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 1009 by zherczeg, Wed Aug 22 12:01:22 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.
13    3.  Single character iterator optimizations in the JIT compiler.
15    4.  Improved JIT compiler optimizations for character ranges.
17    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
18        Reported by Giuseppe D'Angelo.
20    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
21        of a line, was being set incorrectly in cases where .* appeared inside
22        atomic brackets at the start of a pattern, or where there was a subsequent
23        *PRUNE or *SKIP.
25    7.  Improved instruction cache flush for POWER/PowerPC.
26        Patch by Daniel Richard G.
28    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
29        grep:
31        (a) There is now no limit to the number of patterns to be matched.
33        (b) An error is given if a pattern is too long.
35        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
36            are now supported.
38        (d) --exclude-from and --include-from (multiple use) have been added.
40        (e) Exclusions and inclusions now apply to all files and directories, not
41            just to those obtained from scanning a directory recursively.
43        (f) Multiple uses of -f and --file-list are now supported.
45        (g) In a Windows environment, the default for -d has been changed from
46            "read" (the GNU grep default) to "skip", because otherwise the presence
47            of a directory in the file list provokes an error.
49        (h) The documentation has been revised and clarified in places.
51    9.  Improve the matching speed of capturing brackets.
54    Version 8.31 06-July-2012
55    -------------------------
57    1.  Fixing a wrong JIT test case and some compiler warnings.
59    2.  Removed a bashism from the RunTest script.
61    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
62        to unsigned type, result still unsigned" that was given by an MS compiler
63        on encountering the code "-sizeof(xxx)".
65    4.  Partial matching support is added to the JIT compiler.
67    5.  Fixed several bugs concerned with partial matching of items that consist
68        of more than one character:
70        (a) /^(..)\1/ did not partially match "aba" because checking references was
71            done on an "all or nothing" basis. This also applied to repeated
72            references.
74        (b) \R did not give a hard partial match if \r was found at the end of the
75            subject.
77        (c) \X did not give a hard partial match after matching one or more
78            characters at the end of the subject.
80        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
81            a partial match for the string "\r".
83        (e) When newline was set to CRLF, the metacharacter "." did not recognize
84            a partial match for a CR character at the end of the subject string.
86    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
87        running pcretest, the text "(JIT)" added to the output whenever JIT is
88        actually used to run the match.
90    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
91        or /S+[+] with a digit between 1 and 7.
93    8.  OP_NOT now supports any UTF character not just single-byte ones.
95    9.  (*MARK) control verb is now supported by the JIT compiler.
97    10. The command "./RunTest list" lists the available tests without actually
98        running any of them. (Because I keep forgetting what they all are.)
102    12. Applied a (slightly modified) user-supplied patch that improves performance
103        when the heap is used for recursion (compiled with --disable-stack-for-
104        recursion). Instead of malloc and free for each heap frame each time a
105        logical recursion happens, frames are retained on a chain and re-used where
106        possible. This sometimes gives as much as 30% improvement.
108    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
109        call.
111    14. As documented, (*COMMIT) is now confined to within a positive assertion.
113    15. It is now possible to link pcretest with libedit as an alternative to
114        libreadline.
116    16. (*COMMIT) control verb is now supported by the JIT compiler.
118    17. The Unicode data tables have been updated to Unicode 6.1.0.
120    18. Added --file-list option to pcregrep.
122    19. Added binary file support to pcregrep, including the -a, --binary-files,
123        -I, and --text options.
125    20. The madvise function is renamed for posix_madvise for QNX compatibility
126        reasons. Fixed by Giuseppe D'Angelo.
128    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
130    22. Changed the option for creating symbolic links for 16-bit man pages from
131        -s to -sf so that re-installing does not cause issues.
133    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
135    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
136        matches in certain environments (the workspace was not being correctly
137        retained). Also added to pcre_dfa_exec() a simple plausibility check on
138        some of the workspace data at the beginning of a restart.
140    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
141        was not doing so when it should - probably a typo introduced by SVN 528
142        (change 8.10/14).
144    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
145        \w+ when the character tables indicated that \x{c4} was a word character.
146        There were several related cases, all because the tests for doing a table
147        lookup were testing for characters less than 127 instead of 255.
149    27. If a pattern contains capturing parentheses that are not used in a match,
150        their slots in the ovector are set to -1. For those that are higher than
151        any matched groups, this happens at the end of processing. In the case when
152        there were back references that the ovector was too small to contain
153        (causing temporary malloc'd memory to be used during matching), and the
154        highest capturing number was not used, memory off the end of the ovector
155        was incorrectly being set to -1. (It was using the size of the temporary
156        memory instead of the true size.)
158    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
159        ovector size, it uses memory at the end of the block that it has got.
161    29. Check for an overlong MARK name and give an error at compile time. The
162        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
164    30. JIT compiler update.
166    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
167        Rill for the patch.
169    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
171    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
173    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
174        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
175        was enabled.
177    35. Improve JIT code generation for greedy plus quantifier.
179    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
180        1 to "aa" instead of to an empty string. The bug affected repeated groups
181        that could potentially match an empty string.
183    37. Optimizing single character iterators in JIT.
185    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
186        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
187        codepoints that are too big for the mode are faulted, and in a UTF mode,
188        disallowed codepoints are also faulted.
190    39. If PCRE was compiled with UTF support, in three places in the DFA
191        matcher there was code that should only have been obeyed in UTF mode, but
192        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
193        processing when bytes with values greater than 127 were present. In 16-bit
194        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
195        both cases the values are those that cannot be the first data item in a UTF
196        character. The three items that might have provoked this were recursions,
197        possessively repeated groups, and atomic groups.
199    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
200        and pcregrep, because some OS require shared objects to be explicitly
201        passed to ld, causing the link step to fail if they are not.
203    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
204        mode, patterns that started with \h* or \R* might be incorrectly matched.
207    Version 8.30 04-February-2012
208    -----------------------------
210    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
211        name is defined in ctype.h.
213    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
214        only in quite long subpatterns.
216    3.  Removed the function pcre_info(), which has been obsolete and deprecated
217        since it was replaced by pcre_fullinfo() in February 2000.
219    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
220        match a (*MARK), and the match failed at the start of the subject, a
221        reference to memory before the start of the subject could occur. This bug
222        was introduced by fix 17 of release 8.21.
224    5.  A reference to an unset group with zero minimum repetition was giving
225        totally wrong answers (in non-JavaScript-compatibility mode). For example,
226        /(another)?(\1?)test/ matched against "hello world test". This bug was
227        introduced in release 8.13.
229    6.  Add support for 16-bit character strings (a large amount of work involving
230        many changes and refactorings).
232    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
233        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
234        from a file.
236    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
237        rounding is not applied in this particular case).
239    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
240        if they appear, or are escaped, in patterns.
242    10. Get rid of a number of -Wunused-but-set-variable warnings.
244    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
245        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
246        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
247        also returns the mark "x". This bug applied to capturing parentheses,
248        non-capturing parentheses, and atomic parentheses. It also applied to some
249        assertions.
251    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
252        information out of configure.ac instead of relying on pcre.h.generic, which
253        is not stored in the repository.
255    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
256        -lreadline.
258    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
260    15. Applied Graycode's patch to put the top-level frame on the stack rather
261        than the heap when not using the stack for recursion. This gives a
262        performance improvement in many cases when recursion is not deep.
264    16. Experimental code added to "pcretest -C" to output the stack frame size.
267    Version 8.21 12-Dec-2011
268    ------------------------
270    1.  Updating the JIT compiler.
272    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
273        are added as well.
275    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
276        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
277        calling _pcre_jit_exec. Some extra comments are added.
279    4.  (*MARK) settings inside atomic groups that do not contain any capturing
280        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
281        was introduced by change 18 for 8.20.
283    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
284        ECMA-262 standard.
286    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
287        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
288        This bug was probably introduced by change 9 of 8.13.
290    7.  While fixing 6 above, I noticed that a number of other items were being
291        incorrectly rejected as "not fixed length". This arose partly because newer
292        opcodes had not been added to the fixed-length checking code. I have (a)
293        corrected the bug and added tests for these items, and (b) arranged for an
294        error to occur if an unknown opcode is encountered while checking for fixed
295        length instead of just assuming "not fixed length". The items that were
296        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
297        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
298        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
300    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
301        being incorrectly compiled and would have given unpredicatble results.
303    9.  A possessively repeated subpattern with minimum repeat count greater than
304        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
305        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
306        the first (A) could occur when it should not.
308    10. Add a cast and remove a redundant test from the code.
310    11. JIT should use pcre_malloc/pcre_free for allocation.
312    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
313        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
314        is anything other than /usr, -L is still shown).
316    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
318    14. Perl does not support \N without a following name in a [] class; PCRE now
319        also gives an error.
321    15. If a forward reference was repeated with an upper limit of around 2000,
322        it caused the error "internal error: overran compiling workspace". The
323        maximum number of forward references (including repeats) was limited by the
324        internal workspace, and dependent on the LINK_SIZE. The code has been
325        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
326        the default depends on LINK_SIZE. There is a new upper limit (for safety)
327        of around 200,000 forward references. While doing this, I also speeded up
328        the filling in of repeated forward references.
330    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
331        incorrectly expecting the subject to contain another "a" after the start.
333    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
334        in the match, the SKIP should be ignored. This was not happening; instead
335        the SKIP was being treated as NOMATCH. For patterns such as
336        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
337        tested.
339    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
340        now much more compatible with Perl, in particular in cases where the result
341        is a non-match for a non-anchored pattern. For example, if
342        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
343        "m", where previously it did not return a name. A side effect of this
344        change is that for partial matches, the last encountered mark name is
345        returned, as for non matches. A number of tests that were previously not
346        Perl-compatible have been moved into the Perl-compatible test files. The
347        refactoring has had the pleasing side effect of removing one argument from
348        the match() function, thus reducing its stack requirements.
350    19. If the /S+ option was used in pcretest to study a pattern using JIT,
351        subsequent uses of /S (without +) incorrectly behaved like /S+.
353    21. Retrieve executable code size support for the JIT compiler and fixing
354        some warnings.
356    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
357        not work when the shorter character appeared right at the end of the
358        subject string.
360    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
361        systems.
363    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
364        output it when the /M option is used in pcretest.
366    25. The CheckMan script was not being included in the distribution. Also, added
367        an explicit "perl" to run Perl scripts from the PrepareRelease script
368        because this is reportedly needed in Windows.
370    26. If study data was being save in a file and studying had not found a set of
371        "starts with" bytes for the pattern, the data written to the file (though
372        never used) was taken from uninitialized memory and so caused valgrind to
373        complain.
375    27. Updated RunTest.bat as provided by Sheri Pierce.
377    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
379    29. Computation of memory usage for the table of capturing group names was
380        giving an unnecessarily large value.
383    Version 8.20 21-Oct-2011
384    ------------------------
386    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
387        a POSIX class. After further experiments with Perl, which convinced me that
388        Perl has bugs and confusions, a closing square bracket is no longer allowed
389        in a POSIX name. This bug also affected patterns with classes that started
390        with full stops.
392    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
393        captured substring, but while checking the failing first alternative,
394        substring 1 is temporarily captured. If the output vector supplied to
395        pcre_exec() was not big enough for this capture, the yield of the function
396        was still zero ("insufficient space for captured substrings"). This cannot
397        be totally fixed without adding another stack variable, which seems a lot
398        of expense for a edge case. However, I have improved the situation in cases
399        such as /(a)(b)x|abc/ matched against "abc", where the return code
400        indicates that fewer than the maximum number of slots in the ovector have
401        been set.
403    3.  Related to (2) above: when there are more back references in a pattern than
404        slots in the output vector, pcre_exec() uses temporary memory during
405        matching, and copies in the captures as far as possible afterwards. It was
406        using the entire output vector, but this conflicts with the specification
407        that only 2/3 is used for passing back captured substrings. Now it uses
408        only the first 2/3, for compatibility. This is, of course, another edge
409        case.
411    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
412        main code base, and can be used by building with --enable-jit. When this is
413        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
414        runtime --no-jit option is given.
416    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
417        ovector, the return from the function was zero, implying that there were
418        other matches that did not fit. The correct "exactly full" value is now
419        returned.
421    6.  If a subpattern that was called recursively or as a subroutine contained
422        (*PRUNE) or any other control that caused it to give a non-standard return,
423        invalid errors such as "Error -26 (nested recursion at the same subject
424        position)" or even infinite loops could occur.
426    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
427        computing the minimum length on reaching *ACCEPT, and so ended up with the
428        wrong value of 1 rather than 0. Further investigation indicates that
429        computing a minimum subject length in the presence of *ACCEPT is difficult
430        (think back references, subroutine calls), and so I have changed the code
431        so that no minimum is registered for a pattern that contains *ACCEPT.
433    8.  If (*THEN) was present in the first (true) branch of a conditional group,
434        it was not handled as intended. [But see 16 below.]
436    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
437        Sheri Pierce.
439    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
440        the first byte in a match must be "a".
442    11. Change 17 for 8.13 increased the recursion depth for patterns like
443        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
444        pattern contains any instances of (*THEN). If it does not, the old
445        optimizations are restored. It would be nice to do this on a per-group
446        basis, but at the moment that is not feasible.
448    12. In some environments, the output of pcretest -C is CRLF terminated. This
449        broke RunTest's code that checks for the link size. A single white space
450        character after the value is now allowed for.
452    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
453        For "fr", it uses the Windows-specific input and output files.
455    14. If (*THEN) appeared in a group that was called recursively or as a
456        subroutine, it did not work as intended. [But see next item.]
458    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
459        pattern fragments (but not containing any | characters). If A and B are
460        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
461        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
462        D. In other words, Perl considers parentheses that do not contain any |
463        characters to be part of a surrounding alternative, whereas PCRE was
464        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
465        differently. PCRE now behaves in the same way as Perl, except in the case
466        of subroutine/recursion calls such as (?1) which have in any case always
467        been different (but PCRE had them first :-).
469    16. Related to 15 above: Perl does not treat the | in a conditional group as
470        creating alternatives. Such a group is treated in the same way as an
471        ordinary group without any | characters when processing (*THEN). PCRE has
472        been changed to match Perl's behaviour.
474    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
475        RunGrepTest script failed.
477    18. Change 22 for version 13 caused atomic groups to use more stack. This is
478        inevitable for groups that contain captures, but it can lead to a lot of
479        stack use in large patterns. The old behaviour has been restored for atomic
480        groups that do not contain any capturing parentheses.
482    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
483        suppress the check for a minimum subject length at run time. (If it was
484        given to pcre_exec() or pcre_dfa_exec() it did work.)
486    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
487        fail to work when decoding hex characters in data strings in EBCDIC
488        environments.
490    21. It appears that in at least one Mac OS environment, the isxdigit() function
491        is implemented as a macro that evaluates to its argument more than once,
492        contravening the C 90 Standard (I haven't checked a later standard). There
493        was an instance in pcretest which caused it to go wrong when processing
494        \x{...} escapes in subject strings. The has been rewritten to avoid using
495        things like p++ in the argument of isxdigit().
498    Version 8.13 16-Aug-2011
499    ------------------------
501    1.  The Unicode data tables have been updated to Unicode 6.0.0.
503    2.  Two minor typos in pcre_internal.h have been fixed.
505    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
506        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
507        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
509    4.  There were a number of related bugs in the code for matching backrefences
510        caselessly in UTF-8 mode when codes for the characters concerned were
511        different numbers of bytes. For example, U+023A and U+2C65 are an upper
512        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
513        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
514        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
515        2-byte code at the end of the subject (it thought there wasn't enough data
516        left).
518    5.  Comprehensive information about what went wrong is now returned by
519        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
520        as the output vector has at least 2 elements. The offset of the start of
521        the failing character and a reason code are placed in the vector.
523    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
524        now returned is for the first byte of the failing character, instead of the
525        last byte inspected. This is an incompatible change, but I hope it is small
526        enough not to be a problem. It makes the returned offset consistent with
527        pcre_exec() and pcre_dfa_exec().
529    7.  pcretest now gives a text phrase as well as the error number when
530        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
531        failure, the offset and reason code are output.
533    8.  When \R was used with a maximizing quantifier it failed to skip backwards
534        over a \r\n pair if the subsequent match failed. Instead, it just skipped
535        back over a single character (\n). This seems wrong (because it treated the
536        two characters as a single entity when going forwards), conflicts with the
537        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
538        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
539        has been changed.
541    9.  Some internal refactoring has changed the processing so that the handling
542        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
543        time (the PCRE_DOTALL option was changed this way some time ago: version
544        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
545        which was always a bit of a fudge. It also means that there is one less
546        argument for the match() function, which reduces its stack requirements
547        slightly. This change also fixes an incompatibility with Perl: the pattern
548        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
550    10. More internal refactoring has drastically reduced the number of recursive
551        calls to match() for possessively repeated groups such as (abc)++ when
552        using pcre_exec().
554    11. While implementing 10, a number of bugs in the handling of groups were
555        discovered and fixed:
557        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
558        (a|)*(?1) gave a compile-time internal error.
559        ((a|)+)+  did not notice that the outer group could match an empty string.
560        (^a|^)+   was not marked as anchored.
561        (.*a|.*)+ was not marked as matching at start or after a newline.
563    12. Yet more internal refactoring has removed another argument from the match()
564        function. Special calls to this function are now indicated by setting a
565        value in a variable in the "match data" data block.
567    13. Be more explicit in pcre_study() instead of relying on "default" for
568        opcodes that mean there is no starting character; this means that when new
569        ones are added and accidentally left out of pcre_study(), testing should
570        pick them up.
572    14. The -s option of pcretest has been documented for ages as being an old
573        synonym of -m (show memory usage). I have changed it to mean "force study
574        for every regex", that is, assume /S for every regex. This is similar to -i
575        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
576        using it. It makes it easier to run collections of tests with and without
577        study enabled, and thereby test pcre_study() more easily. All the standard
578        tests are now run with and without -s (but some patterns can be marked as
579        "never study" - see 20 below).
581    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
582        restoration of the capturing data to the outer values was not happening
583        correctly.
585    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
586        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
587        pattern had matched an empty string, and so incorrectly returned a no
588        match.
590    17. There was optimizing code for the last branch of non-capturing parentheses,
591        and also for the obeyed branch of a conditional subexpression, which used
592        tail recursion to cut down on stack usage. Unfortunately, now that there is
593        the possibility of (*THEN) occurring in these branches, tail recursion is
594        no longer possible because the return has to be checked for (*THEN). These
595        two optimizations have therefore been removed. [But see 8.20/11 above.]
597    18. If a pattern containing \R was studied, it was assumed that \R always
598        matched two bytes, thus causing the minimum subject length to be
599        incorrectly computed because \R can also match just one byte.
601    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
602        was incorrectly computed.
604    20. If /S is present twice on a test pattern in pcretest input, it now
605        *disables* studying, thereby overriding the use of -s on the command line
606        (see 14 above). This is necessary for one or two tests to keep the output
607        identical in both cases.
609    21. When (*ACCEPT) was used in an assertion that matched an empty string and
610        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
612    22. When an atomic group that contained a capturing parenthesis was
613        successfully matched, but the branch in which it appeared failed, the
614        capturing was not being forgotten if a higher numbered group was later
615        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
616        group 1 to "a", when in fact it should be unset. This applied to multi-
617        branched capturing and non-capturing groups, repeated or not, and also to
618        positive assertions (capturing in negative assertions does not happen
619        in PCRE) and also to nested atomic groups.
621    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
622        subject after a captured substring, to make it easier to tell which of a
623        number of identical substrings has been captured.
625    24. The way atomic groups are processed by pcre_exec() has been changed so that
626        if they are repeated, backtracking one repetition now resets captured
627        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
628        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
629        "aaa". Previously, it would have been "a". As part of this code
630        refactoring, the way recursive calls are handled has also been changed.
632    25. If an assertion condition captured any substrings, they were not passed
633        back unless some other capturing happened later. For example, if
634        (?(?=(a))a) was matched against "a", no capturing was returned.
636    26. When studying a pattern that contained subroutine calls or assertions,
637        the code for finding the minimum length of a possible match was handling
638        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
639        group 1 called group 2 while simultaneously a separate group 2 called group
640        1). A stack overflow occurred in this case. I have fixed this by limiting
641        the recursion depth to 10.
643    27. Updated RunTest.bat in the distribution to the version supplied by Tom
644        Fortmann. This supports explicit test numbers on the command line, and has
645        argument validation and error reporting.
647    28. An instance of \X with an unlimited repeat could fail if at any point the
648        first character it looked at was a mark character.
650    29. Some minor code refactoring concerning Unicode properties and scripts
651        should reduce the stack requirement of match() slightly.
653    30. Added the '=' option to pcretest to check the setting of unused capturing
654        slots at the end of the pattern, which are documented as being -1, but are
655        not included in the return count.
657    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
658        compiled something random. Now it gives a compile-time error (as does
659        Perl).
661    32. A *MARK encountered during the processing of a positive assertion is now
662        recorded and passed back (compatible with Perl).
664    33. If --only-matching or --colour was set on a pcregrep call whose pattern
665        had alternative anchored branches, the search for a second match in a line
666        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
667        matched the line "0102" twice. The same bug affected patterns that started
668        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
669        twice.
671    34. Previously, PCRE did not allow quantification of assertions. However, Perl
672        does, and because of capturing effects, quantifying parenthesized
673        assertions may at times be useful. Quantifiers are now allowed for
674        parenthesized assertions.
676    35. A minor code tidy in pcre_compile() when checking options for \R usage.
678    36. \g was being checked for fancy things in a character class, when it should
679        just be a literal "g".
681    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
682        appearance of a nested POSIX class supersedes an apparent external class.
683        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
684        unescaped square brackets may also appear as part of class names. For
685        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
686        more like Perl. (But see 8.20/1 above.)
688    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
689        was because it thought it was \N{name}, which is not supported).
691    39. Add minix to OS list not supporting the -S option in pcretest.
693    40. PCRE tries to detect cases of infinite recursion at compile time, but it
694        cannot analyze patterns in sufficient detail to catch mutual recursions
695        such as ((?1))((?2)). There is now a runtime test that gives an error if a
696        subgroup is called recursively as a subpattern for a second time at the
697        same position in the subject string. In previous releases this might have
698        been caught by the recursion limit, or it might have run out of stack.
700    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
701        happen only once. PCRE was, however incorrectly giving a compile time error
702        "recursive call could loop indefinitely" because it cannot analyze the
703        pattern in sufficient detail. The compile time test no longer happens when
704        PCRE is compiling a conditional subpattern, but actual runaway loops are
705        now caught at runtime (see 40 above).
707    42. It seems that Perl allows any characters other than a closing parenthesis
708        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
709        has been changed to be the same.
711    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
712        as not to get warnings when autogen.sh is called. Also changed
713        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
715    44. To help people who use pcregrep to scan files containing exceedingly long
716        lines, the following changes have been made:
718        (a) The default value of the buffer size parameter has been increased from
719            8K to 20K. (The actual buffer used is three times this size.)
721        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
722            PCRE is built.
724        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
725            to be set at run time.
727        (d) Numerical values in pcregrep options can be followed by K or M, for
728            example --buffer-size=50K.
730        (e) If a line being scanned overflows pcregrep's buffer, an error is now
731            given and the return code is set to 2.
733    45. Add a pointer to the latest mark to the callout data block.
735    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
736        partial match of an empty string instead of no match. This was specific to
737        the use of ".".
739    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
740        complete match instead of a partial match. This bug was dependent on both
741        the PCRE_UTF8 and PCRE_DOTALL options being set.
743    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
744        starting byte set, because \b was not being ignored.
747    Version 8.12 15-Jan-2011
748    ------------------------
750    1.  Fixed some typos in the markup of the man pages, and wrote a script that
751        checks for such things as part of the documentation building process.
753    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
754        --match-limit and --recursion-limit options (added for 8.11). In
755        particular, this made one of the standard tests fail. (The integer value
756        went into the wrong half of a long int.)
758    3.  If the --colour option was given to pcregrep with -v (invert match), it
759        did strange things, either producing crazy output, or crashing. It should,
760        of course, ignore a request for colour when reporting lines that do not
761        match.
763    4.  Another pcregrep bug caused similar problems if --colour was specified with
764        -M (multiline) and the pattern match finished with a line ending.
766    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
767        matched in multiline mode, the following line was shown as part of the
768        match. This seems wrong, so I have changed it.
770    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
771        the check for further matches in the same line (so they could be coloured)
772        to overrun the end of the current line. If another match was found, it was
773        incorrectly shown (and then shown again when found in the next line).
775    7.  If pcregrep was compiled under Windows, there was a reference to the
776        function pcregrep_exit() before it was defined. I am assuming this was
777        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
778        reported by a user. I've moved the definition above the reference.
781    Version 8.11 10-Dec-2010
782    ------------------------
784    1.  (*THEN) was not working properly if there were untried alternatives prior
785        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
786        backtracked to try for "b" instead of moving to the next alternative branch
787        at the same level (in this case, to look for "c"). The Perl documentation
788        is clear that when (*THEN) is backtracked onto, it goes to the "next
789        alternative in the innermost enclosing group".
791    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
792        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
793        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
794        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
795        (*THEN).
797    3.  If \s appeared in a character class, it removed the VT character from
798        the class, even if it had been included by some previous item, for example
799        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
800        of \s, but is part of the POSIX "space" class.)
802    4.  A partial match never returns an empty string (because you can always
803        match an empty string at the end of the subject); however the checking for
804        an empty string was starting at the "start of match" point. This has been
805        changed to the "earliest inspected character" point, because the returned
806        data for a partial match starts at this character. This means that, for
807        example, /(?<=abc)def/ gives a partial match for the subject "abc"
808        (previously it gave "no match").
810    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
811        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
812        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
813        has an implication that the given string is incomplete (because a partial
814        match is preferred over a full match). For this reason, these items now
815        give a partial match in this situation. [Aside: previously, the one case
816        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
817        match rather than a full match, which was wrong by the old rules, but is
818        now correct.]
820    6.  There was a bug in the handling of #-introduced comments, recognized when
821        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
822        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
823        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
824        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
825        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
826        places in pcre_compile().
828    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
829        comments when looking ahead for named forward references to subpatterns,
830        the only newline sequence it recognized was NL. It now handles newlines
831        according to the set newline convention.
833    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
834        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
835        cater for a lack of strerror(). These oversights have been fixed.
837    9.  Added --match-limit and --recursion-limit to pcregrep.
839    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
841    11. When the -o option was used, pcregrep was setting a return code of 1, even
842        when matches were found, and --line-buffered was not being honoured.
844    12. Added an optional parentheses number to the -o and --only-matching options
845        of pcregrep.
847    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
848        can match an empty string. The code to do it in pcretest and pcredemo
849        needed fixing:
851        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
852            only one byte after an empty string match just before CRLF (this case
853            just got forgotten; "any" and "anycrlf" were OK).
855        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
856            mode when an empty string match preceded an ASCII character followed by
857            a non-ASCII character. (The code for advancing by one character rather
858            than one byte was nonsense.)
860        (c) The pcredemo.c sample program did not have any code at all to handle
861            the cases when CRLF is a valid newline sequence.
863    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
864        as a starting offset was within the subject string. There is now a new
865        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
866        negative or greater than the length of the string. In order to test this,
867        pcretest is extended to allow the setting of negative starting offsets.
869    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
870        starting offset points to the beginning of a UTF-8 character was
871        unnecessarily clumsy. I tidied it up.
873    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
874        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
876    17. Nobody had reported that the --include_dir option, which was added in
877        release 7.7 should have been called --include-dir (hyphen, not underscore)
878        for compatibility with GNU grep. I have changed it to --include-dir, but
879        left --include_dir as an undocumented synonym, and the same for
880        --exclude-dir, though that is not available in GNU grep, at least as of
881        release 2.5.4.
883    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
884        characters from a string of bytes) have been redefined so as not to use
885        loops, in order to improve performance in some environments. At the same
886        time, I abstracted some of the common code into auxiliary macros to save
887        repetition (this should not affect the compiled code).
889    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
890        compile-time error is now given if \c is not followed by an ASCII
891        character, that is, a byte less than 128. (In EBCDIC mode, the code is
892        different, and any byte value is allowed.)
894    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
895        START_OPTIMIZE option, which is now allowed at compile time - but just
896        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
897        to pcregrep and other applications that have no direct access to PCRE
898        options. The new /Y option in pcretest sets this option when calling
899        pcre_compile().
901    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
902        back references. Groups containing recursive back references were forced to
903        be atomic by that change, but in the case of named groups, the amount of
904        memory required was incorrectly computed, leading to "Failed: internal
905        error: code overflow". This has been fixed.
907    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
908        pcretest.c, to avoid build problems in some Borland environments.
911    Version 8.10 25-Jun-2010
912    ------------------------
914    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
915        THEN.
917    2.  (*ACCEPT) was not working when inside an atomic group.
919    3.  Inside a character class, \B is treated as a literal by default, but
920        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
921        causes the error). The code is unchanged, but I tidied the documentation.
923    4.  Inside a character class, PCRE always treated \R and \X as literals,
924        whereas Perl faults them if its -w option is set. I have changed PCRE so
925        that it faults them when PCRE_EXTRA is set.
927    5.  Added support for \N, which always matches any character other than
928        newline. (It is the same as "." when PCRE_DOTALL is not set.)
930    6.  When compiling pcregrep with newer versions of gcc which may have
931        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
932        declared with attribute warn_unused_result" were given. Just casting the
933        result to (void) does not stop the warnings; a more elaborate fudge is
934        needed. I've used a macro to implement this.
936    7.  Minor change to pcretest.c to avoid a compiler warning.
938    8.  Added four artifical Unicode properties to help with an option to make
939        \s etc use properties (see next item). The new properties are: Xan
940        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
942    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
943        use Unicode properties. (*UCP) at the start of a pattern can be used to set
944        this option. Modified pcretest to add /W to test this facility. Added
945        REG_UCP to make it available via the POSIX interface.
947    10. Added --line-buffered to pcregrep.
949    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
950        studied, and the match started with a letter with a code point greater than
951        127 whose first byte was different to the first byte of the other case of
952        the letter, the other case of this starting letter was not recognized
953        (#976).
955    12. If a pattern that was studied started with a repeated Unicode property
956        test, for example, \p{Nd}+, there was the theoretical possibility of
957        setting up an incorrect bitmap of starting bytes, but fortunately it could
958        not have actually happened in practice until change 8 above was made (it
959        added property types that matched character-matching opcodes).
961    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
962        possible starting bytes for non-anchored patterns.
964    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
965        \R, and also a number of cases that involve Unicode properties, both
966        explicit and implicit when PCRE_UCP is set.
968    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
969        input, it could crash or give wrong results if characters with values
970        greater than 0xc0 were present in the subject string. (Detail: it assumed
971        UTF-8 input when processing these items.)
973    16. Added a lot of (int) casts to avoid compiler warnings in systems where
974        size_t is 64-bit (#991).
976    17. Added a check for running out of memory when PCRE is compiled with
977        --disable-stack-for-recursion (#990).
979    18. If the last data line in a file for pcretest does not have a newline on
980        the end, a newline was missing in the output.
982    19. The default pcre_chartables.c file recognizes only ASCII characters (values
983        less than 128) in its various bitmaps. However, there is a facility for
984        generating tables according to the current locale when PCRE is compiled. It
985        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
986        characters, are recognized by isspace() and therefore were getting set in
987        these tables, and indeed these tables seem to approximate to ISO 8859. This
988        caused a problem in UTF-8 mode when pcre_study() was used to create a list
989        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
990        which of course cannot start UTF-8 characters. I have changed the code so
991        that only real ASCII characters (less than 128) and the correct starting
992        bytes for UTF-8 encodings are set for characters greater than 127 when in
993        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
994        altogether.)
996    20. Added the /T option to pcretest so as to be able to run tests with non-
997        standard character tables, thus making it possible to include the tests
998        used for 19 above in the standard set of tests.
1000    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1001        reference to a subpattern the other side of a comment that contains an
1002        opening parenthesis caused either an internal compiling error, or a
1003        reference to the wrong subpattern.
1006    Version 8.02 19-Mar-2010
1007    ------------------------
1009    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1011    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1012        configured.
1014    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1015        original author of that file, following a query about its status.
1017    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1018        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1020    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1021        quantifier applied to a forward-referencing subroutine call, could compile
1022        incorrect code or give the error "internal error: previously-checked
1023        referenced subpattern not found".
1025    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1026        variables to point to external functions. For these systems, therefore,
1027        pcre_malloc etc. are now initialized to local functions that call the
1028        relevant global functions.
1030    7.  There were two entries missing in the vectors called coptable and poptable
1031        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1032        I've fixed the data, and added a kludgy way of testing at compile time that
1033        the lengths are correct (equal to the number of opcodes).
1035    8.  Following on from 7, I added a similar kludge to check the length of the
1036        eint vector in pcreposix.c.
1038    9.  Error texts for pcre_compile() are held as one long string to avoid too
1039        much relocation at load time. To find a text, the string is searched,
1040        counting zeros. There was no check for running off the end of the string,
1041        which could happen if a new error number was added without updating the
1042        string.
1044    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1046    11. \K was not working if it appeared in an atomic group or in a group that
1047        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1048        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1049        the assertion is positive, but not if it is negative.
1051    12. Change 11 fortuitously reduced the size of the stack frame used in the
1052        "match()" function of pcre_exec.c by one pointer. Forthcoming
1053        implementation of support for (*MARK) will need an extra pointer on the
1054        stack; I have reserved it now, so that the stack frame size does not
1055        decrease.
1057    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1058        item in branch that calls a recursion is a subroutine call - as in the
1059        second branch in the above example - was incorrectly given the compile-
1060        time error "recursive call could loop indefinitely" because pcre_compile()
1061        was not correctly checking the subroutine for matching a non-empty string.
1063    14. The checks for overrunning compiling workspace could trigger after an
1064        overrun had occurred. This is a "should never occur" error, but it can be
1065        triggered by pathological patterns such as hundreds of nested parentheses.
1066        The checks now trigger 100 bytes before the end of the workspace.
1068    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1071    Version 8.01 19-Jan-2010
1072    ------------------------
1074    1.  If a pattern contained a conditional subpattern with only one branch (in
1075        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1076        computed the wrong minimum data length (which is of course zero for such
1077        subpatterns). This could cause incorrect "no match" results.
1079    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1080        the pattern is reset in the first branch, pcre_compile() failed with
1081        "internal error: code overflow at offset...". This happened only when
1082        the reset was to the original external option setting. (An optimization
1083        abstracts leading options settings into an external setting, which was the
1084        cause of this.)
1086    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1087        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1088        assertion pattern did not match (meaning that the assertion was true), it
1089        was incorrectly treated as false if the SKIP had been reached during the
1090        matching. This also applied to assertions used as conditions.
1092    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1093        assertion subpattern, including such a pattern used as a condition,
1094        unpredictable results occurred, instead of the error return
1097    5.  The C++ GlobalReplace function was not working like Perl for the special
1098        situation when an empty string is matched. It now does the fancy magic
1099        stuff that is necessary.
1101    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1102        removed. (These were left over from very, very early versions of PCRE.)
1104    7.  Some cosmetic changes to the code to make life easier when compiling it
1105        as part of something else:
1107        (a) Change DEBUG to PCRE_DEBUG.
1109        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1110            called "current" as "current_branch", to prevent a collision with the
1111            Linux macro when compiled as a kernel module.
1113        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1114            prevent a collision with the Linux macro when compiled as a kernel
1115            module.
1117    8.  In pcre_compile() there are some checks for integer overflows that used to
1118        cast potentially large values to (double). This has been changed to that
1119        when building, a check for int64_t is made, and if it is found, it is used
1120        instead, thus avoiding the use of floating point arithmetic. (There is no
1121        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1122        double.
1124    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1125        2005 (difference between two addresses compared to an unsigned value).
1127    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1128        custom one, because of the following reported problem in Windows:
1130          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1131              under Win32.
1132          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1133              therefore missing the function definition.
1134          - The compiler thus generates a "C" signature for the test function.
1135          - The linker fails to find the "C" function.
1136          - PCRE fails to configure if asked to do so against libbz2.
1138    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1139        messages were output:
1141          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1142          rerunning libtoolize, to keep the correct libtool macros in-tree.
1143          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1145        I have done both of these things.
1147    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1148        most of the time, it *can* run out if it is given a pattern that contains a
1149        runaway infinite recursion. I updated the discussion in the pcrestack man
1150        page.
1152    13. Now that we have gone to the x.xx style of version numbers, the minor
1153        version may start with zero. Using 08 or 09 is a bad idea because users
1154        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1155        interpreted as invalid octal numbers. I've updated the previous comment in
1156        configure.ac, and also added a check that gives an error if 08 or 09 are
1157        used.
1159    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1160        causing partial matching to fail when the end of the subject matched \W
1161        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1163    15. There were some discrepancies between the declarations in pcre_internal.h
1164        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1165        their definitions. The declarations used "const uschar *" and the
1166        definitions used USPTR. Even though USPTR is normally defined as "const
1167        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1168        reported that: "This difference in casting confuses some C++ compilers, for
1169        example, SunCC recognizes above declarations as different functions and
1170        generates broken code for hbpcre." I have changed the declarations to use
1171        USPTR.
1173    16. GNU libtool is named differently on some systems. The autogen.sh script now
1174        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1175        (FreeBSD).
1177    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1178        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1179        comment: "Figure out how to create a longlong from a string: strtoll and
1180        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1181        instance, but it only takes 2 args instead of 3!"
1183    18. A subtle bug concerned with back references has been fixed by a change of
1184        specification, with a corresponding code fix. A pattern such as
1185        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1186        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1187        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1188        same bug. Such groups have to be quantified to be useful, or contained
1189        inside another quantified group. (If there's no repetition, the reference
1190        can never match.) The problem arises because, having left the group and
1191        moved on to the rest of the pattern, a later failure that backtracks into
1192        the group uses the captured value from the final iteration of the group
1193        rather than the correct earlier one. I have fixed this in PCRE by forcing
1194        any group that contains a reference to itself to be an atomic group; that
1195        is, there cannot be any backtracking into it once it has completed. This is
1196        similar to recursive and subroutine calls.
1199    Version 8.00 19-Oct-09
1200    ----------------------
1202    1.  The table for translating pcre_compile() error codes into POSIX error codes
1203        was out-of-date, and there was no check on the pcre_compile() error code
1204        being within the table. This could lead to an OK return being given in
1205        error.
1207    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1208        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1209        in a Windows environment.
1211    3.  The pcregrep --count option prints the count for each file even when it is
1212        zero, as does GNU grep. However, pcregrep was also printing all files when
1213        --files-with-matches was added. Now, when both options are given, it prints
1214        counts only for those files that have at least one match. (GNU grep just
1215        prints the file name in this circumstance, but including the count seems
1216        more useful - otherwise, why use --count?) Also ensured that the
1217        combination -clh just lists non-zero counts, with no names.
1219    4.  The long form of the pcregrep -F option was incorrectly implemented as
1220        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1221        but it seems right to fix it, and I didn't think it was worth preserving
1222        the old behaviour.
1224    5.  The command line items --regex=pattern and --regexp=pattern were not
1225        recognized by pcregrep, which required --regex pattern or --regexp pattern
1226        (with a space rather than an '='). The man page documented the '=' forms,
1227        which are compatible with GNU grep; these now work.
1229    6.  No libpcreposix.pc file was created for pkg-config; there was just
1230        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1232    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1233        when UCP support is not needed, by modifying the Python script that
1234        generates it from Unicode data files. This should not matter if the module
1235        is correctly used as a library, but I received one complaint about 50K of
1236        unwanted data. My guess is that the person linked everything into his
1237        program rather than using a library. Anyway, it does no harm.
1239    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1240        was a minimum greater than 1 for a wide character in a possessive
1241        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1242        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1243        character. Chaos in the form of incorrect output or a compiling loop could
1244        result.
1246    9.  The restrictions on what a pattern can contain when partial matching is
1247        requested for pcre_exec() have been removed. All patterns can now be
1248        partially matched by this function. In addition, if there are at least two
1249        slots in the offset vector, the offset of the earliest inspected character
1250        for the match and the offset of the end of the subject are set in them when
1251        PCRE_ERROR_PARTIAL is returned.
1253    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1254        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1255        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1256        and may be more useful for multi-segment matching.
1258    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1259        used to be given if ever the end of the subject was reached; now it is
1260        given only if matching could not proceed because another character was
1261        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1262        string "Z", which now yields "no match" instead of "partial match". In the
1263        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1264        final character ended with (*FAIL).
1266    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1267        if the pattern had a "must contain" character that was already found in the
1268        earlier partial match, unless partial matching was again requested. For
1269        example, with the pattern /dog.(body)?/, the "must contain" character is
1270        "g". If the first part-match was for the string "dog", restarting with
1271        "sbody" failed. This bug has been fixed.
1273    13. The string returned by pcre_dfa_exec() after a partial match has been
1274        changed so that it starts at the first inspected character rather than the
1275        first character of the match. This makes a difference only if the pattern
1276        starts with a lookbehind assertion or \b or \B (\K is not supported by
1277        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1278        matching functions compatible, and I think it's the right thing to do.
1280    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1281        so that the demonstration program is easily available in environments where
1282        PCRE has not been installed from source.
1284    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1285        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1286        library.
1288    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1289        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1290        is not the first non-POSIX option to be added. Clearly some people find
1291        these options useful.
1293    17. If a caller to the POSIX matching function regexec() passes a non-zero
1294        value for nmatch with a NULL value for pmatch, the value of
1295        nmatch is forced to zero.
1297    18. RunGrepTest did not have a test for the availability of the -u option of
1298        the diff command, as RunTest does. It now checks in the same way as
1299        RunTest, and also checks for the -b option.
1301    19. If an odd number of negated classes containing just a single character
1302        interposed, within parentheses, between a forward reference to a named
1303        subpattern and the definition of the subpattern, compilation crashed with
1304        an internal error, complaining that it could not find the referenced
1305        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1306        [The bug was that it was starting one character too far in when skipping
1307        over the character class, thus treating the ] as data rather than
1308        terminating the class. This meant it could skip too much.]
1310    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1311        /g option in pcretest when the pattern contains \K, which makes it possible
1312        to have an empty string match not at the start, even when the pattern is
1313        anchored. Updated pcretest and pcredemo to use this option.
1315    21. If the maximum number of capturing subpatterns in a recursion was greater
1316        than the maximum at the outer level, the higher number was returned, but
1317        with unset values at the outer level. The correct (outer level) value is
1318        now given.
1320    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1321        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1322        make it do so. The string so far is captured, making this feature
1323        compatible with Perl.
1325    23. The tests have been re-organized, adding tests 11 and 12, to make it
1326        possible to check the Perl 5.10 features against Perl 5.10.
1328    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1329        pattern matches a fixed length string. PCRE did not allow this; now it
1330        does. Neither allows recursion.
1332    25. I finally figured out how to implement a request to provide the minimum
1333        length of subject string that was needed in order to match a given pattern.
1334        (It was back references and recursion that I had previously got hung up
1335        on.) This code has now been added to pcre_study(); it finds a lower bound
1336        to the length of subject needed. It is not necessarily the greatest lower
1337        bound, but using it to avoid searching strings that are too short does give
1338        some useful speed-ups. The value is available to calling programs via
1339        pcre_fullinfo().
1341    26. While implementing 25, I discovered to my embarrassment that pcretest had
1342        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1343        study optimizations had never been tested with that matching function.
1344        Oops. What is worse, even when it was passed study data, there was a bug in
1345        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1346        were also very few tests of studied patterns with pcre_dfa_exec().
1348    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1349        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1350        on the other side of the coin, they are no longer allowed to have different
1351        names, because these cannot be distinguished in PCRE, and this has caused
1352        confusion. (This is a difference from Perl.)
1354    28. When duplicate subpattern names are present (necessarily with different
1355        numbers, as required by 27 above), and a test is made by name in a
1356        conditional pattern, either for a subpattern having been matched, or for
1357        recursion in such a pattern, all the associated numbered subpatterns are
1358        tested, and the overall condition is true if the condition is true for any
1359        one of them. This is the way Perl works, and is also more like the way
1360        testing by number works.
1363    Version 7.9 11-Apr-09
1364    ---------------------
1366    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1367        (pcretest), all targets were linked against these libraries. This included
1368        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1369        libraries. This caused unwanted dependencies to be created. This problem
1370        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1371        pcretest is linked with readline.
1373    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1374        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1375        moved outside it again, because FALSE and TRUE are already defined in AIX,
1376        but BOOL is not.
1378    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1379        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1381    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1382        hyphens) following file names and line numbers when outputting matching
1383        lines. This is not true; no spaces are inserted. I have also clarified the
1384        wording for the --colour (or --color) option.
1386    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1387        was not coloured; this is different to GNU grep, so I have changed it to be
1388        the same.
1390    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1391        each matching line was coloured. Now it goes on to look for further matches
1392        of any of the test patterns, which is the same behaviour as GNU grep.
1394    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1395        doesn't make sense to accept an empty string match in pcregrep, so I have
1396        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1397        seems to be how GNU grep behaves.
1399    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1400        start or after a newline", because the conditional assertion was not being
1401        correctly handled. The rule now is that both the assertion and what follows
1402        in the first alternative must satisfy the test.
1404    9.  If auto-callout was enabled in a pattern with a conditional group whose
1405        condition was an assertion, PCRE could crash during matching, both with
1406        pcre_exec() and pcre_dfa_exec().
1408    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1409        used for matching.
1411    11. Unicode property support in character classes was not working for
1412        characters (bytes) greater than 127 when not in UTF-8 mode.
1414    12. Added the -M command line option to pcretest.
1416    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1418    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1420    16. Added comments and documentation about mis-use of no_arg in the C++
1421        wrapper.
1423    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1424        from Martin Jerabek that uses macro names for all relevant character and
1425        string constants.
1427    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1428        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1429        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1430        these, but not everybody uses configure.
1432    19. A conditional group that had only one branch was not being correctly
1433        recognized as an item that could match an empty string. This meant that an
1434        enclosing group might also not be so recognized, causing infinite looping
1435        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1436        with the subject "ab", where knowledge that the repeated group can match
1437        nothing is needed in order to break the loop.
1439    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1440        exec(), but without supplying a callout function, matching went wrong.
1442    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1443        leak if the size of the offset vector was greater than 30. When the vector
1444        is smaller, the saved offsets during recursion go onto a local stack
1445        vector, but for larger vectors malloc() is used. It was failing to free
1446        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1447        error, in fact).
1449    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1450        heapframe that is used only when UTF-8 support is enabled. This caused no
1451        problem, but was untidy.
1453    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1454        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1455        included within another project.
1457    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1458        slightly modified by me:
1460          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1461              not building pcregrep.
1463          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1464              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1466    25. Forward references, both numeric and by name, in patterns that made use of
1467        duplicate group numbers, could behave incorrectly or give incorrect errors,
1468        because when scanning forward to find the reference group, PCRE was not
1469        taking into account the duplicate group numbers. A pattern such as
1470        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1472    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1473        the feature of a custom pointer more persuasive (as requested by a user).
1475    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1476        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1477        pre-defined.
1479    28. Added support for (*UTF8) at the start of a pattern.
1481    29. Arrange for flags added by the "release type" setting in CMake to be shown
1482        in the configuration summary.
1485    Version 7.8 05-Sep-08
1486    ---------------------
1488    1.  Replaced UCP searching code with optimized version as implemented for Ad
1489        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1490        stage table and inline lookup instead of a function, giving speed ups of 2
1491        to 5 times on some simple patterns that I tested. Permission was given to
1492        distribute the MultiStage2.py script that generates the tables (it's not in
1493        the tarball, but is in the Subversion repository).
1495    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1496        scripts.
1498    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1499        a group with a zero qualifier. The result of the study could be incorrect,
1500        or the function might crash, depending on the pattern.
1502    4.  Caseless matching was not working for non-ASCII characters in back
1503        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1504        It now works when Unicode Property Support is available.
1506    5.  In pcretest, an escape such as \x{de} in the data was always generating
1507        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1508        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1509        truncation.
1511    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1513    7.  Added two (int) casts to pcregrep when printing the difference of two
1514        pointers, in case they are 64-bit values.
1516    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1517        test 2 if it fails.
1519    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1520        and a #define of that name to empty if it is not externally set. This is to
1521        allow users of MSVC to set it if necessary.
1523    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1524        the convenience functions in the pcre_get.c source file.
1526    11. An option change at the start of a pattern that had top-level alternatives
1527        could cause overwriting and/or a crash. This command provoked a crash in
1528        some environments:
1530          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1532        This potential security problem was recorded as CVE-2008-2371.
1534    12. For a pattern where the match had to start at the beginning or immediately
1535        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1536        pcre_dfa_exec() could read past the end of the passed subject if there was
1537        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1538        pcretest so that it places the subject at the end of its malloc-ed buffer.
1540    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1541        exec() might read past the end of the data buffer in UTF-8 mode.
1543    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1544        the data contained the byte 0x85 as part of a UTF-8 character within its
1545        first line. This applied both to normal and DFA matching.
1547    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1548        /^[^d]*?$/8 failed to match "abc".
1550    16. Added a missing copyright notice to pcrecpp_internal.h.
1552    17. Make it more clear in the documentation that values returned from
1553        pcre_exec() in ovector are byte offsets, not character counts.
1555    18. Tidied a few places to stop certain compilers from issuing warnings.
1557    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1558        supplied by Stefan Weber. I made a further small update for 7.8 because
1559        there is a change of source arrangements: the pcre_searchfuncs.c module is
1560        replaced by pcre_ucd.c.
1563    Version 7.7 07-May-08
1564    ---------------------
1566    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1567        a string to a long long, pretend we don't even have a long long." This is
1568        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1570    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1571        pre-7.6 versions, which defined a global no_arg variable instead of putting
1572        it in the RE class. (See also #8 below.)
1574    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1575        Lopes.
1577    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1579        (1) The include/exclude patterns were being applied to the whole pathnames
1580            of files, instead of just to the final components.
1582        (2) If there was more than one level of directory, the subdirectories were
1583            skipped unless they satisfied the include/exclude conditions. This is
1584            inconsistent with GNU grep (and could even be seen as contrary to the
1585            pcregrep specification - which I improved to make it absolutely clear).
1586            The action now is always to scan all levels of directory, and just
1587            apply the include/exclude patterns to regular files.
1589    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1590        --exclude_dir in the tests to avoid scanning .svn directories.
1592    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1593        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1594        doesn't support NULs in patterns.
1596    7.  Added some missing "const"s to declarations of static tables in
1597        pcre_compile.c and pcre_dfa_exec.c.
1599    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1600        caused by fix #2  above. (Subsequently also a second patch to fix the
1601        first patch. And a third patch - this was a messy problem.)
1603    9.  Applied Craig's patch to remove the use of push_back().
1605    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1606        matching function regexec().
1608    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1609        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1610        references. PCRE supports relative numbers with this syntax (I don't think
1611        Oniguruma does).
1613    12. Previously, a group with a zero repeat such as (...){0} was completely
1614        omitted from the compiled regex. However, this means that if the group
1615        was called as a subroutine from elsewhere in the pattern, things went wrong
1616        (an internal error was given). Such groups are now left in the compiled
1617        pattern, with a new opcode that causes them to be skipped at execution
1618        time.
1620    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1621        to the way PCRE behaves:
1623        (a) A lone ] character is dis-allowed (Perl treats it as data).
1625        (b) A back reference to an unmatched subpattern matches an empty string
1626            (Perl fails the current match path).
1628        (c) A data ] in a character class must be notated as \] because if the
1629            first data character in a class is ], it defines an empty class. (In
1630            Perl it is not possible to have an empty class.) The empty class []
1631            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1632            The negative empty class [^] matches any one character, independently
1633            of the DOTALL setting.
1635    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1636        non-existent subpattern following a character class starting with ']' and
1637        containing () gave an internal compiling error instead of "reference to
1638        non-existent subpattern". Fortunately, when the pattern did exist, the
1639        compiled code was correct. (When scanning forwards to check for the
1640        existencd of the subpattern, it was treating the data ']' as terminating
1641        the class, so got the count wrong. When actually compiling, the reference
1642        was subsequently set up correctly.)
1644    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1645        it was being rejected as not supported by pcre_dfa_exec(), even though
1646        other assertions are supported. I have made pcre_dfa_exec() support
1647        (*FAIL).
1649    16. The implementation of 13c above involved the invention of a new opcode,
1650        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1651        cannot be changed at match time, I realized I could make a small
1652        improvement to matching performance by compiling OP_ALLANY instead of
1653        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1654        on the OP_ANY path.
1656    17. Compiling pcretest on Windows with readline support failed without the
1657        following two fixes: (1) Make the unistd.h include conditional on
1658        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1660    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1661        ncurses library to be included for pcretest when ReadLine support is
1662        requested, but also to allow for it to be overridden. This patch came from
1663        Daniel Bergström.
1665    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1666        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1667        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1668        spotting this.
1671    Version 7.6 28-Jan-08
1672    ---------------------
1674    1.  A character class containing a very large number of characters with
1675        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1676        overflow.
1678    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1679        HAVE_LONG_LONG is not defined.
1681    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1682        bring them up to date and include new features. This patch includes:
1684        - Fixed PH's badly added libz and libbz2 support.
1685        - Fixed a problem with static linking.
1686        - Added pcredemo. [But later removed - see 7 below.]
1687        - Fixed dftables problem and added an option.
1688        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1689            HAVE_LONG_LONG.
1690        - Added readline support for pcretest.
1691        - Added an listing of the option settings after cmake has run.
1693    4.  A user submitted a patch to Makefile that makes it easy to create
1694        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1695        Makefile.am that cause it to include this special target, without
1696        affecting anything else. Note that the same mingw target plus all
1697        the other distribution libraries and programs are now supported
1698        when configuring with CMake (see 6 below) instead of with
1699        Configure/Make.
1701    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1702        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1703        exported in the Windows port". It has not yet been confirmed that the patch
1704        solves the problem, but it does no harm.
1706    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1707        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1708        with CMake, and also correct the comment about stack recursion.
1710    7.  Remove the automatic building of pcredemo from the ./configure system and
1711        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1712        of a program that users should build themselves after PCRE is installed, so
1713        building it automatically is not really right. What is more, it gave
1714        trouble in some build environments.
1716    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1719    Version 7.5 10-Jan-08
1720    ---------------------
1722    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1723        values in parens when parsing an RE using the C++ wrapper."
1725    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1726        Characters greater than 255 were excluded from the class instead of being
1727        included.
1729    3.  The same bug as (2) above applied to negated POSIX classes such as
1730        [:^space:].
1732    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1733        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1734        I have changed it.
1736    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1737        first named subpattern) and a construct such as (?&a) would reference the
1738        first named subpattern whose name started with "a" (in other words, the
1739        length check was missing). Both these problems are fixed. "Subpattern name
1740        expected" is now given for (?&) (a zero-length name), and this patch also
1741        makes it give the same error for \k'' (previously it complained that that
1742        was a reference to a non-existent subpattern).
1744    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1745        this is right because (?- can be followed by option settings as well as by
1746        digits. I have, however, made the messages clearer.
1748    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1749        than the number used in the conditional) now cause a compile-time error.
1750        This is actually not compatible with Perl, which accepts such patterns, but
1751        treats the conditional as always being FALSE (as PCRE used to), but it
1752        seems to me that giving a diagnostic is better.
1754    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1755        and messages.
1757    9.  Fix two occurrences of "backslash" in comments that should have been
1758        "backspace".
1760    10. Remove two redundant lines of code that can never be obeyed (their function
1761        was moved elsewhere).
1763    11. The program that makes PCRE's Unicode character property table had a bug
1764        which caused it to generate incorrect table entries for sequences of
1765        characters that have the same character type, but are in different scripts.
1766        It amalgamated them into a single range, with the script of the first of
1767        them. In other words, some characters were in the wrong script. There were
1768        thirteen such cases, affecting characters in the following ranges:
1770          U+002b0 - U+002c1
1771          U+0060c - U+0060d
1772          U+0061e - U+00612
1773          U+0064b - U+0065e
1774          U+0074d - U+0076d
1775          U+01800 - U+01805
1776          U+01d00 - U+01d77
1777          U+01d9b - U+01dbf
1778          U+0200b - U+0200f
1779          U+030fc - U+030fe
1780          U+03260 - U+0327f
1781          U+0fb46 - U+0fbb1
1782          U+10450 - U+1049d
1784    12. The -o option (show only the matching part of a line) for pcregrep was not
1785        compatible with GNU grep in that, if there was more than one match in a
1786        line, it showed only the first of them. It now behaves in the same way as
1787        GNU grep.
1789    13. If the -o and -v options were combined for pcregrep, it printed a blank
1790        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1791        does the same. The return code can be used to tell if there were any
1792        non-matching lines.
1794    14. Added --file-offsets and --line-offsets to pcregrep.
1796    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1797        infinitely looping recursion. The bug was that positive lookaheads were not
1798        being skipped when checking for a possible empty match (negative lookaheads
1799        and both kinds of lookbehind were skipped).
1801    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1802        inclusion of <windows.h> to before rather than after the definition of
1803        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1805    17. Specifying a possessive quantifier with a specific limit for a Unicode
1806        character property caused pcre_compile() to compile bad code, which led at
1807        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1808        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1809        caused the error; without that there was no problem.
1811    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1813    19. Added --enable-pcretest-libreadline.
1815    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1816        RE::GlobalReplace(). As a result, the number of replacements returned was
1817        double what it should be. I removed one of the increments, but Craig sent a
1818        later patch that removed the other one (the right fix) and added unit tests
1819        that check the return values (which was not done before).
1821    21. Several CMake things:
1823        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1824            the names libpcre and libpcreposix, not just pcre and pcreposix.
1826        (2) The above change means that pcretest and pcregrep are now correctly
1827            linked with the newly-built libraries, not previously installed ones.
1831    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1832        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1833        UTF-8 newline character). The key issue is that the pattern starts .*;
1834        this means that the match must be either at the beginning, or after a
1835        newline. The bug was in the code for advancing after a failed match and
1836        checking that the new position followed a newline. It was not taking
1837        account of UTF-8 characters correctly.
1839    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1840        character classes. PCRE was not treating the sequence [:...:] as a
1841        character class unless the ... were all letters. Perl, however, seems to
1842        allow any characters between [: and :], though of course it rejects as
1843        unknown any "names" that contain non-letters, because all the known class
1844        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1845        for example, whereas PCRE did not - it did not recognize a POSIX character
1846        class. This seemed a bit dangerous, so the code has been changed to be
1847        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1848        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1849        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1850        Perl does, and where it didn't before.
1852    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1853        Windows environments %n is disabled by default.
1856    Version 7.4 21-Sep-07
1857    ---------------------
1859    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1860        means that a class such as [\s] counted as "explicit reference to CR or
1861        LF". That isn't really right - the whole point of the change was to try to
1862        help when there was an actual mention of one of the two characters. So now
1863        the change happens only if \r or \n (or a literal CR or LF) character is
1864        encountered.
1866    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1867        of both had grown to the point where there were only 3 bits left.
1868        Fortunately, there was spare space in the data structure, and so I have
1869        moved the internal flags into a new 16-bit field to free up more option
1870        bits.
1872    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1873        but did not set the internal JCHANGED flag - either of these is enough to
1874        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1875        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1876        start sets both bits.
1878    4.  Added options (at build time, compile time, exec time) to change \R from
1879        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1881    5.  doc/pcresyntax.html was missing from the distribution.
1883    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1884        compatibility, even though it is no longer used.
1886    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1887        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1888        windows.h file is present (where different names are used). [This was
1889        reversed later after testing - see 16 below.]
1891    8.  Changed all #include <config.h> to #include "config.h". There were also
1892        some further <pcre.h> cases that I changed to "pcre.h".
1894    9.  When pcregrep was used with the --colour option, it missed the line ending
1895        sequence off the lines that it output.
1897    10. It was pointed out to me that arrays of string pointers cause lots of
1898        relocations when a shared library is dynamically loaded. A technique of
1899        using a single long string with a table of offsets can drastically reduce
1900        these. I have refactored PCRE in four places to do this. The result is
1901        dramatic:
1903          Originally:                          290
1904          After changing UCP table:            187
1905          After changing error message table:   43
1906          After changing table of "verbs"       36
1907          After changing table of Posix names   22
1909        Thanks to the folks working on Gregex for glib for this insight.
1911    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1912        unicode-properties was also set.
1914    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1916    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1917        checked only for CRLF.
1919    14. Added casts to pcretest.c to avoid compiler warnings.
1921    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1923    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1924        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1925        entirely. This removes changes made in 7 above.
1927    17. The CMake files have been updated, and there is now more information about
1928        building with CMake in the NON-UNIX-USE document.
1931    Version 7.3 28-Aug-07
1932    ---------------------
1934     1. In the rejigging of the build system that eventually resulted in 7.1, the
1935        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1936        brackets there is not right, since it causes compilers to look for an
1937        installed pcre.h, not the version that is in the source that is being
1938        compiled (which of course may be different). I have changed it back to:
1940          #include "pcre.h"
1942        I have a vague recollection that the change was concerned with compiling in
1943        different directories, but in the new build system, that is taken care of
1944        by the VPATH setting the Makefile.
1946     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1947        when the subject happened to end in the byte 0x85 (e.g. if the last
1948        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1949        characters but of course it shouldn't be taken as a newline when it is part
1950        of another character. The bug was that, for an unlimited repeat of . in
1951        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1952        characters when looking for a newline.
1954     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1956     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1957        in debug output.
1959     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1960        long printing in the pcrecpp unittest when running under MinGW.
1962     6. ESC_K was left out of the EBCDIC table.
1964     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1965        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1966        limit also applies to "virtual nesting" when a pattern is recursive, and in
1967        this case 1000 isn't so big. I have been able to remove this limit at the
1968        expense of backing off one optimization in certain circumstances. Normally,
1969        when pcre_exec() would call its internal match() function recursively and
1970        immediately return the result unconditionally, it uses a "tail recursion"
1971        feature to save stack. However, when a subpattern that can match an empty
1972        string has an unlimited repetition quantifier, it no longer makes this
1973        optimization. That gives it a stack frame in which to save the data for
1974        checking that an empty string has been matched. Previously this was taken
1975        from the 1000-entry workspace that had been reserved. So now there is no
1976        explicit limit, but more stack is used.
1978     8. Applied Daniel's patches to solve problems with the import/export magic
1979        syntax that is required for Windows, and which was going wrong for the
1980        pcreposix and pcrecpp parts of the library. These were overlooked when this
1981        problem was solved for the main library.
1983     9. There were some crude static tests to avoid integer overflow when computing
1984        the size of patterns that contain repeated groups with explicit upper
1985        limits. As the maximum quantifier is 65535, the maximum group length was
1986        set at 30,000 so that the product of these two numbers did not overflow a
1987        32-bit integer. However, it turns out that people want to use groups that
1988        are longer than 30,000 bytes (though not repeat them that many times).
1989        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1990        made it possible to implement the integer overflow checks in a much more
1991        dynamic way, which I have now done. The artificial limitation on group
1992        length has been removed - we now have only the limit on the total length of
1993        the compiled pattern, which depends on the LINK_SIZE setting.
1995    10. Fixed a bug in the documentation for get/copy named substring when
1996        duplicate names are permitted. If none of the named substrings are set, the
1997        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1998        empty string.
2000    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2001        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2002        because the ] is interpreted as the first data character and the
2003        terminating ] is not found. PCRE has been made compatible with Perl in this
2004        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2005        cause memory overwriting.
2007    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2008        string has been matched (to stop an infinite loop). It was not recognizing
2009        a conditional subpattern that could match an empty string if that
2010        subpattern was within another subpattern. For example, it looped when
2011        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2012        condition was not nested. This bug has been fixed.
2014    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2015        past the start of the subject in the presence of bytes with the top bit
2016        set, for example "\x8aBCD".
2018    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2019        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2021    14. Optimized (?!) to (*FAIL).
2023    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2024        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2025        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2026        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2027        does: it's just the validity check that is more restrictive.
2029    16. Inserted checks for integer overflows during escape sequence (backslash)
2030        processing, and also fixed erroneous offset values for syntax errors during
2031        backslash processing.
2033    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2034        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2036    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2037        caused an overrun.
2039    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2040        something other than just ASCII characters) inside a group that had an
2041        unlimited repeat caused a loop at compile time (while checking to see
2042        whether the group could match an empty string).
2044    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2045        [\P{Any}] did so. (Error in the code for printing property names.)
2047    21. An orphan \E inside a character class could cause a crash.
2049    22. A repeated capturing bracket such as (A)? could cause a wild memory
2050        reference during compilation.
2052    23. There are several functions in pcre_compile() that scan along a compiled
2053        expression for various reasons (e.g. to see if it's fixed length for look
2054        behind). There were bugs in these functions when a repeated \p or \P was
2055        present in the pattern. These operators have additional parameters compared
2056        with \d, etc, and these were not being taken into account when moving along
2057        the compiled data. Specifically:
2059        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2060            length.
2062        (b) An item such as \pL+ within a repeated group could cause crashes or
2063            loops.
2065        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2066            "reference to non-existent subpattern" error.
2068        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2070    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2071        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2073    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2075    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2076        character were causing crashes (broken optimization).
2078    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2079        \p or \P) caused a compile-time loop.
2081    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2082        break. For example, the unstudied pattern [\r\n]A does not match the string
2083        "\r\nA" because change 7.0/46 below moves the current point on by two
2084        characters after failing to match at the start. However, the pattern \nA
2085        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2086        the same is true. There doesn't seem any very clean way out of this, but
2087        what I have chosen to do makes the common cases work: PCRE now takes note
2088        of whether there can be an explicit match for \r or \n anywhere in the
2089        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2090        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2091        pattern has explicit CR or LF references.
2093    29. Added (*CR) etc for changing newline setting at start of pattern.
2096    Version 7.2 19-Jun-07
2097    ---------------------
2099     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2100        which is apparently normally available under Windows.
2102     2. Re-jig the pcregrep tests with different newline settings in an attempt
2103        to make them independent of the local environment's newline setting.
2105     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2107     4. Some of the "internals" tests were previously cut out when the link size
2108        was not 2, because the output contained actual offsets. The recent new
2109        "Z" feature of pcretest means that these can be cut out, making the tests
2110        usable with all link sizes.
2112     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2113        stack recursion. This gives a massive performance boost under BSD, but just
2114        a small improvement under Linux. However, it saves one field in the frame
2115        in all cases.
2117     6. Added more features from the forthcoming Perl 5.10:
2119        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2120            recursion call. It refers to the nth most recently opened parentheses.
2122        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2123            to be opened parentheses.
2125        (c) Conditions that refer to capturing parentheses can be specified
2126            relatively, for example, (?(-2)... or (?(+3)...
2128        (d) \K resets the start of the current match so that everything before
2129            is not part of it.
2131        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2133        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2134            reference syntax.
2136        (g) (?| introduces a group in which the numbering of parentheses in each
2137            alternative starts with the same number.
2139        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2141     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2144     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2145        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2146        for detecting groups that can match an empty string.
2148     9. A pattern with a very large number of alternatives (more than several
2149        hundred) was running out of internal workspace during the pre-compile
2150        phase, where pcre_compile() figures out how much memory will be needed. A
2151        bit of new cunning has reduced the workspace needed for groups with
2152        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2153        workspace instead of running out of the 4096 that are available.
2155    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2157    11. Applied patch from Google to remove an optimization that didn't quite work.
2158        The report of the bug said:
2160          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2161          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2162          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2164    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2165        it matched the wrong number of bytes.
2168    Version 7.1 24-Apr-07
2169    ---------------------
2171     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2172        that is more "standard", making use of automake and other Autotools. There
2173        is some re-arrangement of the files and adjustment of comments consequent
2174        on this.
2176     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2177        for recursive directory scanning broke on some systems because the files
2178        are not scanned in any specific order and on different systems the order
2179        was different. A call to "sort" has been inserted into RunGrepTest for the
2180        approprate test as a short-term fix. In the longer term there may be an
2181        alternative.
2183     3. I had an email from Eric Raymond about problems translating some of PCRE's
2184        man pages to HTML (despite the fact that I distribute HTML pages, some
2185        people do their own conversions for various reasons). The problems
2186        concerned the use of low-level troff macros .br and .in. I have therefore
2187        removed all such uses from the man pages (some were redundant, some could
2188        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2189        HTML has been updated to handle .nf/.fi and to complain if it encounters
2190        .br or .in.
2192     4. Updated comments in configure.ac that get placed in config.h.in and also
2193        arranged for config.h to be included in the distribution, with the name
2194        config.h.generic, for the benefit of those who have to compile without
2195        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2197     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2198        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2199        makevp.bat for the current PCRE, using the additional files
2200        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2202     6. A Windows user reported a minor discrepancy with test 2, which turned out
2203        to be caused by a trailing space on an input line that had got lost in his
2204        copy. The trailing space was an accident, so I've just removed it.
2206     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2207        that is needed.
2209     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2210        as "const" (a) because they are and (b) because it helps the PHP
2211        maintainers who have recently made a script to detect big data structures
2212        in the php code that should be moved to the .rodata section. I remembered
2213        to update Builducptable as well, so it won't revert if ucptable.h is ever
2214        re-created.
2216     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2217        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2218        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2219        support is not required. This saves 1.5-2K of code, which is important in
2220        some applications.
2222        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2223        so as not to refer to the tables, even though these functions will never be
2224        called when UTF-8 support is disabled. Otherwise there are problems with a
2225        shared library.
2227    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2229        (a) It was defining its arguments as char * instead of void *.
2231        (b) It was assuming that all moves were upwards in memory; this was true
2232            a long time ago when I wrote it, but is no longer the case.
2234        The emulated memove() is provided for those environments that have neither
2235        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2236        is clearly not the case, as these two bugs were recently reported.
2238    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2239        and Detrail to create the HTML documentation, the .txt form of the man
2240        pages, and it removes trailing spaces from listed files. It also creates
2241        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2242        case, it wraps all the #defines with #ifndefs. This script should be run
2243        before "make dist".
2245    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2246        with Unicode property support.
2248        (a) For a maximizing quantifier, if the two different cases of the
2249            character were of different lengths in their UTF-8 codings (there are
2250            some cases like this - I found 11), and the matching function had to
2251            back up over a mixture of the two cases, it incorrectly assumed they
2252            were both the same length.
2254        (b) When PCRE was configured to use the heap rather than the stack for
2255            recursion during matching, it was not correctly preserving the data for
2256            the other case of a UTF-8 character when checking ahead for a match
2257            while processing a minimizing repeat. If the check also involved
2258            matching a wide character, but failed, corruption could cause an
2259            erroneous result when trying to check for a repeat of the original
2260            character.
2262    13. Some tidying changes to the testing mechanism:
2264        (a) The RunTest script now detects the internal link size and whether there
2265            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2266            values substituted by "configure". (The RunGrepTest script already did
2267            this for UTF-8.) The configure.ac script no longer substitutes the
2268            relevant variables.
2270        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2271            with length and offset values. This means that the output is different
2272            for different internal link sizes. Test 2 is skipped for link sizes
2273            other than 2 because of this, bypassing the problem. Unfortunately,
2274            there was also a test in test 3 (the locale tests) that used /B and
2275            failed for link sizes other than 2. Rather than cut the whole test out,
2276            I have added a new /Z option to pcretest that replaces the length and
2277            offset values with spaces. This is now used to make test 3 independent
2278            of link size. (Test 2 will be tidied up later.)
2280    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2281        segmentation fault instead of returning the appropriate error message.
2283    15. In multiline mode when the newline sequence was set to "any", the pattern
2284        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2285        This doesn't seem right; it now treats the CRLF combination as the line
2286        ending, and so does not match in that case. It's only a pattern such as ^$
2287        that would hit this one: something like ^ABC$ would have failed after \r
2288        and then tried again after \r\n.
2290    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2291        in an attempt to make files that differ only in their line terminators
2292        compare equal. This works on Linux.
2294    17. Under certain error circumstances pcregrep might try to free random memory
2295        as it exited. This is now fixed, thanks to valgrind.
2297    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2298        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2299        was because its rules for how to advance for /g after matching an empty
2300        string at the end of a line did not allow for this case. They now check for
2301        it specially.
2303    20. pcretest is supposed to handle patterns and data of any length, by
2304        extending its buffers when necessary. It was getting this wrong when the
2305        buffer for a data line had to be extended.
2307    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2308        CRLF as a newline sequence.
2310    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2311        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2312        I have nevertheless tidied it up.
2314    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2316    24. Added a man page for pcre-config.
2319    Version 7.0 19-Dec-06
2320    ---------------------
2322     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2323        moving to gcc 4.1.1.
2325     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2326        sys/time.h, which is documented as needed for this function. It doesn't
2327        seem to matter on Linux, but it showed up on some releases of OS X.
2329     3. It seems that there are systems where bytes whose values are greater than
2330        127 match isprint() in the "C" locale. The "C" locale should be the
2331        default when a C program starts up. In most systems, only ASCII printing
2332        characters match isprint(). This difference caused the output from pcretest
2333        to vary, making some of the tests fail. I have changed pcretest so that:
2335        (a) When it is outputting text in the compiled version of a pattern, bytes
2336            other than 32-126 are always shown as hex escapes.
2338        (b) When it is outputting text that is a matched part of a subject string,
2339            it does the same, unless a different locale has been set for the match
2340            (using the /L modifier). In this case, it uses isprint() to decide.
2342     4. Fixed a major bug that caused incorrect computation of the amount of memory
2343        required for a compiled pattern when options that changed within the
2344        pattern affected the logic of the preliminary scan that determines the
2345        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2346        that the computed length was too small. The symptoms of this bug were
2347        either the PCRE error "internal error: code overflow" from pcre_compile(),
2348        or a glibc crash with a message such as "pcretest: free(): invalid next
2349        size (fast)". Examples of patterns that provoked this bug (shown in
2350        pcretest format) are:
2352          /(?-x: )/x
2353          /(?x)(?-x: \s*#\s*)/
2354          /((?i)[\x{c0}])/8
2355          /(?i:[\x{c0}])/8
2357        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2358        is now done differently.
2360     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2361        wrapper classes; (b) implement a new function in the C++ scanner that is
2362        more efficient than the old way of doing things because it avoids levels of
2363        recursion in the regex matching; (c) add a paragraph to the documentation
2364        for the FullMatch() function.
2366     6. The escape sequence \n was being treated as whatever was defined as
2367        "newline". Not only was this contrary to the documentation, which states
2368        that \n is character 10 (hex 0A), but it also went horribly wrong when
2369        "newline" was defined as CRLF. This has been fixed.
2371     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2372        was being set to -1 for the "end of line" case (supposedly a value that no
2373        character can have). Though this value is never used (the check for end of
2374        line is "zero bytes in current character"), it caused compiler complaints.
2375        I've changed it to 0xffffffff.
2377     8. In pcre_version.c, the version string was being built by a sequence of
2378        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2379        string (as it is for production releases) called a macro with an empty
2380        argument. The C standard says the result of this is undefined. The gcc
2381        compiler treats it as an empty string (which was what was wanted) but it is
2382        reported that Visual C gives an error. The source has been hacked around to
2383        avoid this problem.
2385     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2386        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2387        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2388        of them did).
2390    10. Originally, pcretest opened its input and output without "b"; then I was
2391        told that "b" was needed in some environments, so it was added for release
2392        5.0 to both the input and output. (It makes no difference on Unix-like
2393        systems.) Later I was told that it is wrong for the input on Windows. I've
2394        now abstracted the modes into two macros, to make it easier to fiddle with
2395        them, and removed "b" from the input mode under Windows.
2397    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2399    12. Added -help and --help to pcretest as an official way of being reminded
2400        of the options.
2402    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2403        and pcrecpp.cc because they annoy compilers at high warning levels.
2405    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2407    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2408        scripts are not C programs :-) and which was not noticed because it works
2409        on Linux.
2411    16. pcretest is supposed to handle any length of pattern and data line (as one
2412        line or as a continued sequence of lines) by extending its input buffer if
2413        necessary. This feature was broken for very long pattern lines, leading to
2414        a string of junk being passed to pcre_compile() if the pattern was longer
2415        than about 50K.
2417    17. I have done a major re-factoring of the way pcre_compile() computes the
2418        amount of memory needed for a compiled pattern. Previously, there was code
2419        that made a preliminary scan of the pattern in order to do this. That was
2420        OK when PCRE was new, but as the facilities have expanded, it has become
2421        harder and harder to keep it in step with the real compile phase, and there
2422        have been a number of bugs (see for example, 4 above). I have now found a
2423        cunning way of running the real compile function in a "fake" mode that
2424        enables it to compute how much memory it would need, while actually only
2425        ever using a few hundred bytes of working memory and without too many
2426        tests of the mode. This should make future maintenance and development
2427        easier. A side effect of this work is that the limit of 200 on the nesting
2428        depth of parentheses has been removed (though this was never a serious
2429        limitation, I suspect). However, there is a downside: pcre_compile() now
2430        runs more slowly than before (30% or more, depending on the pattern). I
2431        hope this isn't a big issue. There is no effect on runtime performance.
2433    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2434        newline (only possible for the last line of a file) and it was a
2435        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2437    19. Added additional timing features to pcretest. (1) The -tm option now times
2438        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2439        separate command line item, by a number that specifies the number of
2440        repeats to use when timing. The default is 50000; this gives better
2441        precision, but takes uncomfortably long for very large patterns.
2443    20. Extended pcre_study() to be more clever in cases where a branch of a
2444        subpattern has no definite first character. For example, (a*|b*)[cd] would
2445        previously give no result from pcre_study(). Now it recognizes that the
2446        first character must be a, b, c, or d.
2448    21. There was an incorrect error "recursive call could loop indefinitely" if
2449        a subpattern (or the entire pattern) that was being tested for matching an
2450        empty string contained only one non-empty item after a nested subpattern.
2451        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2452        incorrectly, because the \d was being skipped in the check.
2454    22. The pcretest program now has a new pattern option /B and a command line
2455        option -b, which is equivalent to adding /B to every pattern. This causes
2456        it to show the compiled bytecode, without the additional information that
2457        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2458        is the same as /B/I).
2460    23. A new optimization is now able automatically to treat some sequences such
2461        as a*b as a*+b. More specifically, if something simple (such as a character
2462        or a simple class like \d) has an unlimited quantifier, and is followed by
2463        something that cannot possibly match the quantified thing, the quantifier
2464        is automatically "possessified".
2466    24. A recursive reference to a subpattern whose number was greater than 39
2467        went wrong under certain circumstances in UTF-8 mode. This bug could also
2468        have affected the operation of pcre_study().
2470    25. Realized that a little bit of performance could be had by replacing
2471        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2473    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2475    27. Possessive quantifiers such as a++ were previously implemented by turning
2476        them into atomic groups such as ($>a+). Now they have their own opcodes,
2477        which improves performance. This includes the automatically created ones
2478        from 23 above.
2480    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2481        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2482        the first matched character to be a colon. This applied both to named and
2483        numbered groups.
2485    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2487    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2488        building PCRE on Windows easier, so I have included it in the distribution.
2490    31. There is now a check in pcretest against a ridiculously large number being
2491        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2492        loop, the loop is abandoned.
2494    32. Forward references to subpatterns in conditions such as (?(2)...) where
2495        subpattern 2 is defined later cause pcre_compile() to search forwards in
2496        the pattern for the relevant set of parentheses. This search went wrong
2497        when there were unescaped parentheses in a character class, parentheses
2498        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2500    33. "Subroutine" calls and backreferences were previously restricted to
2501        referencing subpatterns earlier in the regex. This restriction has now
2502        been removed.
2504    34. Added a number of extra features that are going to be in Perl 5.10. On the
2505        whole, these are just syntactic alternatives for features that PCRE had
2506        previously implemented using the Python syntax or my own invention. The
2507        other formats are all retained for compatibility.
2509        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2510            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2511            also .NET compatible.
2513        (b) A recursion or subroutine call to a named group can now be defined as
2514            (?&name) as well as (?P>name).
2516        (c) A backreference to a named group can now be defined as \k<name> or
2517            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2518            5.10, are also .NET compatible.
2520        (d) A conditional reference to a named group can now use the syntax
2521            (?(<name>) or (?('name') as well as (?(name).
2523        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2524            groups (named and numbered) that are never evaluated inline, but can be
2525            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2526            is always false. There may be only one alternative in such a group.
2528        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2529            as the simple (?(R). The condition is true only if the most recent
2530            recursion is that of the given number or name. It does not search out
2531            through the entire recursion stack.
2533        (g) The escape \gN or \g{N} has been added, where N is a positive or
2534            negative number, specifying an absolute or relative reference.
2536    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2537        some "unreachable code" warnings.
2539    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2540        things, this adds five new scripts.
2542    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2543        There were also incompatibilities regarding the handling of \Q..\E inside
2544        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2545        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2547    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2548        matches an empty string, and forcibly breaks the loop. There were bugs in
2549        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2550        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2551        separate and independent bugs (that affected different cases) have been
2552        fixed.
2554    39. Refactored the code to abolish the use of different opcodes for small
2555        capturing bracket numbers. This is a tidy that I avoided doing when I
2556        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2557        The new approach is not only tidier, it makes it possible to reduce the
2558        memory needed to fix the previous bug (38).
2560    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2561        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2562        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2563        mode.
2565    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2566        report.
2568    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2569        copy construction and assignment in the C++ wrapper.
2571    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2572        couple of bugs that could have given wrong results in the "--newline=crlf"
2573        case.
2575    44. Added a number of casts and did some reorganization of signed/unsigned int
2576        variables following suggestions from Dair Grant. Also renamed the variable
2577        "this" as "item" because it is a C++ keyword.
2579    45. Arranged for dftables to add
2581          #include "pcre_internal.h"
2583        to pcre_chartables.c because without it, gcc 4.x may remove the array
2584        definition from the final binary if PCRE is built into a static library and
2585        dead code stripping is activated.
2587    46. For an unanchored pattern, if a match attempt fails at the start of a
2588        newline sequence, and the newline setting is CRLF or ANY, and the next two
2589        characters are CRLF, advance by two characters instead of one.
2592    Version 6.7 04-Jul-06
2593    ---------------------
2595     1. In order to handle tests when input lines are enormously long, pcretest has
2596        been re-factored so that it automatically extends its buffers when
2597        necessary. The code is crude, but this _is_ just a test program. The
2598        default size has been increased from 32K to 50K.
2600     2. The code in pcre_study() was using the value of the re argument before
2601        testing it for NULL. (Of course, in any sensible call of the function, it
2602        won't be NULL.)
2604     3. The memmove() emulation function in pcre_internal.h, which is used on
2605        systems that lack both memmove() and bcopy() - that is, hardly ever -
2606        was missing a "static" storage class specifier.
2608     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2609        containing an extended class (one that cannot be represented by a bitmap
2610        because it contains high-valued characters or Unicode property items, e.g.
2611        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2612        pattern, but PCRE should not loop if you do not (it no longer does).
2613        [Detail: two cases were found: (a) a repeated subpattern containing an
2614        extended class; (b) a recursive reference to a subpattern that followed a
2615        previous extended class. It wasn't skipping over the extended class
2616        correctly when UTF-8 mode was not set.]
2618     5. A negated single-character class was not being recognized as fixed-length
2619        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2620        compile error "lookbehind assertion is not fixed length".
2622     6. The RunPerlTest auxiliary script was showing an unexpected difference
2623        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2624        write a Perl script that can interpret lines of an input file either as
2625        byte characters or as UTF-8, which is what "perltest" was being required to
2626        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2627        can't do is switch easily at run time between having the "use utf8;" pragma
2628        or not. In the end, I fudged it by using the RunPerlTest script to insert
2629        "use utf8;" explicitly for the UTF-8 tests.
2631     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2632        the end of the subject string, contrary to the documentation and to what
2633        Perl does. This was true of both matching functions. Now it matches only at
2634        the start of the subject and immediately after *internal* newlines.
2636     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2637        a pointer to an int instead of a pointer to an unsigned long int. This
2638        caused problems on 64-bit systems.
2640     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2641        instance of the 'standard' template library not being so standard".
2643    10. There was no check on the number of named subpatterns nor the maximum
2644        length of a subpattern name. The product of these values is used to compute
2645        the size of the memory block for a compiled pattern. By supplying a very
2646        long subpattern name and a large number of named subpatterns, the size
2647        computation could be caused to overflow. This is now prevented by limiting
2648        the length of names to 32 characters, and the number of named subpatterns
2649        to 10,000.
2651    11. Subpatterns that are repeated with specific counts have to be replicated in
2652        the compiled pattern. The size of memory for this was computed from the
2653        length of the subpattern and the repeat count. The latter is limited to
2654        65535, but there was no limit on the former, meaning that integer overflow
2655        could in principle occur. The compiled length of a repeated subpattern is
2656        now limited to 30,000 bytes in order to prevent this.
2658    12. Added the optional facility to have named substrings with the same name.
2660    13. Added the ability to use a named substring as a condition, using the
2661        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2662        are numbers (not recommended). Forward references are permitted.
2664    14. Added forward references in named backreferences (if you see what I mean).
2666    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2667        pattern could run off the end of the subject. For example, the pattern
2668        "(?s)(.{1,5})"8 did this with the subject "ab".
2670    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2671        PCRE_CASELESS was set when matching characters that were quantified with ?
2672        or *.
2674    17. A character class other than a single negated character that had a minimum
2675        but no maximum quantifier - for example [ab]{6,} - was not handled
2676        correctly by pce_dfa_exec(). It would match only one character.
2678    18. A valid (though odd) pattern that looked like a POSIX character
2679        class but used an invalid character after [ (for example [[,abc,]]) caused
2680        pcre_compile() to give the error "Failed: internal error: code overflow" or
2681        in some cases to crash with a glibc free() error. This could even happen if
2682        the pattern terminated after [[ but there just happened to be a sequence of
2683        letters, a binary zero, and a closing ] in the memory that followed.
2685    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2686        over the years. Originally (before any Unicode support), just the bottom 8
2687        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2688        output from "man perlunicode" includes this:
2690          The regular expression compiler produces polymorphic opcodes.  That
2691          is, the pattern adapts to the data and automatically switches to
2692          the Unicode character scheme when presented with Unicode data--or
2693          instead uses a traditional byte scheme when presented with byte
2694          data.
2696        Sadly, a wide octal escape does not cause a switch, and in a string with
2697        no other multibyte characters, these octal escapes are treated as before.
2698        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2699        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2700        Unicode string.
2702        I have not perpetrated such confusion in PCRE. Up till now, it took just
2703        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2704        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2705        translate to the appropriate multibyte character.
2707    29. Applied some refactoring to reduce the number of warnings from Microsoft
2708        and Borland compilers. This has included removing the fudge introduced
2709        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2710        a warning about an unused variable.
2712    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2713        characters since release 4.0, because Perl (from release 5.004) does not.
2714        [Or at least, is documented not to: some releases seem to be in conflict
2715        with the documentation.] However, when a pattern was studied with
2716        pcre_study() and all its branches started with \s, PCRE still included VT
2717        as a possible starting character. Of course, this did no harm; it just
2718        caused an unnecessary match attempt.
2720    22. Removed a now-redundant internal flag bit that recorded the fact that case
2721        dependency changed within the pattern. This was once needed for "required
2722        byte" processing, but is no longer used. This recovers a now-scarce options
2723        bit. Also moved the least significant internal flag bit to the most-
2724        significant bit of the word, which was not previously used (hangover from
2725        the days when it was an int rather than a uint) to free up another bit for
2726        the future.
2728    23. Added support for CRLF line endings as well as CR and LF. As well as the
2729        default being selectable at build time, it can now be changed at runtime
2730        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2731        specify that it is scanning data with non-default line endings.
2733    24. Changed the definition of CXXLINK to make it agree with the definition of
2734        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2736    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2737        recursions. This makes a big different to stack usage for some patterns.
2739    26. If a subpattern containing a named recursion or subroutine reference such
2740        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2741        the space required for the compiled pattern went wrong and gave too small a
2742        value. Depending on the environment, this could lead to "Failed: internal
2743        error: code overflow at offset 49" or "glibc detected double free or
2744        corruption" errors.
2746    27. Applied patches from Google (a) to support the new newline modes and (b) to
2747        advance over multibyte UTF-8 characters in GlobalReplace.
2749    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2750        difference for some implementation of PCRE in some Windows version.
2752    29. Added some extra testing facilities to pcretest:
2754        \q<number>   in a data line sets the "match limit" value
2755        \Q<number>   in a data line sets the "match recursion limt" value
2756        -S <number>  sets the stack size, where <number> is in megabytes
2758        The -S option isn't available for Windows.
2761    Version 6.6 06-Feb-06
2762    ---------------------
2764     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2765        in pcreposix.h. I have copied the definition from pcre.h.
2767     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2768        because pcre.h is no longer a built file.
2770     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2771        not normally included in the compiled code.
2774    Version 6.5 01-Feb-06
2775    ---------------------
2777     1. When using the partial match feature with pcre_dfa_exec(), it was not
2778        anchoring the second and subsequent partial matches at the new starting
2779        point. This could lead to incorrect results. For example, with the pattern
2780        /1234/, partially matching against "123" and then "a4" gave a match.
2782     2. Changes to pcregrep:
2784        (a) All non-match returns from pcre_exec() were being treated as failures
2785            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2786            error message is output. Some extra information is given for the
2787            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2788            probably the only errors that are likely to be caused by users (by
2789            specifying a regex that has nested indefinite repeats, for instance).
2790            If there are more than 20 of these errors, pcregrep is abandoned.
2792        (b) A binary zero was treated as data while matching, but terminated the
2793            output line if it was written out. This has been fixed: binary zeroes
2794            are now no different to any other data bytes.
2796        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2797            used to set a locale for matching. The --locale=xxxx long option has
2798            been added (no short equivalent) to specify a locale explicitly on the
2799            pcregrep command, overriding the environment variables.
2801        (d) When -B was used with -n, some line numbers in the output were one less
2802            than they should have been.
2804        (e) Added the -o (--only-matching) option.
2806        (f) If -A or -C was used with -c (count only), some lines of context were
2807            accidentally printed for the final match.
2809        (g) Added the -H (--with-filename) option.
2811        (h) The combination of options -rh failed to suppress file names for files
2812            that were found from directory arguments.
2814        (i) Added the -D (--devices) and -d (--directories) options.
2816        (j) Added the -F (--fixed-strings) option.
2818        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2820        (l) Added the --colo(u)r option.
2822        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2823            is not present by default.
2825     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2826        items such as (?R) or (?1), when the recursion could match a number of
2827        alternatives. If it matched one of the alternatives, but subsequently,
2828        outside the recursion, there was a failure, the code tried to back up into
2829        the recursion. However, because of the way PCRE is implemented, this is not
2830        possible, and the result was an incorrect result from the match.
2832        In order to prevent this happening, the specification of recursion has
2833        been changed so that all such subpatterns are automatically treated as
2834        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2836     4. I had overlooked the fact that, in some locales, there are characters for
2837        which isalpha() is true but neither isupper() nor islower() are true. In
2838        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2839        and ordfeminine) are like this. This affected the treatment of \w and \W
2840        when they appeared in character classes, but not when they appeared outside
2841        a character class. The bit map for "word" characters is now created
2842        separately from the results of isalnum() instead of just taking it from the
2843        upper, lower, and digit maps. (Plus the underscore character, of course.)
2845     5. The above bug also affected the handling of POSIX character classes such as
2846        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2847        permanent tables. Instead, the bit maps for such a class were previously
2848        created as the appropriate unions of the upper, lower, and digit bitmaps.
2849        Now they are created by subtraction from the [[:word:]] class, which has
2850        its own bitmap.
2852     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2853        It is created by subtracting the vertical space characters (\x09, \x0a,
2854        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2855        subtraction was done in the overall bitmap for a character class, meaning
2856        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2857        be recognized. This bug has been fixed.
2859     7. Patches from the folks at Google:
2861          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2862          real life, but is still worth protecting against".
2864          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2865          regular expressions".
2867          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2868          have it.
2870          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2871          "configure" and the latter not, in order to fix a problem somebody had
2872          with compiling the Arg class on HP-UX.
2874          (e) Improve the error-handling of the C++ wrapper a little bit.
2876          (f) New tests for checking recursion limiting.
2878     8. The pcre_memmove() function, which is used only if the environment does not
2879        have a standard memmove() function (and is therefore rarely compiled),
2880        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2881        returning a result (though PCRE never actually uses the result).
2883     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2884        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2885        returned instead of calling malloc() with an overflowing number that would
2886        most likely cause subsequent chaos.
2888    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2890    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2891        with this option is matched, the nmatch and pmatch options of regexec() are
2892        ignored.
2894    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2895        provided in case anyone wants to the the POSIX interface with UTF-8
2896        strings.
2898    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2899        C++ linking (needed for some HP-UX environments).
2901    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2902        (unused parameter) and in the pcre_printint() function (omitted "default"
2903        switch label when the default is to do nothing).
2905    15. Added some code to make it possible, when PCRE is compiled as a C++
2906        library, to replace subject pointers for pcre_exec() with a smart pointer
2907        class, thus making it possible to process discontinuous strings.
2909    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2910        much the same function. They were added by different people who were trying
2911        to make PCRE easy to compile on non-Unix systems. It has been suggested
2912        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2913        for compiling on Windows systems. I have therefore replaced it with
2914        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2915        defaults to "extern" for C or "extern C" for C++, which works fine on
2916        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2917        SCOPE with something explicit in config.h. In addition:
2919        (a) pcreposix.h still had just "extern" instead of either of these macros;
2920            I have replaced it with PCRE_DATA_SCOPE.
2922        (b) Functions such as _pcre_xclass(), which are internal to the library,
2923            but external in the C sense, all had PCRE_EXPORT in their definitions.
2924            This is apparently wrong for the Windows case, so I have removed it.
2925            (It makes no difference on Unix-like systems.)
2927    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2928        of recursive calls to match(). This is different to MATCH_LIMIT because
2929        that limits the total number of calls to match(), not all of which increase
2930        the depth of recursion. Limiting the recursion depth limits the amount of
2931        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2932        when PCRE is compiled, and changed at run time. A patch from Google adds
2933        this functionality to the C++ interface.
2935    18. Changes to the handling of Unicode character properties:
2937        (a) Updated the table to Unicode 4.1.0.
2939        (b) Recognize characters that are not in the table as "Cn" (undefined).
2941        (c) I revised the way the table is implemented to a much improved format
2942            which includes recognition of ranges. It now supports the ranges that
2943            are defined in UnicodeData.txt, and it also amalgamates other
2944            characters into ranges. This has reduced the number of entries in the
2945            table from around 16,000 to around 3,000, thus reducing its size
2946            considerably. I realized I did not need to use a tree structure after
2947            all - a binary chop search is just as efficient. Having reduced the
2948            number of entries, I extended their size from 6 bytes to 8 bytes to
2949            allow for more data.
2951        (d) Added support for Unicode script names via properties such as \p{Han}.
2953    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2954        matching that character.
2956    20. When matching a repeated Unicode property with a minimum greater than zero,
2957        (for example \pL{2,}), PCRE could look past the end of the subject if it
2958        reached it while seeking the minimum number of characters. This could
2959        happen only if some of the characters were more than one byte long, because
2960        there is a check for at least the minimum number of bytes.
2962    21. Refactored the implementation of \p and \P so as to be more general, to
2963        allow for more different types of property in future. This has changed the
2964        compiled form incompatibly. Anybody with saved compiled patterns that use
2965        \p or \P will have to recompile them.
2967    22. Added "Any" and "L&" to the supported property types.
2969    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2970        but give a compile time error if the value is greater than 0xff.
2972    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2973        accidentally not being installed or uninstalled.
2975    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2976        made were to insert the current release number. This seemed silly, because
2977        it made things harder for people building PCRE on systems that don't run
2978        "configure". I have turned pcre.h into a distributed file, no longer built
2979        by "configure", with the version identification directly included. There is
2980        no longer a pcre.h.in file.
2982        However, this change necessitated a change to the pcre-config script as
2983        well. It is built from pcre-config.in, and one of the substitutions was the
2984        release number. I have updated configure.ac so that ./configure now finds
2985        the release number by grepping pcre.h.
2987    26. Added the ability to run the tests under valgrind.
2990    Version 6.4 05-Sep-05
2991    ---------------------
2993     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2994        "--" to be printed when multiple files were scanned, even when none of the
2995        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2996        consider it to be a bug, and have restored the previous behaviour.
2998     2. A couple of code tidies to get rid of compiler warnings.
3000     3. The pcretest program used to cheat by referring to symbols in the library
3001        whose names begin with _pcre_. These are internal symbols that are not
3002        really supposed to be visible externally, and in some environments it is
3003        possible to suppress them. The cheating is now confined to including
3004        certain files from the library's source, which is a bit cleaner.
3006     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3007        file's purpose clearer.
3009     5. Reorganized pcre_ucp_findchar().
3012    Version 6.3 15-Aug-05
3013    ---------------------
3015     1. The file libpcre.pc.in did not have general read permission in the tarball.
3017     2. There were some problems when building without C++ support:
3019        (a) If C++ support was not built, "make install" and "make test" still
3020            tried to test it.
3022        (b) There were problems when the value of CXX was explicitly set. Some
3023            changes have been made to try to fix these, and ...
3025        (c) --disable-cpp can now be used to explicitly disable C++ support.
3027        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3028            backslash in a target when C++ was disabled. This confuses some
3029            versions of "make", apparently. Using an intermediate variable solves
3030            this. (Same for CPP_LOBJ.)
3032     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3033        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3034        necessary on certain architectures.
3036     4. Added a setting of -export-symbols-regex to the link command to remove
3037        those symbols that are exported in the C sense, but actually are local
3038        within the library, and not documented. Their names all begin with
3039        "_pcre_". This is not a perfect job, because (a) we have to except some
3040        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3041        available (and never for static libraries). I have made a note to try to
3042        find a way round (a) in the future.
3045    Version 6.2 01-Aug-05
3046    ---------------------
3048     1. There was no test for integer overflow of quantifier values. A construction
3049        such as {1111111111111111} would give undefined results. What is worse, if
3050        a minimum quantifier for a parenthesized subpattern overflowed and became
3051        negative, the calculation of the memory size went wrong. This could have
3052        led to memory overwriting.
3054     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3056     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3057        operating environments where this matters.
3059     4. Applied Giuseppe Maxia's patch to add additional features for controlling
3060        PCRE options from within the C++ wrapper.
3062     5. Named capturing subpatterns were not being correctly counted when a pattern
3063        was compiled. This caused two problems: (a) If there were more than 100
3064        such subpatterns, the calculation of the memory needed for the whole
3065        compiled pattern went wrong, leading to an overflow error. (b) Numerical
3066        back references of the form \12, where the number was greater than 9, were
3067        not recognized as back references, even though there were sufficient
3068        previous subpatterns.
3070     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3071        versions of gcc, e.g. 2.95.4.
3074    Version 6.1 21-Jun-05
3075    ---------------------
3077     1. There was one reference to the variable "posix" in pcretest.c that was not
3078        surrounded by "#if !defined NOPOSIX".
3080     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3081        the cross-check on the old pcre_info() function, for the benefit of the
3082        cut-down version of PCRE that is currently imported into Exim.
3084     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3085        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3086        patterns that start (?i) but I don't think that matters. The use of (?i) is
3087        just an example; this all applies to the other options as well.
3089     4. Since libtool seems to echo the compile commands it is issuing, the output
3090        from "make" can be reduced a bit by putting "@" in front of each libtool
3091        compile command.
3093     5. Patch from the folks at Google for configure.in to be a bit more thorough
3094        in checking for a suitable C++ installation before trying to compile the
3095        C++ stuff. This should fix a reported problem when a compiler was present,
3096        but no suitable headers.
3098     6. The man pages all had just "PCRE" as their title. I have changed them to
3099        be the relevant file name. I have also arranged that these names are
3100        retained in the file doc/pcre.txt, which is a concatenation in text format
3101        of all the man pages except the little individual ones for each function.
3103     7. The NON-UNIX-USE file had not been updated for the different set of source
3104        files that come with release 6. I also added a few comments about the C++
3105        wrapper.
3108  Version 6.0 07-Jun-05  Version 6.0 07-Jun-05
3109  ---------------------  ---------------------

Removed from v.77  
changed lines
  Added in v.1009

  ViewVC Help
Powered by ViewVC 1.1.5