.
1244    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1245        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1246        library.
1248    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1249        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1250        is not the first non-POSIX option to be added. Clearly some people find
1251        these options useful.
1253    17. If a caller to the POSIX matching function regexec() passes a non-zero
1254        value for nmatch with a NULL value for pmatch, the value of
1255        nmatch is forced to zero.
1257    18. RunGrepTest did not have a test for the availability of the -u option of
1258        the diff command, as RunTest does. It now checks in the same way as
1259        RunTest, and also checks for the -b option.
1261    19. If an odd number of negated classes containing just a single character
1262        interposed, within parentheses, between a forward reference to a named
1263        subpattern and the definition of the subpattern, compilation crashed with
1264        an internal error, complaining that it could not find the referenced
1265        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1266        [The bug was that it was starting one character too far in when skipping
1267        over the character class, thus treating the ] as data rather than
1268        terminating the class. This meant it could skip too much.]
1270    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1271        /g option in pcretest when the pattern contains \K, which makes it possible
1272        to have an empty string match not at the start, even when the pattern is
1273        anchored. Updated pcretest and pcredemo to use this option.
1275    21. If the maximum number of capturing subpatterns in a recursion was greater
1276        than the maximum at the outer level, the higher number was returned, but
1277        with unset values at the outer level. The correct (outer level) value is
1278        now given.
1280    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1281        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1282        make it do so. The string so far is captured, making this feature
1283        compatible with Perl.
1285    23. The tests have been re-organized, adding tests 11 and 12, to make it
1286        possible to check the Perl 5.10 features against Perl 5.10.
1288    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1289        pattern matches a fixed length string. PCRE did not allow this; now it
1290        does. Neither allows recursion.
1292    25. I finally figured out how to implement a request to provide the minimum
1293        length of subject string that was needed in order to match a given pattern.
1294        (It was back references and recursion that I had previously got hung up
1295        on.) This code has now been added to pcre_study(); it finds a lower bound
1296        to the length of subject needed. It is not necessarily the greatest lower
1297        bound, but using it to avoid searching strings that are too short does give
1298        some useful speed-ups. The value is available to calling programs via
1299        pcre_fullinfo().
1301    26. While implementing 25, I discovered to my embarrassment that pcretest had
1302        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1303        study optimizations had never been tested with that matching function.
1304        Oops. What is worse, even when it was passed study data, there was a bug in
1305        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1306        were also very few tests of studied patterns with pcre_dfa_exec().
1308    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1309        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1310        on the other side of the coin, they are no longer allowed to have different
1311        names, because these cannot be distinguished in PCRE, and this has caused
1312        confusion. (This is a difference from Perl.)
1314    28. When duplicate subpattern names are present (necessarily with different
1315        numbers, as required by 27 above), and a test is made by name in a
1316        conditional pattern, either for a subpattern having been matched, or for
1317        recursion in such a pattern, all the associated numbered subpatterns are
1318        tested, and the overall condition is true if the condition is true for any
1319        one of them. This is the way Perl works, and is also more like the way
1320        testing by number works.
1323    Version 7.9 11-Apr-09
1324    ---------------------
1326    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1327        (pcretest), all targets were linked against these libraries. This included
1328        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1329        libraries. This caused unwanted dependencies to be created. This problem
1330        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1331        pcretest is linked with readline.
1333    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1334        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1335        moved outside it again, because FALSE and TRUE are already defined in AIX,
1336        but BOOL is not.
1338    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1339        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1341    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1342        hyphens) following file names and line numbers when outputting matching
1343        lines. This is not true; no spaces are inserted. I have also clarified the
1344        wording for the --colour (or --color) option.
1346    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1347        was not coloured; this is different to GNU grep, so I have changed it to be
1348        the same.
1350    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1351        each matching line was coloured. Now it goes on to look for further matches
1352        of any of the test patterns, which is the same behaviour as GNU grep.
1354    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1355        doesn't make sense to accept an empty string match in pcregrep, so I have
1356        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1357        seems to be how GNU grep behaves.
1359    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1360        start or after a newline", because the conditional assertion was not being
1361        correctly handled. The rule now is that both the assertion and what follows
1362        in the first alternative must satisfy the test.
1364    9.  If auto-callout was enabled in a pattern with a conditional group whose
1365        condition was an assertion, PCRE could crash during matching, both with
1366        pcre_exec() and pcre_dfa_exec().
1368    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1369        used for matching.
1371    11. Unicode property support in character classes was not working for
1372        characters (bytes) greater than 127 when not in UTF-8 mode.
1374    12. Added the -M command line option to pcretest.
1376    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1378    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1380    16. Added comments and documentation about mis-use of no_arg in the C++
1381        wrapper.
1383    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1384        from Martin Jerabek that uses macro names for all relevant character and
1385        string constants.
1387    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1388        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1389        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1390        these, but not everybody uses configure.
1392    19. A conditional group that had only one branch was not being correctly
1393        recognized as an item that could match an empty string. This meant that an
1394        enclosing group might also not be so recognized, causing infinite looping
1395        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1396        with the subject "ab", where knowledge that the repeated group can match
1397        nothing is needed in order to break the loop.
1399    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1400        exec(), but without supplying a callout function, matching went wrong.
1402    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1403        leak if the size of the offset vector was greater than 30. When the vector
1404        is smaller, the saved offsets during recursion go onto a local stack
1405        vector, but for larger vectors malloc() is used. It was failing to free
1406        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1407        error, in fact).
1409    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1410        heapframe that is used only when UTF-8 support is enabled. This caused no
1411        problem, but was untidy.
1413    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1414        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1415        included within another project.
1417    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1418        slightly modified by me:
1420          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1421              not building pcregrep.
1423          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1424              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1426    25. Forward references, both numeric and by name, in patterns that made use of
1427        duplicate group numbers, could behave incorrectly or give incorrect errors,
1428        because when scanning forward to find the reference group, PCRE was not
1429        taking into account the duplicate group numbers. A pattern such as
1430        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1432    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1433        the feature of a custom pointer more persuasive (as requested by a user).
1435    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1436        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1437        pre-defined.
1439    28. Added support for (*UTF8) at the start of a pattern.
1441    29. Arrange for flags added by the "release type" setting in CMake to be shown
1442        in the configuration summary.
1445    Version 7.8 05-Sep-08
1446    ---------------------
1448    1.  Replaced UCP searching code with optimized version as implemented for Ad
1449        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1450        stage table and inline lookup instead of a function, giving speed ups of 2
1451        to 5 times on some simple patterns that I tested. Permission was given to
1452        distribute the MultiStage2.py script that generates the tables (it's not in
1453        the tarball, but is in the Subversion repository).
1455    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1456        scripts.
1458    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1459        a group with a zero qualifier. The result of the study could be incorrect,
1460        or the function might crash, depending on the pattern.
1462    4.  Caseless matching was not working for non-ASCII characters in back
1463        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1464        It now works when Unicode Property Support is available.
1466    5.  In pcretest, an escape such as \x{de} in the data was always generating
1467        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1468        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1469        truncation.
1471    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1473    7.  Added two (int) casts to pcregrep when printing the difference of two
1474        pointers, in case they are 64-bit values.
1476    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1477        test 2 if it fails.
1479    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1480        and a #define of that name to empty if it is not externally set. This is to
1481        allow users of MSVC to set it if necessary.
1483    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1484        the convenience functions in the pcre_get.c source file.
1486    11. An option change at the start of a pattern that had top-level alternatives
1487        could cause overwriting and/or a crash. This command provoked a crash in
1488        some environments:
1490          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1492        This potential security problem was recorded as CVE-2008-2371.
1494    12. For a pattern where the match had to start at the beginning or immediately
1495        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1496        pcre_dfa_exec() could read past the end of the passed subject if there was
1497        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1498        pcretest so that it places the subject at the end of its malloc-ed buffer.
1500    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1501        exec() might read past the end of the data buffer in UTF-8 mode.
1503    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1504        the data contained the byte 0x85 as part of a UTF-8 character within its
1505        first line. This applied both to normal and DFA matching.
1507    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1508        /^[^d]*?$/8 failed to match "abc".
1510    16. Added a missing copyright notice to pcrecpp_internal.h.
1512    17. Make it more clear in the documentation that values returned from
1513        pcre_exec() in ovector are byte offsets, not character counts.
1515    18. Tidied a few places to stop certain compilers from issuing warnings.
1517    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1518        supplied by Stefan Weber. I made a further small update for 7.8 because
1519        there is a change of source arrangements: the pcre_searchfuncs.c module is
1520        replaced by pcre_ucd.c.
1523    Version 7.7 07-May-08
1524    ---------------------
1526    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1527        a string to a long long, pretend we don't even have a long long." This is
1528        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1530    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1531        pre-7.6 versions, which defined a global no_arg variable instead of putting
1532        it in the RE class. (See also #8 below.)
1534    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1535        Lopes.
1537    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1539        (1) The include/exclude patterns were being applied to the whole pathnames
1540            of files, instead of just to the final components.
1542        (2) If there was more than one level of directory, the subdirectories were
1543            skipped unless they satisfied the include/exclude conditions. This is
1544            inconsistent with GNU grep (and could even be seen as contrary to the
1545            pcregrep specification - which I improved to make it absolutely clear).
1546            The action now is always to scan all levels of directory, and just
1547            apply the include/exclude patterns to regular files.
1549    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1550        --exclude_dir in the tests to avoid scanning .svn directories.
1552    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1553        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1554        doesn't support NULs in patterns.
1556    7.  Added some missing "const"s to declarations of static tables in
1557        pcre_compile.c and pcre_dfa_exec.c.
1559    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1560        caused by fix #2  above. (Subsequently also a second patch to fix the
1561        first patch. And a third patch - this was a messy problem.)
1563    9.  Applied Craig's patch to remove the use of push_back().
1565    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1566        matching function regexec().
1568    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1569        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1570        references. PCRE supports relative numbers with this syntax (I don't think
1571        Oniguruma does).
1573    12. Previously, a group with a zero repeat such as (...){0} was completely
1574        omitted from the compiled regex. However, this means that if the group
1575        was called as a subroutine from elsewhere in the pattern, things went wrong
1576        (an internal error was given). Such groups are now left in the compiled
1577        pattern, with a new opcode that causes them to be skipped at execution
1578        time.
1580    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1581        to the way PCRE behaves:
1583        (a) A lone ] character is dis-allowed (Perl treats it as data).
1585        (b) A back reference to an unmatched subpattern matches an empty string
1586            (Perl fails the current match path).
1588        (c) A data ] in a character class must be notated as \] because if the
1589            first data character in a class is ], it defines an empty class. (In
1590            Perl it is not possible to have an empty class.) The empty class []
1591            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1592            The negative empty class [^] matches any one character, independently
1593            of the DOTALL setting.
1595    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1596        non-existent subpattern following a character class starting with ']' and
1597        containing () gave an internal compiling error instead of "reference to
1598        non-existent subpattern". Fortunately, when the pattern did exist, the
1599        compiled code was correct. (When scanning forwards to check for the
1600        existencd of the subpattern, it was treating the data ']' as terminating
1601        the class, so got the count wrong. When actually compiling, the reference
1602        was subsequently set up correctly.)
1604    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1605        it was being rejected as not supported by pcre_dfa_exec(), even though
1606        other assertions are supported. I have made pcre_dfa_exec() support
1607        (*FAIL).
1609    16. The implementation of 13c above involved the invention of a new opcode,
1610        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1611        cannot be changed at match time, I realized I could make a small
1612        improvement to matching performance by compiling OP_ALLANY instead of
1613        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1614        on the OP_ANY path.
1616    17. Compiling pcretest on Windows with readline support failed without the
1617        following two fixes: (1) Make the unistd.h include conditional on
1618        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1620    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1621        ncurses library to be included for pcretest when ReadLine support is
1622        requested, but also to allow for it to be overridden. This patch came from
1623        Daniel Bergström.
1625    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1626        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1627        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1628        spotting this.
1631    Version 7.6 28-Jan-08
1632    ---------------------
1634    1.  A character class containing a very large number of characters with
1635        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1636        overflow.
1638    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1639        HAVE_LONG_LONG is not defined.
1641    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1642        bring them up to date and include new features. This patch includes:
1644        - Fixed PH's badly added libz and libbz2 support.
1645        - Fixed a problem with static linking.
1646        - Added pcredemo. [But later removed - see 7 below.]
1647        - Fixed dftables problem and added an option.
1648        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1649            HAVE_LONG_LONG.
1650        - Added readline support for pcretest.
1651        - Added an listing of the option settings after cmake has run.
1653    4.  A user submitted a patch to Makefile that makes it easy to create
1654        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1655        Makefile.am that cause it to include this special target, without
1656        affecting anything else. Note that the same mingw target plus all
1657        the other distribution libraries and programs are now supported
1658        when configuring with CMake (see 6 below) instead of with
1659        Configure/Make.
1661    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1662        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1663        exported in the Windows port". It has not yet been confirmed that the patch
1664        solves the problem, but it does no harm.
1666    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1667        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1668        with CMake, and also correct the comment about stack recursion.
1670    7.  Remove the automatic building of pcredemo from the ./configure system and
1671        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1672        of a program that users should build themselves after PCRE is installed, so
1673        building it automatically is not really right. What is more, it gave
1674        trouble in some build environments.
1676    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1679    Version 7.5 10-Jan-08
1680    ---------------------
1682    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1683        values in parens when parsing an RE using the C++ wrapper."
1685    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1686        Characters greater than 255 were excluded from the class instead of being
1687        included.
1689    3.  The same bug as (2) above applied to negated POSIX classes such as
1690        [:^space:].
1692    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1693        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1694        I have changed it.
1696    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1697        first named subpattern) and a construct such as (?&a) would reference the
1698        first named subpattern whose name started with "a" (in other words, the
1699        length check was missing). Both these problems are fixed. "Subpattern name
1700        expected" is now given for (?&) (a zero-length name), and this patch also
1701        makes it give the same error for \k'' (previously it complained that that
1702        was a reference to a non-existent subpattern).
1704    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1705        this is right because (?- can be followed by option settings as well as by
1706        digits. I have, however, made the messages clearer.
1708    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1709        than the number used in the conditional) now cause a compile-time error.
1710        This is actually not compatible with Perl, which accepts such patterns, but
1711        treats the conditional as always being FALSE (as PCRE used to), but it
1712        seems to me that giving a diagnostic is better.
1714    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1715        and messages.
1717    9.  Fix two occurrences of "backslash" in comments that should have been
1718        "backspace".
1720    10. Remove two redundant lines of code that can never be obeyed (their function
1721        was moved elsewhere).
1723    11. The program that makes PCRE's Unicode character property table had a bug
1724        which caused it to generate incorrect table entries for sequences of
1725        characters that have the same character type, but are in different scripts.
1726        It amalgamated them into a single range, with the script of the first of
1727        them. In other words, some characters were in the wrong script. There were
1728        thirteen such cases, affecting characters in the following ranges:
1730          U+002b0 - U+002c1
1731          U+0060c - U+0060d
1732          U+0061e - U+00612
1733          U+0064b - U+0065e
1734          U+0074d - U+0076d
1735          U+01800 - U+01805
1736          U+01d00 - U+01d77
1737          U+01d9b - U+01dbf
1738          U+0200b - U+0200f
1739          U+030fc - U+030fe
1740          U+03260 - U+0327f
1741          U+0fb46 - U+0fbb1
1742          U+10450 - U+1049d
1744    12. The -o option (show only the matching part of a line) for pcregrep was not
1745        compatible with GNU grep in that, if there was more than one match in a
1746        line, it showed only the first of them. It now behaves in the same way as
1747        GNU grep.
1749    13. If the -o and -v options were combined for pcregrep, it printed a blank
1750        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1751        does the same. The return code can be used to tell if there were any
1752        non-matching lines.
1754    14. Added --file-offsets and --line-offsets to pcregrep.
1756    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1757        infinitely looping recursion. The bug was that positive lookaheads were not
1758        being skipped when checking for a possible empty match (negative lookaheads
1759        and both kinds of lookbehind were skipped).
1761    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1762        inclusion of <windows.h> to before rather than after the definition of
1763        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1765    17. Specifying a possessive quantifier with a specific limit for a Unicode
1766        character property caused pcre_compile() to compile bad code, which led at
1767        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1768        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1769        caused the error; without that there was no problem.
1771    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1773    19. Added --enable-pcretest-libreadline.
1775    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1776        RE::GlobalReplace(). As a result, the number of replacements returned was
1777        double what it should be. I removed one of the increments, but Craig sent a
1778        later patch that removed the other one (the right fix) and added unit tests
1779        that check the return values (which was not done before).
1781    21. Several CMake things:
1783        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1784            the names libpcre and libpcreposix, not just pcre and pcreposix.
1786        (2) The above change means that pcretest and pcregrep are now correctly
1787            linked with the newly-built libraries, not previously installed ones.
1791    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1792        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1793        UTF-8 newline character). The key issue is that the pattern starts .*;
1794        this means that the match must be either at the beginning, or after a
1795        newline. The bug was in the code for advancing after a failed match and
1796        checking that the new position followed a newline. It was not taking
1797        account of UTF-8 characters correctly.
1799    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1800        character classes. PCRE was not treating the sequence [:...:] as a
1801        character class unless the ... were all letters. Perl, however, seems to
1802        allow any characters between [: and :], though of course it rejects as
1803        unknown any "names" that contain non-letters, because all the known class
1804        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1805        for example, whereas PCRE did not - it did not recognize a POSIX character
1806        class. This seemed a bit dangerous, so the code has been changed to be
1807        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1808        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1809        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1810        Perl does, and where it didn't before.
1812    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1813        Windows environments %n is disabled by default.
1816    Version 7.4 21-Sep-07
1817    ---------------------
1819    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1820        means that a class such as [\s] counted as "explicit reference to CR or
1821        LF". That isn't really right - the whole point of the change was to try to
1822        help when there was an actual mention of one of the two characters. So now
1823        the change happens only if \r or \n (or a literal CR or LF) character is
1824        encountered.
1826    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1827        of both had grown to the point where there were only 3 bits left.
1828        Fortunately, there was spare space in the data structure, and so I have
1829        moved the internal flags into a new 16-bit field to free up more option
1830        bits.
1832    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1833        but did not set the internal JCHANGED flag - either of these is enough to
1834        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1835        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1836        start sets both bits.
1838    4.  Added options (at build time, compile time, exec time) to change \R from
1839        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1841    5.  doc/pcresyntax.html was missing from the distribution.
1843    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1844        compatibility, even though it is no longer used.
1846    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1847        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1848        windows.h file is present (where different names are used). [This was
1849        reversed later after testing - see 16 below.]
1851    8.  Changed all #include <config.h> to #include "config.h". There were also
1852        some further <pcre.h> cases that I changed to "pcre.h".
1854    9.  When pcregrep was used with the --colour option, it missed the line ending
1855        sequence off the lines that it output.
1857    10. It was pointed out to me that arrays of string pointers cause lots of
1858        relocations when a shared library is dynamically loaded. A technique of
1859        using a single long string with a table of offsets can drastically reduce
1860        these. I have refactored PCRE in four places to do this. The result is
1861        dramatic:
1863          Originally:                          290
1864          After changing UCP table:            187
1865          After changing error message table:   43
1866          After changing table of "verbs"       36
1867          After changing table of Posix names   22
1869        Thanks to the folks working on Gregex for glib for this insight.
1871    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1872        unicode-properties was also set.
1874    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1876    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1877        checked only for CRLF.
1879    14. Added casts to pcretest.c to avoid compiler warnings.
1881    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1883    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1884        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1885        entirely. This removes changes made in 7 above.
1887    17. The CMake files have been updated, and there is now more information about
1888        building with CMake in the NON-UNIX-USE document.
1891    Version 7.3 28-Aug-07
1892    ---------------------
1894     1. In the rejigging of the build system that eventually resulted in 7.1, the
1895        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1896        brackets there is not right, since it causes compilers to look for an
1897        installed pcre.h, not the version that is in the source that is being
1898        compiled (which of course may be different). I have changed it back to:
1900          #include "pcre.h"
1902        I have a vague recollection that the change was concerned with compiling in
1903        different directories, but in the new build system, that is taken care of
1904        by the VPATH setting the Makefile.
1906     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1907        when the subject happened to end in the byte 0x85 (e.g. if the last
1908        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1909        characters but of course it shouldn't be taken as a newline when it is part
1910        of another character. The bug was that, for an unlimited repeat of . in
1911        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1912        characters when looking for a newline.
1914     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1916     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1917        in debug output.
1919     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1920        long printing in the pcrecpp unittest when running under MinGW.
1922     6. ESC_K was left out of the EBCDIC table.
1924     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1925        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1926        limit also applies to "virtual nesting" when a pattern is recursive, and in
1927        this case 1000 isn't so big. I have been able to remove this limit at the
1928        expense of backing off one optimization in certain circumstances. Normally,
1929        when pcre_exec() would call its internal match() function recursively and
1930        immediately return the result unconditionally, it uses a "tail recursion"
1931        feature to save stack. However, when a subpattern that can match an empty
1932        string has an unlimited repetition quantifier, it no longer makes this
1933        optimization. That gives it a stack frame in which to save the data for
1934        checking that an empty string has been matched. Previously this was taken
1935        from the 1000-entry workspace that had been reserved. So now there is no
1936        explicit limit, but more stack is used.
1938     8. Applied Daniel's patches to solve problems with the import/export magic
1939        syntax that is required for Windows, and which was going wrong for the
1940        pcreposix and pcrecpp parts of the library. These were overlooked when this
1941        problem was solved for the main library.
1943     9. There were some crude static tests to avoid integer overflow when computing
1944        the size of patterns that contain repeated groups with explicit upper
1945        limits. As the maximum quantifier is 65535, the maximum group length was
1946        set at 30,000 so that the product of these two numbers did not overflow a
1947        32-bit integer. However, it turns out that people want to use groups that
1948        are longer than 30,000 bytes (though not repeat them that many times).
1949        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1950        made it possible to implement the integer overflow checks in a much more
1951        dynamic way, which I have now done. The artificial limitation on group
1952        length has been removed - we now have only the limit on the total length of
1953        the compiled pattern, which depends on the LINK_SIZE setting.
1955    10. Fixed a bug in the documentation for get/copy named substring when
1956        duplicate names are permitted. If none of the named substrings are set, the
1957        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1958        empty string.
1960    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1961        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1962        because the ] is interpreted as the first data character and the
1963        terminating ] is not found. PCRE has been made compatible with Perl in this
1964        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1965        cause memory overwriting.
1967    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1968        string has been matched (to stop an infinite loop). It was not recognizing
1969        a conditional subpattern that could match an empty string if that
1970        subpattern was within another subpattern. For example, it looped when
1971        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1972        condition was not nested. This bug has been fixed.
1974    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1975        past the start of the subject in the presence of bytes with the top bit
1976        set, for example "\x8aBCD".
1978    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1979        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1981    14. Optimized (?!) to (*FAIL).
1983    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1984        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1985        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1986        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1987        does: it's just the validity check that is more restrictive.
1989    16. Inserted checks for integer overflows during escape sequence (backslash)
1990        processing, and also fixed erroneous offset values for syntax errors during
1991        backslash processing.
1993    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1994        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1996    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1997        caused an overrun.
1999    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2000        something other than just ASCII characters) inside a group that had an
2001        unlimited repeat caused a loop at compile time (while checking to see
2002        whether the group could match an empty string).
2004    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2005        [\P{Any}] did so. (Error in the code for printing property names.)
2007    21. An orphan \E inside a character class could cause a crash.
2009    22. A repeated capturing bracket such as (A)? could cause a wild memory
2010        reference during compilation.
2012    23. There are several functions in pcre_compile() that scan along a compiled
2013        expression for various reasons (e.g. to see if it's fixed length for look
2014        behind). There were bugs in these functions when a repeated \p or \P was
2015        present in the pattern. These operators have additional parameters compared
2016        with \d, etc, and these were not being taken into account when moving along
2017        the compiled data. Specifically:
2019        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2020            length.
2022        (b) An item such as \pL+ within a repeated group could cause crashes or
2023            loops.
2025        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2026            "reference to non-existent subpattern" error.
2028        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2030    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2031        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2033    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2035    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2036        character were causing crashes (broken optimization).
2038    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2039        \p or \P) caused a compile-time loop.
2041    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2042        break. For example, the unstudied pattern [\r\n]A does not match the string
2043        "\r\nA" because change 7.0/46 below moves the current point on by two
2044        characters after failing to match at the start. However, the pattern \nA
2045        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2046        the same is true. There doesn't seem any very clean way out of this, but
2047        what I have chosen to do makes the common cases work: PCRE now takes note
2048        of whether there can be an explicit match for \r or \n anywhere in the
2049        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2050        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2051        pattern has explicit CR or LF references.
2053    29. Added (*CR) etc for changing newline setting at start of pattern.
2056    Version 7.2 19-Jun-07
2057    ---------------------
2059     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2060        which is apparently normally available under Windows.
2062     2. Re-jig the pcregrep tests with different newline settings in an attempt
2063        to make them independent of the local environment's newline setting.
2065     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2067     4. Some of the "internals" tests were previously cut out when the link size
2068        was not 2, because the output contained actual offsets. The recent new
2069        "Z" feature of pcretest means that these can be cut out, making the tests
2070        usable with all link sizes.
2072     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2073        stack recursion. This gives a massive performance boost under BSD, but just
2074        a small improvement under Linux. However, it saves one field in the frame
2075        in all cases.
2077     6. Added more features from the forthcoming Perl 5.10:
2079        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2080            recursion call. It refers to the nth most recently opened parentheses.
2082        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2083            to be opened parentheses.
2085        (c) Conditions that refer to capturing parentheses can be specified
2086            relatively, for example, (?(-2)... or (?(+3)...
2088        (d) \K resets the start of the current match so that everything before
2089            is not part of it.
2091        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2093        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2094            reference syntax.
2096        (g) (?| introduces a group in which the numbering of parentheses in each
2097            alternative starts with the same number.
2099        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2101     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2104     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2105        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2106        for detecting groups that can match an empty string.
2108     9. A pattern with a very large number of alternatives (more than several
2109        hundred) was running out of internal workspace during the pre-compile
2110        phase, where pcre_compile() figures out how much memory will be needed. A
2111        bit of new cunning has reduced the workspace needed for groups with
2112        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2113        workspace instead of running out of the 4096 that are available.
2115    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2117    11. Applied patch from Google to remove an optimization that didn't quite work.
2118        The report of the bug said:
2120          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2121          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2122          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2124    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2125        it matched the wrong number of bytes.
2128    Version 7.1 24-Apr-07
2129    ---------------------
2131     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2132        that is more "standard", making use of automake and other Autotools. There
2133        is some re-arrangement of the files and adjustment of comments consequent
2134        on this.
2136     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2137        for recursive directory scanning broke on some systems because the files
2138        are not scanned in any specific order and on different systems the order
2139        was different. A call to "sort" has been inserted into RunGrepTest for the
2140        approprate test as a short-term fix. In the longer term there may be an
2141        alternative.
2143     3. I had an email from Eric Raymond about problems translating some of PCRE's
2144        man pages to HTML (despite the fact that I distribute HTML pages, some
2145        people do their own conversions for various reasons). The problems
2146        concerned the use of low-level troff macros .br and .in. I have therefore
2147        removed all such uses from the man pages (some were redundant, some could
2148        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2149        HTML has been updated to handle .nf/.fi and to complain if it encounters
2150        .br or .in.
2152     4. Updated comments in configure.ac that get placed in config.h.in and also
2153        arranged for config.h to be included in the distribution, with the name
2154        config.h.generic, for the benefit of those who have to compile without
2155        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2157     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2158        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2159        makevp.bat for the current PCRE, using the additional files
2160        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2162     6. A Windows user reported a minor discrepancy with test 2, which turned out
2163        to be caused by a trailing space on an input line that had got lost in his
2164        copy. The trailing space was an accident, so I've just removed it.
2166     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2167        that is needed.
2169     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2170        as "const" (a) because they are and (b) because it helps the PHP
2171        maintainers who have recently made a script to detect big data structures
2172        in the php code that should be moved to the .rodata section. I remembered
2173        to update Builducptable as well, so it won't revert if ucptable.h is ever
2174        re-created.
2176     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2177        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2178        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2179        support is not required. This saves 1.5-2K of code, which is important in
2180        some applications.
2182        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2183        so as not to refer to the tables, even though these functions will never be
2184        called when UTF-8 support is disabled. Otherwise there are problems with a
2185        shared library.
2187    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2189        (a) It was defining its arguments as char * instead of void *.
2191        (b) It was assuming that all moves were upwards in memory; this was true
2192            a long time ago when I wrote it, but is no longer the case.
2194        The emulated memove() is provided for those environments that have neither
2195        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2196        is clearly not the case, as these two bugs were recently reported.
2198    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2199        and Detrail to create the HTML documentation, the .txt form of the man
2200        pages, and it removes trailing spaces from listed files. It also creates
2201        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2202        case, it wraps all the #defines with #ifndefs. This script should be run
2203        before "make dist".
2205    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2206        with Unicode property support.
2208        (a) For a maximizing quantifier, if the two different cases of the
2209            character were of different lengths in their UTF-8 codings (there are
2210            some cases like this - I found 11), and the matching function had to
2211            back up over a mixture of the two cases, it incorrectly assumed they
2212            were both the same length.
2214        (b) When PCRE was configured to use the heap rather than the stack for
2215            recursion during matching, it was not correctly preserving the data for
2216            the other case of a UTF-8 character when checking ahead for a match
2217            while processing a minimizing repeat. If the check also involved
2218            matching a wide character, but failed, corruption could cause an
2219            erroneous result when trying to check for a repeat of the original
2220            character.
2222    13. Some tidying changes to the testing mechanism:
2224        (a) The RunTest script now detects the internal link size and whether there
2225            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2226            values substituted by "configure". (The RunGrepTest script already did
2227            this for UTF-8.) The configure.ac script no longer substitutes the
2228            relevant variables.
2230        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2231            with length and offset values. This means that the output is different
2232            for different internal link sizes. Test 2 is skipped for link sizes
2233            other than 2 because of this, bypassing the problem. Unfortunately,
2234            there was also a test in test 3 (the locale tests) that used /B and
2235            failed for link sizes other than 2. Rather than cut the whole test out,
2236            I have added a new /Z option to pcretest that replaces the length and
2237            offset values with spaces. This is now used to make test 3 independent
2238            of link size. (Test 2 will be tidied up later.)
2240    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2241        segmentation fault instead of returning the appropriate error message.
2243    15. In multiline mode when the newline sequence was set to "any", the pattern
2244        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2245        This doesn't seem right; it now treats the CRLF combination as the line
2246        ending, and so does not match in that case. It's only a pattern such as ^$
2247        that would hit this one: something like ^ABC$ would have failed after \r
2248        and then tried again after \r\n.
2250    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2251        in an attempt to make files that differ only in their line terminators
2252        compare equal. This works on Linux.
2254    17. Under certain error circumstances pcregrep might try to free random memory
2255        as it exited. This is now fixed, thanks to valgrind.
2257    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2258        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2259        was because its rules for how to advance for /g after matching an empty
2260        string at the end of a line did not allow for this case. They now check for
2261        it specially.
2263    20. pcretest is supposed to handle patterns and data of any length, by
2264        extending its buffers when necessary. It was getting this wrong when the
2265        buffer for a data line had to be extended.
2267    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2268        CRLF as a newline sequence.
2270    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2271        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2272        I have nevertheless tidied it up.
2274    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2276    24. Added a man page for pcre-config.
2279    Version 7.0 19-Dec-06
2280    ---------------------
2282     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2283        moving to gcc 4.1.1.
2285     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2286        sys/time.h, which is documented as needed for this function. It doesn't
2287        seem to matter on Linux, but it showed up on some releases of OS X.
2289     3. It seems that there are systems where bytes whose values are greater than
2290        127 match isprint() in the "C" locale. The "C" locale should be the
2291        default when a C program starts up. In most systems, only ASCII printing
2292        characters match isprint(). This difference caused the output from pcretest
2293        to vary, making some of the tests fail. I have changed pcretest so that:
2295        (a) When it is outputting text in the compiled version of a pattern, bytes
2296            other than 32-126 are always shown as hex escapes.
2298        (b) When it is outputting text that is a matched part of a subject string,
2299            it does the same, unless a different locale has been set for the match
2300            (using the /L modifier). In this case, it uses isprint() to decide.
2302     4. Fixed a major bug that caused incorrect computation of the amount of memory
2303        required for a compiled pattern when options that changed within the
2304        pattern affected the logic of the preliminary scan that determines the
2305        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2306        that the computed length was too small. The symptoms of this bug were
2307        either the PCRE error "internal error: code overflow" from pcre_compile(),
2308        or a glibc crash with a message such as "pcretest: free(): invalid next
2309        size (fast)". Examples of patterns that provoked this bug (shown in
2310        pcretest format) are:
2312          /(?-x: )/x
2313          /(?x)(?-x: \s*#\s*)/
2314          /((?i)[\x{c0}])/8
2315          /(?i:[\x{c0}])/8
2317        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2318        is now done differently.
2320     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2321        wrapper classes; (b) implement a new function in the C++ scanner that is
2322        more efficient than the old way of doing things because it avoids levels of
2323        recursion in the regex matching; (c) add a paragraph to the documentation
2324        for the FullMatch() function.
2326     6. The escape sequence \n was being treated as whatever was defined as
2327        "newline". Not only was this contrary to the documentation, which states
2328        that \n is character 10 (hex 0A), but it also went horribly wrong when
2329        "newline" was defined as CRLF. This has been fixed.
2331     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2332        was being set to -1 for the "end of line" case (supposedly a value that no
2333        character can have). Though this value is never used (the check for end of
2334        line is "zero bytes in current character"), it caused compiler complaints.
2335        I've changed it to 0xffffffff.
2337     8. In pcre_version.c, the version string was being built by a sequence of
2338        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2339        string (as it is for production releases) called a macro with an empty
2340        argument. The C standard says the result of this is undefined. The gcc
2341        compiler treats it as an empty string (which was what was wanted) but it is
2342        reported that Visual C gives an error. The source has been hacked around to
2343        avoid this problem.
2345     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2346        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2347        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2348        of them did).
2350    10. Originally, pcretest opened its input and output without "b"; then I was
2351        told that "b" was needed in some environments, so it was added for release
2352        5.0 to both the input and output. (It makes no difference on Unix-like
2353        systems.) Later I was told that it is wrong for the input on Windows. I've
2354        now abstracted the modes into two macros, to make it easier to fiddle with
2355        them, and removed "b" from the input mode under Windows.
2357    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2359    12. Added -help and --help to pcretest as an official way of being reminded
2360        of the options.
2362    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2363        and pcrecpp.cc because they annoy compilers at high warning levels.
2365    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2367    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2368        scripts are not C programs :-) and which was not noticed because it works
2369        on Linux.
2371    16. pcretest is supposed to handle any length of pattern and data line (as one
2372        line or as a continued sequence of lines) by extending its input buffer if
2373        necessary. This feature was broken for very long pattern lines, leading to
2374        a string of junk being passed to pcre_compile() if the pattern was longer
2375        than about 50K.
2377    17. I have done a major re-factoring of the way pcre_compile() computes the
2378        amount of memory needed for a compiled pattern. Previously, there was code
2379        that made a preliminary scan of the pattern in order to do this. That was
2380        OK when PCRE was new, but as the facilities have expanded, it has become
2381        harder and harder to keep it in step with the real compile phase, and there
2382        have been a number of bugs (see for example, 4 above). I have now found a
2383        cunning way of running the real compile function in a "fake" mode that
2384        enables it to compute how much memory it would need, while actually only
2385        ever using a few hundred bytes of working memory and without too many
2386        tests of the mode. This should make future maintenance and development
2387        easier. A side effect of this work is that the limit of 200 on the nesting
2388        depth of parentheses has been removed (though this was never a serious
2389        limitation, I suspect). However, there is a downside: pcre_compile() now
2390        runs more slowly than before (30% or more, depending on the pattern). I
2391        hope this isn't a big issue. There is no effect on runtime performance.
2393    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2394        newline (only possible for the last line of a file) and it was a
2395        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2397    19. Added additional timing features to pcretest. (1) The -tm option now times
2398        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2399        separate command line item, by a number that specifies the number of
2400        repeats to use when timing. The default is 50000; this gives better
2401        precision, but takes uncomfortably long for very large patterns.
2403    20. Extended pcre_study() to be more clever in cases where a branch of a
2404        subpattern has no definite first character. For example, (a*|b*)[cd] would
2405        previously give no result from pcre_study(). Now it recognizes that the
2406        first character must be a, b, c, or d.
2408    21. There was an incorrect error "recursive call could loop indefinitely" if
2409        a subpattern (or the entire pattern) that was being tested for matching an
2410        empty string contained only one non-empty item after a nested subpattern.
2411        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2412        incorrectly, because the \d was being skipped in the check.
2414    22. The pcretest program now has a new pattern option /B and a command line
2415        option -b, which is equivalent to adding /B to every pattern. This causes
2416        it to show the compiled bytecode, without the additional information that
2417        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2418        is the same as /B/I).
2420    23. A new optimization is now able automatically to treat some sequences such
2421        as a*b as a*+b. More specifically, if something simple (such as a character
2422        or a simple class like \d) has an unlimited quantifier, and is followed by
2423        something that cannot possibly match the quantified thing, the quantifier
2424        is automatically "possessified".
2426    24. A recursive reference to a subpattern whose number was greater than 39
2427        went wrong under certain circumstances in UTF-8 mode. This bug could also
2428        have affected the operation of pcre_study().
2430    25. Realized that a little bit of performance could be had by replacing
2431        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2433    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2435    27. Possessive quantifiers such as a++ were previously implemented by turning
2436        them into atomic groups such as ($>a+). Now they have their own opcodes,
2437        which improves performance. This includes the automatically created ones
2438        from 23 above.
2440    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2441        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2442        the first matched character to be a colon. This applied both to named and
2443        numbered groups.
2445    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2447    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2448        building PCRE on Windows easier, so I have included it in the distribution.
2450    31. There is now a check in pcretest against a ridiculously large number being
2451        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2452        loop, the loop is abandoned.
2454    32. Forward references to subpatterns in conditions such as (?(2)...) where
2455        subpattern 2 is defined later cause pcre_compile() to search forwards in
2456        the pattern for the relevant set of parentheses. This search went wrong
2457        when there were unescaped parentheses in a character class, parentheses
2458        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2460    33. "Subroutine" calls and backreferences were previously restricted to
2461        referencing subpatterns earlier in the regex. This restriction has now
2462        been removed.
2464    34. Added a number of extra features that are going to be in Perl 5.10. On the
2465        whole, these are just syntactic alternatives for features that PCRE had
2466        previously implemented using the Python syntax or my own invention. The
2467        other formats are all retained for compatibility.
2469        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2470            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2471            also .NET compatible.
2473        (b) A recursion or subroutine call to a named group can now be defined as
2474            (?&name) as well as (?P>name).
2476        (c) A backreference to a named group can now be defined as \k<name> or
2477            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2478            5.10, are also .NET compatible.
2480        (d) A conditional reference to a named group can now use the syntax
2481            (?(<name>) or (?('name') as well as (?(name).
2483        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2484            groups (named and numbered) that are never evaluated inline, but can be
2485            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2486            is always false. There may be only one alternative in such a group.
2488        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2489            as the simple (?(R). The condition is true only if the most recent
2490            recursion is that of the given number or name. It does not search out
2491            through the entire recursion stack.
2493        (g) The escape \gN or \g{N} has been added, where N is a positive or
2494            negative number, specifying an absolute or relative reference.
2496    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2497        some "unreachable code" warnings.
2499    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2500        things, this adds five new scripts.
2502    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2503        There were also incompatibilities regarding the handling of \Q..\E inside
2504        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2505        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2507    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2508        matches an empty string, and forcibly breaks the loop. There were bugs in
2509        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2510        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2511        separate and independent bugs (that affected different cases) have been
2512        fixed.
2514    39. Refactored the code to abolish the use of different opcodes for small
2515        capturing bracket numbers. This is a tidy that I avoided doing when I
2516        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2517        The new approach is not only tidier, it makes it possible to reduce the
2518        memory needed to fix the previous bug (38).
2520    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2521        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2522        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2523        mode.
2525    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2526        report.
2528    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2529        copy construction and assignment in the C++ wrapper.
2531    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2532        couple of bugs that could have given wrong results in the "--newline=crlf"
2533        case.
2535    44. Added a number of casts and did some reorganization of signed/unsigned int
2536        variables following suggestions from Dair Grant. Also renamed the variable
2537        "this" as "item" because it is a C++ keyword.
2539    45. Arranged for dftables to add
2541          #include "pcre_internal.h"
2543        to pcre_chartables.c because without it, gcc 4.x may remove the array
2544        definition from the final binary if PCRE is built into a static library and
2545        dead code stripping is activated.
2547    46. For an unanchored pattern, if a match attempt fails at the start of a
2548        newline sequence, and the newline setting is CRLF or ANY, and the next two
2549        characters are CRLF, advance by two characters instead of one.
2552    Version 6.7 04-Jul-06
2553    ---------------------
2555     1. In order to handle tests when input lines are enormously long, pcretest has
2556        been re-factored so that it automatically extends its buffers when
2557        necessary. The code is crude, but this _is_ just a test program. The
2558        default size has been increased from 32K to 50K.
2560     2. The code in pcre_study() was using the value of the re argument before
2561        testing it for NULL. (Of course, in any sensible call of the function, it
2562        won't be NULL.)
2564     3. The memmove() emulation function in pcre_internal.h, which is used on
2565        systems that lack both memmove() and bcopy() - that is, hardly ever -
2566        was missing a "static" storage class specifier.
2568     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2569        containing an extended class (one that cannot be represented by a bitmap
2570        because it contains high-valued characters or Unicode property items, e.g.
2571        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2572        pattern, but PCRE should not loop if you do not (it no longer does).
2573        [Detail: two cases were found: (a) a repeated subpattern containing an
2574        extended class; (b) a recursive reference to a subpattern that followed a
2575        previous extended class. It wasn't skipping over the extended class
2576        correctly when UTF-8 mode was not set.]
2578     5. A negated single-character class was not being recognized as fixed-length
2579        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2580        compile error "lookbehind assertion is not fixed length".
2582     6. The RunPerlTest auxiliary script was showing an unexpected difference
2583        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2584        write a Perl script that can interpret lines of an input file either as
2585        byte characters or as UTF-8, which is what "perltest" was being required to
2586        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2587        can't do is switch easily at run time between having the "use utf8;" pragma
2588        or not. In the end, I fudged it by using the RunPerlTest script to insert
2589        "use utf8;" explicitly for the UTF-8 tests.
2591     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2592        the end of the subject string, contrary to the documentation and to what
2593        Perl does. This was true of both matching functions. Now it matches only at
2594        the start of the subject and immediately after *internal* newlines.
2596     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2597        a pointer to an int instead of a pointer to an unsigned long int. This
2598        caused problems on 64-bit systems.
2600     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2601        instance of the 'standard' template library not being so standard".
2603    10. There was no check on the number of named subpatterns nor the maximum
2604        length of a subpattern name. The product of these values is used to compute
2605        the size of the memory block for a compiled pattern. By supplying a very
2606        long subpattern name and a large number of named subpatterns, the size
2607        computation could be caused to overflow. This is now prevented by limiting
2608        the length of names to 32 characters, and the number of named subpatterns
2609        to 10,000.
2611    11. Subpatterns that are repeated with specific counts have to be replicated in
2612        the compiled pattern. The size of memory for this was computed from the
2613        length of the subpattern and the repeat count. The latter is limited to
2614        65535, but there was no limit on the former, meaning that integer overflow
2615        could in principle occur. The compiled length of a repeated subpattern is
2616        now limited to 30,000 bytes in order to prevent this.
2618    12. Added the optional facility to have named substrings with the same name.
2620    13. Added the ability to use a named substring as a condition, using the
2621        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2622        are numbers (not recommended). Forward references are permitted.
2624    14. Added forward references in named backreferences (if you see what I mean).
2626    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2627        pattern could run off the end of the subject. For example, the pattern
2628        "(?s)(.{1,5})"8 did this with the subject "ab".
2630    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2631        PCRE_CASELESS was set when matching characters that were quantified with ?
2632        or *.
2634    17. A character class other than a single negated character that had a minimum
2635        but no maximum quantifier - for example [ab]{6,} - was not handled
2636        correctly by pce_dfa_exec(). It would match only one character.
2638    18. A valid (though odd) pattern that looked like a POSIX character
2639        class but used an invalid character after [ (for example [[,abc,]]) caused
2640        pcre_compile() to give the error "Failed: internal error: code overflow" or
2641        in some cases to crash with a glibc free() error. This could even happen if
2642        the pattern terminated after [[ but there just happened to be a sequence of
2643        letters, a binary zero, and a closing ] in the memory that followed.
2645    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2646        over the years. Originally (before any Unicode support), just the bottom 8
2647        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2648        output from "man perlunicode" includes this:
2650          The regular expression compiler produces polymorphic opcodes.  That
2651          is, the pattern adapts to the data and automatically switches to
2652          the Unicode character scheme when presented with Unicode data--or
2653          instead uses a traditional byte scheme when presented with byte
2654          data.
2656        Sadly, a wide octal escape does not cause a switch, and in a string with
2657        no other multibyte characters, these octal escapes are treated as before.
2658        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2659        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2660        Unicode string.
2662        I have not perpetrated such confusion in PCRE. Up till now, it took just
2663        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2664        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2665        translate to the appropriate multibyte character.
2667    29. Applied some refactoring to reduce the number of warnings from Microsoft
2668        and Borland compilers. This has included removing the fudge introduced
2669        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2670        a warning about an unused variable.
2672    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2673        characters since release 4.0, because Perl (from release 5.004) does not.
2674        [Or at least, is documented not to: some releases seem to be in conflict
2675        with the documentation.] However, when a pattern was studied with
2676        pcre_study() and all its branches started with \s, PCRE still included VT
2677        as a possible starting character. Of course, this did no harm; it just
2678        caused an unnecessary match attempt.
2680    22. Removed a now-redundant internal flag bit that recorded the fact that case
2681        dependency changed within the pattern. This was once needed for "required
2682        byte" processing, but is no longer used. This recovers a now-scarce options
2683        bit. Also moved the least significant internal flag bit to the most-
2684        significant bit of the word, which was not previously used (hangover from
2685        the days when it was an int rather than a uint) to free up another bit for
2686        the future.
2688    23. Added support for CRLF line endings as well as CR and LF. As well as the
2689        default being selectable at build time, it can now be changed at runtime
2690        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2691        specify that it is scanning data with non-default line endings.
2693    24. Changed the definition of CXXLINK to make it agree with the definition of
2694        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2696    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2697        recursions. This makes a big different to stack usage for some patterns.
2699    26. If a subpattern containing a named recursion or subroutine reference such
2700        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2701        the space required for the compiled pattern went wrong and gave too small a
2702        value. Depending on the environment, this could lead to "Failed: internal
2703        error: code overflow at offset 49" or "glibc detected double free or
2704        corruption" errors.
2706    27. Applied patches from Google (a) to support the new newline modes and (b) to
2707        advance over multibyte UTF-8 characters in GlobalReplace.
2709    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2710        difference for some implementation of PCRE in some Windows version.
2712    29. Added some extra testing facilities to pcretest:
2714        \q<number>   in a data line sets the "match limit" value
2715        \Q<number>   in a data line sets the "match recursion limt" value
2716        -S <number>  sets the stack size, where <number> is in megabytes
2718        The -S option isn't available for Windows.
2721    Version 6.6 06-Feb-06
2722    ---------------------
2724     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2725        in pcreposix.h. I have copied the definition from pcre.h.
2727     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2728        because pcre.h is no longer a built file.
2730     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2731        not normally included in the compiled code.
2734    Version 6.5 01-Feb-06
2735    ---------------------
2737     1. When using the partial match feature with pcre_dfa_exec(), it was not
2738        anchoring the second and subsequent partial matches at the new starting
2739        point. This could lead to incorrect results. For example, with the pattern
2740        /1234/, partially matching against "123" and then "a4" gave a match.
2742     2. Changes to pcregrep:
2744        (a) All non-match returns from pcre_exec() were being treated as failures
2745            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2746            error message is output. Some extra information is given for the
2747            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2748            probably the only errors that are likely to be caused by users (by
2749            specifying a regex that has nested indefinite repeats, for instance).
2750            If there are more than 20 of these errors, pcregrep is abandoned.
2752        (b) A binary zero was treated as data while matching, but terminated the
2753            output line if it was written out. This has been fixed: binary zeroes
2754            are now no different to any other data bytes.
2756        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2757            used to set a locale for matching. The --locale=xxxx long option has
2758            been added (no short equivalent) to specify a locale explicitly on the
2759            pcregrep command, overriding the environment variables.
2761        (d) When -B was used with -n, some line numbers in the output were one less
2762            than they should have been.
2764        (e) Added the -o (--only-matching) option.
2766        (f) If -A or -C was used with -c (count only), some lines of context were
2767            accidentally printed for the final match.
2769        (g) Added the -H (--with-filename) option.
2771        (h) The combination of options -rh failed to suppress file names for files
2772            that were found from directory arguments.
2774        (i) Added the -D (--devices) and -d (--directories) options.
2776        (j) Added the -F (--fixed-strings) option.
2778        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2780        (l) Added the --colo(u)r option.
2782        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2783            is not present by default.
2785     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2786        items such as (?R) or (?1), when the recursion could match a number of
2787        alternatives. If it matched one of the alternatives, but subsequently,
2788        outside the recursion, there was a failure, the code tried to back up into
2789        the recursion. However, because of the way PCRE is implemented, this is not
2790        possible, and the result was an incorrect result from the match.
2792        In order to prevent this happening, the specification of recursion has
2793        been changed so that all such subpatterns are automatically treated as
2794        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2796     4. I had overlooked the fact that, in some locales, there are characters for
2797        which isalpha() is true but neither isupper() nor islower() are true. In
2798        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2799        and ordfeminine) are like this. This affected the treatment of \w and \W
2800        when they appeared in character classes, but not when they appeared outside
2801        a character class. The bit map for "word" characters is now created
2802        separately from the results of isalnum() instead of just taking it from the
2803        upper, lower, and digit maps. (Plus the underscore character, of course.)
2805     5. The above bug also affected the handling of POSIX character classes such as
2806        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2807        permanent tables. Instead, the bit maps for such a class were previously
2808        created as the appropriate unions of the upper, lower, and digit bitmaps.
2809        Now they are created by subtraction from the [[:word:]] class, which has
2810        its own bitmap.
2812     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2813        It is created by subtracting the vertical space characters (\x09, \x0a,
2814        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2815        subtraction was done in the overall bitmap for a character class, meaning
2816        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2817        be recognized. This bug has been fixed.
2819     7. Patches from the folks at Google:
2821          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2822          real life, but is still worth protecting against".
2824          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2825          regular expressions".
2827          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2828          have it.
2830          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2831          "configure" and the latter not, in order to fix a problem somebody had
2832          with compiling the Arg class on HP-UX.
2834          (e) Improve the error-handling of the C++ wrapper a little bit.
2836          (f) New tests for checking recursion limiting.
2838     8. The pcre_memmove() function, which is used only if the environment does not
2839        have a standard memmove() function (and is therefore rarely compiled),
2840        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2841        returning a result (though PCRE never actually uses the result).
2843     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2844        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2845        returned instead of calling malloc() with an overflowing number that would
2846        most likely cause subsequent chaos.
2848    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2850    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2851        with this option is matched, the nmatch and pmatch options of regexec() are
2852        ignored.
2854    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2855        provided in case anyone wants to the the POSIX interface with UTF-8
2856        strings.
2858    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2859        C++ linking (needed for some HP-UX environments).
2861    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2862        (unused parameter) and in the pcre_printint() function (omitted "default"
2863        switch label when the default is to do nothing).
2865    15. Added some code to make it possible, when PCRE is compiled as a C++
2866        library, to replace subject pointers for pcre_exec() with a smart pointer
2867        class, thus making it possible to process discontinuous strings.
2869    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2870        much the same function. They were added by different people who were trying
2871        to make PCRE easy to compile on non-Unix systems. It has been suggested
2872        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2873        for compiling on Windows systems. I have therefore replaced it with
2874        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2875        defaults to "extern" for C or "extern C" for C++, which works fine on
2876        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2877        SCOPE with something explicit in config.h. In addition:
2879        (a) pcreposix.h still had just "extern" instead of either of these macros;
2880            I have replaced it with PCRE_DATA_SCOPE.
2882        (b) Functions such as _pcre_xclass(), which are internal to the library,
2883            but external in the C sense, all had PCRE_EXPORT in their definitions.
2884            This is apparently wrong for the Windows case, so I have removed it.
2885            (It makes no difference on Unix-like systems.)
2887    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2888        of recursive calls to match(). This is different to MATCH_LIMIT because
2889        that limits the total number of calls to match(), not all of which increase
2890        the depth of recursion. Limiting the recursion depth limits the amount of
2891        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2892        when PCRE is compiled, and changed at run time. A patch from Google adds
2893        this functionality to the C++ interface.
2895    18. Changes to the handling of Unicode character properties:
2897        (a) Updated the table to Unicode 4.1.0.
2899        (b) Recognize characters that are not in the table as "Cn" (undefined).
2901        (c) I revised the way the table is implemented to a much improved format
2902            which includes recognition of ranges. It now supports the ranges that
2903            are defined in UnicodeData.txt, and it also amalgamates other
2904            characters into ranges. This has reduced the number of entries in the
2905            table from around 16,000 to around 3,000, thus reducing its size
2906            considerably. I realized I did not need to use a tree structure after
2907            all - a binary chop search is just as efficient. Having reduced the
2908            number of entries, I extended their size from 6 bytes to 8 bytes to
2909            allow for more data.
2911        (d) Added support for Unicode script names via properties such as \p{Han}.
2913    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2914        matching that character.
2916    20. When matching a repeated Unicode property with a minimum greater than zero,
2917        (for example \pL{2,}), PCRE could look past the end of the subject if it
2918        reached it while seeking the minimum number of characters. This could
2919        happen only if some of the characters were more than one byte long, because
2920        there is a check for at least the minimum number of bytes.
2922    21. Refactored the implementation of \p and \P so as to be more general, to
2923        allow for more different types of property in future. This has changed the
2924        compiled form incompatibly. Anybody with saved compiled patterns that use
2925        \p or \P will have to recompile them.
2927    22. Added "Any" and "L&" to the supported property types.
2929    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2930        but give a compile time error if the value is greater than 0xff.
2932    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2933        accidentally not being installed or uninstalled.
2935    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2936        made were to insert the current release number. This seemed silly, because
2937        it made things harder for people building PCRE on systems that don't run
2938        "configure". I have turned pcre.h into a distributed file, no longer built
2939        by "configure", with the version identification directly included. There is
2940        no longer a pcre.h.in file.
2942        However, this change necessitated a change to the pcre-config script as
2943        well. It is built from pcre-config.in, and one of the substitutions was the
2944        release number. I have updated configure.ac so that ./configure now finds
2945        the release number by grepping pcre.h.
2947    26. Added the ability to run the tests under valgrind.
2950    Version 6.4 05-Sep-05
2951    ---------------------
2953     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2954        "--" to be printed when multiple files were scanned, even when none of the
2955        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2956        consider it to be a bug, and have restored the previous behaviour.
2958     2. A couple of code tidies to get rid of compiler warnings.
2960     3. The pcretest program used to cheat by referring to symbols in the library
2961        whose names begin with _pcre_. These are internal symbols that are not
2962        really supposed to be visible externally, and in some environments it is
2963        possible to suppress them. The cheating is now confined to including
2964        certain files from the library's source, which is a bit cleaner.
2966     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2967        file's purpose clearer.
2969     5. Reorganized pcre_ucp_findchar().
2972    Version 6.3 15-Aug-05
2973    ---------------------
2975     1. The file libpcre.pc.in did not have general read permission in the tarball.
2977     2. There were some problems when building without C++ support:
2979        (a) If C++ support was not built, "make install" and "make test" still
2980            tried to test it.
2982        (b) There were problems when the value of CXX was explicitly set. Some
2983            changes have been made to try to fix these, and ...
2985        (c) --disable-cpp can now be used to explicitly disable C++ support.
2987        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2988            backslash in a target when C++ was disabled. This confuses some
2989            versions of "make", apparently. Using an intermediate variable solves
2990            this. (Same for CPP_LOBJ.)
2992     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2993        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2994        necessary on certain architectures.
2996     4. Added a setting of -export-symbols-regex to the link command to remove
2997        those symbols that are exported in the C sense, but actually are local
2998        within the library, and not documented. Their names all begin with
2999        "_pcre_". This is not a perfect job, because (a) we have to except some
3000        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3001        available (and never for static libraries). I have made a note to try to
3002        find a way round (a) in the future.
3005    Version 6.2 01-Aug-05
3006    ---------------------
3008     1. There was no test for integer overflow of quantifier values. A construction
3009        such as {1111111111111111} would give undefined results. What is worse, if
3010        a minimum quantifier for a parenthesized subpattern overflowed and became
3011        negative, the calculation of the memory size went wrong. This could have
3012        led to memory overwriting.
3014     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3016     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3017        operating environments where this matters.
3019     4. Applied Giuseppe Maxia's patch to add additional features for controlling
3020        PCRE options from within the C++ wrapper.
3022     5. Named capturing subpatterns were not being correctly counted when a pattern
3023        was compiled. This caused two problems: (a) If there were more than 100
3024        such subpatterns, the calculation of the memory needed for the whole
3025        compiled pattern went wrong, leading to an overflow error. (b) Numerical
3026        back references of the form \12, where the number was greater than 9, were
3027        not recognized as back references, even though there were sufficient
3028        previous subpatterns.
3030     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3031        versions of gcc, e.g. 2.95.4.
3034    Version 6.1 21-Jun-05
3035    ---------------------
3037     1. There was one reference to the variable "posix" in pcretest.c that was not
3038        surrounded by "#if !defined NOPOSIX".
3040     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3041        the cross-check on the old pcre_info() function, for the benefit of the
3042        cut-down version of PCRE that is currently imported into Exim.
3044     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3045        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3046        patterns that start (?i) but I don't think that matters. The use of (?i) is
3047        just an example; this all applies to the other options as well.
3049     4. Since libtool seems to echo the compile commands it is issuing, the output
3050        from "make" can be reduced a bit by putting "@" in front of each libtool
3051        compile command.
3053     5. Patch from the folks at Google for configure.in to be a bit more thorough
3054        in checking for a suitable C++ installation before trying to compile the
3055        C++ stuff. This should fix a reported problem when a compiler was present,
3056        but no suitable headers.
3058     6. The man pages all had just "PCRE" as their title. I have changed them to
3059        be the relevant file name. I have also arranged that these names are
3060        retained in the file doc/pcre.txt, which is a concatenation in text format
3061        of all the man pages except the little individual ones for each function.
3063     7. The NON-UNIX-USE file had not been updated for the different set of source
3064        files that come with release 6. I also added a few comments about the C++
3065        wrapper.
3068  Version 6.0 07-Jun-05  Version 6.0 07-Jun-05
3069  ---------------------  ---------------------

Removed from v.77  
changed lines
  Added in v.987

  ViewVC Help
Powered by ViewVC 1.1.5