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

Diff of /code/trunk/ChangeLog

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

revision 277 by ph10, Mon Nov 26 17:03:45 2007 UTC revision 1038 by zherczeg, Wed Sep 12 09:03:23 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.5 12-Nov-07  Version 8.32
5    ------------
6    
7    1.  Improved JIT compiler optimizations for first character search and single
8        character iterators.
9    
10    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
11        Patch by Daniel Richard G.
12    
13    3.  Single character iterator optimizations in the JIT compiler.
14    
15    4.  Improved JIT compiler optimizations for character ranges.
16    
17    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
18        Reported by Giuseppe D'Angelo.
19    
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.
24    
25    7.  Improved instruction cache flush for POWER/PowerPC.
26        Patch by Daniel Richard G.
27    
28    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
29        grep:
30    
31        (a) There is now no limit to the number of patterns to be matched.
32    
33        (b) An error is given if a pattern is too long.
34    
35        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
36            are now supported.
37    
38        (d) --exclude-from and --include-from (multiple use) have been added.
39    
40        (e) Exclusions and inclusions now apply to all files and directories, not
41            just to those obtained from scanning a directory recursively.
42    
43        (f) Multiple uses of -f and --file-list are now supported.
44    
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.
48    
49        (h) The documentation has been revised and clarified in places.
50    
51    9.  Improve the matching speed of capturing brackets.
52    
53    10. Changed the meaning of \X so that it now matches a Unicode extended
54        grapheme cluster.
55    
56    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
57        out POSIX threads when JIT support is configured.
58    
59    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
60    
61    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
62        structure could go wrong in environments where size_t is not the same size
63        as int.
64    
65    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
66        set.
67    
68    15. The EBCDIC support had decayed; later updates to the code had included
69        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
70        general tidy up of EBCDIC-related issues, and the documentation was also
71        not quite right. There is now a test that can be run on ASCII systems to
72        check some of the EBCDIC-related things (but is it not a full test).
73    
74    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
75        in a small tidy to the code.
76    
77    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
78    
79    
80    Version 8.31 06-July-2012
81    -------------------------
82    
83    1.  Fixing a wrong JIT test case and some compiler warnings.
84    
85    2.  Removed a bashism from the RunTest script.
86    
87    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
88        to unsigned type, result still unsigned" that was given by an MS compiler
89        on encountering the code "-sizeof(xxx)".
90    
91    4.  Partial matching support is added to the JIT compiler.
92    
93    5.  Fixed several bugs concerned with partial matching of items that consist
94        of more than one character:
95    
96        (a) /^(..)\1/ did not partially match "aba" because checking references was
97            done on an "all or nothing" basis. This also applied to repeated
98            references.
99    
100        (b) \R did not give a hard partial match if \r was found at the end of the
101            subject.
102    
103        (c) \X did not give a hard partial match after matching one or more
104            characters at the end of the subject.
105    
106        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
107            a partial match for the string "\r".
108    
109        (e) When newline was set to CRLF, the metacharacter "." did not recognize
110            a partial match for a CR character at the end of the subject string.
111    
112    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
113        running pcretest, the text "(JIT)" added to the output whenever JIT is
114        actually used to run the match.
115    
116    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
117        or /S+[+] with a digit between 1 and 7.
118    
119    8.  OP_NOT now supports any UTF character not just single-byte ones.
120    
121    9.  (*MARK) control verb is now supported by the JIT compiler.
122    
123    10. The command "./RunTest list" lists the available tests without actually
124        running any of them. (Because I keep forgetting what they all are.)
125    
126    11. Add PCRE_INFO_MAXLOOKBEHIND.
127    
128    12. Applied a (slightly modified) user-supplied patch that improves performance
129        when the heap is used for recursion (compiled with --disable-stack-for-
130        recursion). Instead of malloc and free for each heap frame each time a
131        logical recursion happens, frames are retained on a chain and re-used where
132        possible. This sometimes gives as much as 30% improvement.
133    
134    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
135        call.
136    
137    14. As documented, (*COMMIT) is now confined to within a positive assertion.
138    
139    15. It is now possible to link pcretest with libedit as an alternative to
140        libreadline.
141    
142    16. (*COMMIT) control verb is now supported by the JIT compiler.
143    
144    17. The Unicode data tables have been updated to Unicode 6.1.0.
145    
146    18. Added --file-list option to pcregrep.
147    
148    19. Added binary file support to pcregrep, including the -a, --binary-files,
149        -I, and --text options.
150    
151    20. The madvise function is renamed for posix_madvise for QNX compatibility
152        reasons. Fixed by Giuseppe D'Angelo.
153    
154    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
155    
156    22. Changed the option for creating symbolic links for 16-bit man pages from
157        -s to -sf so that re-installing does not cause issues.
158    
159    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
160    
161    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
162        matches in certain environments (the workspace was not being correctly
163        retained). Also added to pcre_dfa_exec() a simple plausibility check on
164        some of the workspace data at the beginning of a restart.
165    
166    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
167        was not doing so when it should - probably a typo introduced by SVN 528
168        (change 8.10/14).
169    
170    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
171        \w+ when the character tables indicated that \x{c4} was a word character.
172        There were several related cases, all because the tests for doing a table
173        lookup were testing for characters less than 127 instead of 255.
174    
175    27. If a pattern contains capturing parentheses that are not used in a match,
176        their slots in the ovector are set to -1. For those that are higher than
177        any matched groups, this happens at the end of processing. In the case when
178        there were back references that the ovector was too small to contain
179        (causing temporary malloc'd memory to be used during matching), and the
180        highest capturing number was not used, memory off the end of the ovector
181        was incorrectly being set to -1. (It was using the size of the temporary
182        memory instead of the true size.)
183    
184    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
185        ovector size, it uses memory at the end of the block that it has got.
186    
187    29. Check for an overlong MARK name and give an error at compile time. The
188        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
189    
190    30. JIT compiler update.
191    
192    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
193        Rill for the patch.
194    
195    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
196    
197    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
198    
199    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
200        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
201        was enabled.
202    
203    35. Improve JIT code generation for greedy plus quantifier.
204    
205    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
206        1 to "aa" instead of to an empty string. The bug affected repeated groups
207        that could potentially match an empty string.
208    
209    37. Optimizing single character iterators in JIT.
210    
211    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
212        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
213        codepoints that are too big for the mode are faulted, and in a UTF mode,
214        disallowed codepoints are also faulted.
215    
216    39. If PCRE was compiled with UTF support, in three places in the DFA
217        matcher there was code that should only have been obeyed in UTF mode, but
218        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
219        processing when bytes with values greater than 127 were present. In 16-bit
220        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
221        both cases the values are those that cannot be the first data item in a UTF
222        character. The three items that might have provoked this were recursions,
223        possessively repeated groups, and atomic groups.
224    
225    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
226        and pcregrep, because some OS require shared objects to be explicitly
227        passed to ld, causing the link step to fail if they are not.
228    
229    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
230        mode, patterns that started with \h* or \R* might be incorrectly matched.
231    
232    
233    Version 8.30 04-February-2012
234    -----------------------------
235    
236    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
237        name is defined in ctype.h.
238    
239    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
240        only in quite long subpatterns.
241    
242    3.  Removed the function pcre_info(), which has been obsolete and deprecated
243        since it was replaced by pcre_fullinfo() in February 2000.
244    
245    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
246        match a (*MARK), and the match failed at the start of the subject, a
247        reference to memory before the start of the subject could occur. This bug
248        was introduced by fix 17 of release 8.21.
249    
250    5.  A reference to an unset group with zero minimum repetition was giving
251        totally wrong answers (in non-JavaScript-compatibility mode). For example,
252        /(another)?(\1?)test/ matched against "hello world test". This bug was
253        introduced in release 8.13.
254    
255    6.  Add support for 16-bit character strings (a large amount of work involving
256        many changes and refactorings).
257    
258    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
259        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
260        from a file.
261    
262    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
263        rounding is not applied in this particular case).
264    
265    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
266        if they appear, or are escaped, in patterns.
267    
268    10. Get rid of a number of -Wunused-but-set-variable warnings.
269    
270    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
271        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
272        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
273        also returns the mark "x". This bug applied to capturing parentheses,
274        non-capturing parentheses, and atomic parentheses. It also applied to some
275        assertions.
276    
277    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
278        information out of configure.ac instead of relying on pcre.h.generic, which
279        is not stored in the repository.
280    
281    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
282        -lreadline.
283    
284    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
285    
286    15. Applied Graycode's patch to put the top-level frame on the stack rather
287        than the heap when not using the stack for recursion. This gives a
288        performance improvement in many cases when recursion is not deep.
289    
290    16. Experimental code added to "pcretest -C" to output the stack frame size.
291    
292    
293    Version 8.21 12-Dec-2011
294    ------------------------
295    
296    1.  Updating the JIT compiler.
297    
298    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
299        are added as well.
300    
301    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
302        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
303        calling _pcre_jit_exec. Some extra comments are added.
304    
305    4.  (*MARK) settings inside atomic groups that do not contain any capturing
306        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
307        was introduced by change 18 for 8.20.
308    
309    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
310        ECMA-262 standard.
311    
312    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
313        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
314        This bug was probably introduced by change 9 of 8.13.
315    
316    7.  While fixing 6 above, I noticed that a number of other items were being
317        incorrectly rejected as "not fixed length". This arose partly because newer
318        opcodes had not been added to the fixed-length checking code. I have (a)
319        corrected the bug and added tests for these items, and (b) arranged for an
320        error to occur if an unknown opcode is encountered while checking for fixed
321        length instead of just assuming "not fixed length". The items that were
322        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
323        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
324        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
325    
326    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
327        being incorrectly compiled and would have given unpredicatble results.
328    
329    9.  A possessively repeated subpattern with minimum repeat count greater than
330        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
331        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
332        the first (A) could occur when it should not.
333    
334    10. Add a cast and remove a redundant test from the code.
335    
336    11. JIT should use pcre_malloc/pcre_free for allocation.
337    
338    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
339        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
340        is anything other than /usr, -L is still shown).
341    
342    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
343    
344    14. Perl does not support \N without a following name in a [] class; PCRE now
345        also gives an error.
346    
347    15. If a forward reference was repeated with an upper limit of around 2000,
348        it caused the error "internal error: overran compiling workspace". The
349        maximum number of forward references (including repeats) was limited by the
350        internal workspace, and dependent on the LINK_SIZE. The code has been
351        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
352        the default depends on LINK_SIZE. There is a new upper limit (for safety)
353        of around 200,000 forward references. While doing this, I also speeded up
354        the filling in of repeated forward references.
355    
356    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
357        incorrectly expecting the subject to contain another "a" after the start.
358    
359    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
360        in the match, the SKIP should be ignored. This was not happening; instead
361        the SKIP was being treated as NOMATCH. For patterns such as
362        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
363        tested.
364    
365    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
366        now much more compatible with Perl, in particular in cases where the result
367        is a non-match for a non-anchored pattern. For example, if
368        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
369        "m", where previously it did not return a name. A side effect of this
370        change is that for partial matches, the last encountered mark name is
371        returned, as for non matches. A number of tests that were previously not
372        Perl-compatible have been moved into the Perl-compatible test files. The
373        refactoring has had the pleasing side effect of removing one argument from
374        the match() function, thus reducing its stack requirements.
375    
376    19. If the /S+ option was used in pcretest to study a pattern using JIT,
377        subsequent uses of /S (without +) incorrectly behaved like /S+.
378    
379    21. Retrieve executable code size support for the JIT compiler and fixing
380        some warnings.
381    
382    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
383        not work when the shorter character appeared right at the end of the
384        subject string.
385    
386    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
387        systems.
388    
389    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
390        output it when the /M option is used in pcretest.
391    
392    25. The CheckMan script was not being included in the distribution. Also, added
393        an explicit "perl" to run Perl scripts from the PrepareRelease script
394        because this is reportedly needed in Windows.
395    
396    26. If study data was being save in a file and studying had not found a set of
397        "starts with" bytes for the pattern, the data written to the file (though
398        never used) was taken from uninitialized memory and so caused valgrind to
399        complain.
400    
401    27. Updated RunTest.bat as provided by Sheri Pierce.
402    
403    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
404    
405    29. Computation of memory usage for the table of capturing group names was
406        giving an unnecessarily large value.
407    
408    
409    Version 8.20 21-Oct-2011
410    ------------------------
411    
412    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
413        a POSIX class. After further experiments with Perl, which convinced me that
414        Perl has bugs and confusions, a closing square bracket is no longer allowed
415        in a POSIX name. This bug also affected patterns with classes that started
416        with full stops.
417    
418    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
419        captured substring, but while checking the failing first alternative,
420        substring 1 is temporarily captured. If the output vector supplied to
421        pcre_exec() was not big enough for this capture, the yield of the function
422        was still zero ("insufficient space for captured substrings"). This cannot
423        be totally fixed without adding another stack variable, which seems a lot
424        of expense for a edge case. However, I have improved the situation in cases
425        such as /(a)(b)x|abc/ matched against "abc", where the return code
426        indicates that fewer than the maximum number of slots in the ovector have
427        been set.
428    
429    3.  Related to (2) above: when there are more back references in a pattern than
430        slots in the output vector, pcre_exec() uses temporary memory during
431        matching, and copies in the captures as far as possible afterwards. It was
432        using the entire output vector, but this conflicts with the specification
433        that only 2/3 is used for passing back captured substrings. Now it uses
434        only the first 2/3, for compatibility. This is, of course, another edge
435        case.
436    
437    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
438        main code base, and can be used by building with --enable-jit. When this is
439        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
440        runtime --no-jit option is given.
441    
442    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
443        ovector, the return from the function was zero, implying that there were
444        other matches that did not fit. The correct "exactly full" value is now
445        returned.
446    
447    6.  If a subpattern that was called recursively or as a subroutine contained
448        (*PRUNE) or any other control that caused it to give a non-standard return,
449        invalid errors such as "Error -26 (nested recursion at the same subject
450        position)" or even infinite loops could occur.
451    
452    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
453        computing the minimum length on reaching *ACCEPT, and so ended up with the
454        wrong value of 1 rather than 0. Further investigation indicates that
455        computing a minimum subject length in the presence of *ACCEPT is difficult
456        (think back references, subroutine calls), and so I have changed the code
457        so that no minimum is registered for a pattern that contains *ACCEPT.
458    
459    8.  If (*THEN) was present in the first (true) branch of a conditional group,
460        it was not handled as intended. [But see 16 below.]
461    
462    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
463        Sheri Pierce.
464    
465    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
466        the first byte in a match must be "a".
467    
468    11. Change 17 for 8.13 increased the recursion depth for patterns like
469        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
470        pattern contains any instances of (*THEN). If it does not, the old
471        optimizations are restored. It would be nice to do this on a per-group
472        basis, but at the moment that is not feasible.
473    
474    12. In some environments, the output of pcretest -C is CRLF terminated. This
475        broke RunTest's code that checks for the link size. A single white space
476        character after the value is now allowed for.
477    
478    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
479        For "fr", it uses the Windows-specific input and output files.
480    
481    14. If (*THEN) appeared in a group that was called recursively or as a
482        subroutine, it did not work as intended. [But see next item.]
483    
484    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
485        pattern fragments (but not containing any | characters). If A and B are
486        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
487        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
488        D. In other words, Perl considers parentheses that do not contain any |
489        characters to be part of a surrounding alternative, whereas PCRE was
490        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
491        differently. PCRE now behaves in the same way as Perl, except in the case
492        of subroutine/recursion calls such as (?1) which have in any case always
493        been different (but PCRE had them first :-).
494    
495    16. Related to 15 above: Perl does not treat the | in a conditional group as
496        creating alternatives. Such a group is treated in the same way as an
497        ordinary group without any | characters when processing (*THEN). PCRE has
498        been changed to match Perl's behaviour.
499    
500    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
501        RunGrepTest script failed.
502    
503    18. Change 22 for version 13 caused atomic groups to use more stack. This is
504        inevitable for groups that contain captures, but it can lead to a lot of
505        stack use in large patterns. The old behaviour has been restored for atomic
506        groups that do not contain any capturing parentheses.
507    
508    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
509        suppress the check for a minimum subject length at run time. (If it was
510        given to pcre_exec() or pcre_dfa_exec() it did work.)
511    
512    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
513        fail to work when decoding hex characters in data strings in EBCDIC
514        environments.
515    
516    21. It appears that in at least one Mac OS environment, the isxdigit() function
517        is implemented as a macro that evaluates to its argument more than once,
518        contravening the C 90 Standard (I haven't checked a later standard). There
519        was an instance in pcretest which caused it to go wrong when processing
520        \x{...} escapes in subject strings. The has been rewritten to avoid using
521        things like p++ in the argument of isxdigit().
522    
523    
524    Version 8.13 16-Aug-2011
525    ------------------------
526    
527    1.  The Unicode data tables have been updated to Unicode 6.0.0.
528    
529    2.  Two minor typos in pcre_internal.h have been fixed.
530    
531    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
532        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
533        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
534    
535    4.  There were a number of related bugs in the code for matching backrefences
536        caselessly in UTF-8 mode when codes for the characters concerned were
537        different numbers of bytes. For example, U+023A and U+2C65 are an upper
538        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
539        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
540        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
541        2-byte code at the end of the subject (it thought there wasn't enough data
542        left).
543    
544    5.  Comprehensive information about what went wrong is now returned by
545        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
546        as the output vector has at least 2 elements. The offset of the start of
547        the failing character and a reason code are placed in the vector.
548    
549    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
550        now returned is for the first byte of the failing character, instead of the
551        last byte inspected. This is an incompatible change, but I hope it is small
552        enough not to be a problem. It makes the returned offset consistent with
553        pcre_exec() and pcre_dfa_exec().
554    
555    7.  pcretest now gives a text phrase as well as the error number when
556        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
557        failure, the offset and reason code are output.
558    
559    8.  When \R was used with a maximizing quantifier it failed to skip backwards
560        over a \r\n pair if the subsequent match failed. Instead, it just skipped
561        back over a single character (\n). This seems wrong (because it treated the
562        two characters as a single entity when going forwards), conflicts with the
563        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
564        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
565        has been changed.
566    
567    9.  Some internal refactoring has changed the processing so that the handling
568        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
569        time (the PCRE_DOTALL option was changed this way some time ago: version
570        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
571        which was always a bit of a fudge. It also means that there is one less
572        argument for the match() function, which reduces its stack requirements
573        slightly. This change also fixes an incompatibility with Perl: the pattern
574        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
575    
576    10. More internal refactoring has drastically reduced the number of recursive
577        calls to match() for possessively repeated groups such as (abc)++ when
578        using pcre_exec().
579    
580    11. While implementing 10, a number of bugs in the handling of groups were
581        discovered and fixed:
582    
583        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
584        (a|)*(?1) gave a compile-time internal error.
585        ((a|)+)+  did not notice that the outer group could match an empty string.
586        (^a|^)+   was not marked as anchored.
587        (.*a|.*)+ was not marked as matching at start or after a newline.
588    
589    12. Yet more internal refactoring has removed another argument from the match()
590        function. Special calls to this function are now indicated by setting a
591        value in a variable in the "match data" data block.
592    
593    13. Be more explicit in pcre_study() instead of relying on "default" for
594        opcodes that mean there is no starting character; this means that when new
595        ones are added and accidentally left out of pcre_study(), testing should
596        pick them up.
597    
598    14. The -s option of pcretest has been documented for ages as being an old
599        synonym of -m (show memory usage). I have changed it to mean "force study
600        for every regex", that is, assume /S for every regex. This is similar to -i
601        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
602        using it. It makes it easier to run collections of tests with and without
603        study enabled, and thereby test pcre_study() more easily. All the standard
604        tests are now run with and without -s (but some patterns can be marked as
605        "never study" - see 20 below).
606    
607    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
608        restoration of the capturing data to the outer values was not happening
609        correctly.
610    
611    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
612        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
613        pattern had matched an empty string, and so incorrectly returned a no
614        match.
615    
616    17. There was optimizing code for the last branch of non-capturing parentheses,
617        and also for the obeyed branch of a conditional subexpression, which used
618        tail recursion to cut down on stack usage. Unfortunately, now that there is
619        the possibility of (*THEN) occurring in these branches, tail recursion is
620        no longer possible because the return has to be checked for (*THEN). These
621        two optimizations have therefore been removed. [But see 8.20/11 above.]
622    
623    18. If a pattern containing \R was studied, it was assumed that \R always
624        matched two bytes, thus causing the minimum subject length to be
625        incorrectly computed because \R can also match just one byte.
626    
627    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
628        was incorrectly computed.
629    
630    20. If /S is present twice on a test pattern in pcretest input, it now
631        *disables* studying, thereby overriding the use of -s on the command line
632        (see 14 above). This is necessary for one or two tests to keep the output
633        identical in both cases.
634    
635    21. When (*ACCEPT) was used in an assertion that matched an empty string and
636        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
637    
638    22. When an atomic group that contained a capturing parenthesis was
639        successfully matched, but the branch in which it appeared failed, the
640        capturing was not being forgotten if a higher numbered group was later
641        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
642        group 1 to "a", when in fact it should be unset. This applied to multi-
643        branched capturing and non-capturing groups, repeated or not, and also to
644        positive assertions (capturing in negative assertions does not happen
645        in PCRE) and also to nested atomic groups.
646    
647    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
648        subject after a captured substring, to make it easier to tell which of a
649        number of identical substrings has been captured.
650    
651    24. The way atomic groups are processed by pcre_exec() has been changed so that
652        if they are repeated, backtracking one repetition now resets captured
653        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
654        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
655        "aaa". Previously, it would have been "a". As part of this code
656        refactoring, the way recursive calls are handled has also been changed.
657    
658    25. If an assertion condition captured any substrings, they were not passed
659        back unless some other capturing happened later. For example, if
660        (?(?=(a))a) was matched against "a", no capturing was returned.
661    
662    26. When studying a pattern that contained subroutine calls or assertions,
663        the code for finding the minimum length of a possible match was handling
664        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
665        group 1 called group 2 while simultaneously a separate group 2 called group
666        1). A stack overflow occurred in this case. I have fixed this by limiting
667        the recursion depth to 10.
668    
669    27. Updated RunTest.bat in the distribution to the version supplied by Tom
670        Fortmann. This supports explicit test numbers on the command line, and has
671        argument validation and error reporting.
672    
673    28. An instance of \X with an unlimited repeat could fail if at any point the
674        first character it looked at was a mark character.
675    
676    29. Some minor code refactoring concerning Unicode properties and scripts
677        should reduce the stack requirement of match() slightly.
678    
679    30. Added the '=' option to pcretest to check the setting of unused capturing
680        slots at the end of the pattern, which are documented as being -1, but are
681        not included in the return count.
682    
683    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
684        compiled something random. Now it gives a compile-time error (as does
685        Perl).
686    
687    32. A *MARK encountered during the processing of a positive assertion is now
688        recorded and passed back (compatible with Perl).
689    
690    33. If --only-matching or --colour was set on a pcregrep call whose pattern
691        had alternative anchored branches, the search for a second match in a line
692        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
693        matched the line "0102" twice. The same bug affected patterns that started
694        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
695        twice.
696    
697    34. Previously, PCRE did not allow quantification of assertions. However, Perl
698        does, and because of capturing effects, quantifying parenthesized
699        assertions may at times be useful. Quantifiers are now allowed for
700        parenthesized assertions.
701    
702    35. A minor code tidy in pcre_compile() when checking options for \R usage.
703    
704    36. \g was being checked for fancy things in a character class, when it should
705        just be a literal "g".
706    
707    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
708        appearance of a nested POSIX class supersedes an apparent external class.
709        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
710        unescaped square brackets may also appear as part of class names. For
711        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
712        more like Perl. (But see 8.20/1 above.)
713    
714    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
715        was because it thought it was \N{name}, which is not supported).
716    
717    39. Add minix to OS list not supporting the -S option in pcretest.
718    
719    40. PCRE tries to detect cases of infinite recursion at compile time, but it
720        cannot analyze patterns in sufficient detail to catch mutual recursions
721        such as ((?1))((?2)). There is now a runtime test that gives an error if a
722        subgroup is called recursively as a subpattern for a second time at the
723        same position in the subject string. In previous releases this might have
724        been caught by the recursion limit, or it might have run out of stack.
725    
726    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
727        happen only once. PCRE was, however incorrectly giving a compile time error
728        "recursive call could loop indefinitely" because it cannot analyze the
729        pattern in sufficient detail. The compile time test no longer happens when
730        PCRE is compiling a conditional subpattern, but actual runaway loops are
731        now caught at runtime (see 40 above).
732    
733    42. It seems that Perl allows any characters other than a closing parenthesis
734        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
735        has been changed to be the same.
736    
737    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
738        as not to get warnings when autogen.sh is called. Also changed
739        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
740    
741    44. To help people who use pcregrep to scan files containing exceedingly long
742        lines, the following changes have been made:
743    
744        (a) The default value of the buffer size parameter has been increased from
745            8K to 20K. (The actual buffer used is three times this size.)
746    
747        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
748            PCRE is built.
749    
750        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
751            to be set at run time.
752    
753        (d) Numerical values in pcregrep options can be followed by K or M, for
754            example --buffer-size=50K.
755    
756        (e) If a line being scanned overflows pcregrep's buffer, an error is now
757            given and the return code is set to 2.
758    
759    45. Add a pointer to the latest mark to the callout data block.
760    
761    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
762        partial match of an empty string instead of no match. This was specific to
763        the use of ".".
764    
765    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
766        complete match instead of a partial match. This bug was dependent on both
767        the PCRE_UTF8 and PCRE_DOTALL options being set.
768    
769    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
770        starting byte set, because \b was not being ignored.
771    
772    
773    Version 8.12 15-Jan-2011
774    ------------------------
775    
776    1.  Fixed some typos in the markup of the man pages, and wrote a script that
777        checks for such things as part of the documentation building process.
778    
779    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
780        --match-limit and --recursion-limit options (added for 8.11). In
781        particular, this made one of the standard tests fail. (The integer value
782        went into the wrong half of a long int.)
783    
784    3.  If the --colour option was given to pcregrep with -v (invert match), it
785        did strange things, either producing crazy output, or crashing. It should,
786        of course, ignore a request for colour when reporting lines that do not
787        match.
788    
789    4.  Another pcregrep bug caused similar problems if --colour was specified with
790        -M (multiline) and the pattern match finished with a line ending.
791    
792    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
793        matched in multiline mode, the following line was shown as part of the
794        match. This seems wrong, so I have changed it.
795    
796    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
797        the check for further matches in the same line (so they could be coloured)
798        to overrun the end of the current line. If another match was found, it was
799        incorrectly shown (and then shown again when found in the next line).
800    
801    7.  If pcregrep was compiled under Windows, there was a reference to the
802        function pcregrep_exit() before it was defined. I am assuming this was
803        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
804        reported by a user. I've moved the definition above the reference.
805    
806    
807    Version 8.11 10-Dec-2010
808    ------------------------
809    
810    1.  (*THEN) was not working properly if there were untried alternatives prior
811        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
812        backtracked to try for "b" instead of moving to the next alternative branch
813        at the same level (in this case, to look for "c"). The Perl documentation
814        is clear that when (*THEN) is backtracked onto, it goes to the "next
815        alternative in the innermost enclosing group".
816    
817    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
818        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
819        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
820        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
821        (*THEN).
822    
823    3.  If \s appeared in a character class, it removed the VT character from
824        the class, even if it had been included by some previous item, for example
825        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
826        of \s, but is part of the POSIX "space" class.)
827    
828    4.  A partial match never returns an empty string (because you can always
829        match an empty string at the end of the subject); however the checking for
830        an empty string was starting at the "start of match" point. This has been
831        changed to the "earliest inspected character" point, because the returned
832        data for a partial match starts at this character. This means that, for
833        example, /(?<=abc)def/ gives a partial match for the subject "abc"
834        (previously it gave "no match").
835    
836    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
837        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
838        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
839        has an implication that the given string is incomplete (because a partial
840        match is preferred over a full match). For this reason, these items now
841        give a partial match in this situation. [Aside: previously, the one case
842        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
843        match rather than a full match, which was wrong by the old rules, but is
844        now correct.]
845    
846    6.  There was a bug in the handling of #-introduced comments, recognized when
847        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
848        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
849        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
850        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
851        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
852        places in pcre_compile().
853    
854    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
855        comments when looking ahead for named forward references to subpatterns,
856        the only newline sequence it recognized was NL. It now handles newlines
857        according to the set newline convention.
858    
859    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
860        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
861        cater for a lack of strerror(). These oversights have been fixed.
862    
863    9.  Added --match-limit and --recursion-limit to pcregrep.
864    
865    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
866    
867    11. When the -o option was used, pcregrep was setting a return code of 1, even
868        when matches were found, and --line-buffered was not being honoured.
869    
870    12. Added an optional parentheses number to the -o and --only-matching options
871        of pcregrep.
872    
873    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
874        can match an empty string. The code to do it in pcretest and pcredemo
875        needed fixing:
876    
877        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
878            only one byte after an empty string match just before CRLF (this case
879            just got forgotten; "any" and "anycrlf" were OK).
880    
881        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
882            mode when an empty string match preceded an ASCII character followed by
883            a non-ASCII character. (The code for advancing by one character rather
884            than one byte was nonsense.)
885    
886        (c) The pcredemo.c sample program did not have any code at all to handle
887            the cases when CRLF is a valid newline sequence.
888    
889    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
890        as a starting offset was within the subject string. There is now a new
891        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
892        negative or greater than the length of the string. In order to test this,
893        pcretest is extended to allow the setting of negative starting offsets.
894    
895    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
896        starting offset points to the beginning of a UTF-8 character was
897        unnecessarily clumsy. I tidied it up.
898    
899    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
900        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
901    
902    17. Nobody had reported that the --include_dir option, which was added in
903        release 7.7 should have been called --include-dir (hyphen, not underscore)
904        for compatibility with GNU grep. I have changed it to --include-dir, but
905        left --include_dir as an undocumented synonym, and the same for
906        --exclude-dir, though that is not available in GNU grep, at least as of
907        release 2.5.4.
908    
909    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
910        characters from a string of bytes) have been redefined so as not to use
911        loops, in order to improve performance in some environments. At the same
912        time, I abstracted some of the common code into auxiliary macros to save
913        repetition (this should not affect the compiled code).
914    
915    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
916        compile-time error is now given if \c is not followed by an ASCII
917        character, that is, a byte less than 128. (In EBCDIC mode, the code is
918        different, and any byte value is allowed.)
919    
920    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
921        START_OPTIMIZE option, which is now allowed at compile time - but just
922        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
923        to pcregrep and other applications that have no direct access to PCRE
924        options. The new /Y option in pcretest sets this option when calling
925        pcre_compile().
926    
927    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
928        back references. Groups containing recursive back references were forced to
929        be atomic by that change, but in the case of named groups, the amount of
930        memory required was incorrectly computed, leading to "Failed: internal
931        error: code overflow". This has been fixed.
932    
933    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
934        pcretest.c, to avoid build problems in some Borland environments.
935    
936    
937    Version 8.10 25-Jun-2010
938    ------------------------
939    
940    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
941        THEN.
942    
943    2.  (*ACCEPT) was not working when inside an atomic group.
944    
945    3.  Inside a character class, \B is treated as a literal by default, but
946        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
947        causes the error). The code is unchanged, but I tidied the documentation.
948    
949    4.  Inside a character class, PCRE always treated \R and \X as literals,
950        whereas Perl faults them if its -w option is set. I have changed PCRE so
951        that it faults them when PCRE_EXTRA is set.
952    
953    5.  Added support for \N, which always matches any character other than
954        newline. (It is the same as "." when PCRE_DOTALL is not set.)
955    
956    6.  When compiling pcregrep with newer versions of gcc which may have
957        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
958        declared with attribute warn_unused_result" were given. Just casting the
959        result to (void) does not stop the warnings; a more elaborate fudge is
960        needed. I've used a macro to implement this.
961    
962    7.  Minor change to pcretest.c to avoid a compiler warning.
963    
964    8.  Added four artifical Unicode properties to help with an option to make
965        \s etc use properties (see next item). The new properties are: Xan
966        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
967    
968    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
969        use Unicode properties. (*UCP) at the start of a pattern can be used to set
970        this option. Modified pcretest to add /W to test this facility. Added
971        REG_UCP to make it available via the POSIX interface.
972    
973    10. Added --line-buffered to pcregrep.
974    
975    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
976        studied, and the match started with a letter with a code point greater than
977        127 whose first byte was different to the first byte of the other case of
978        the letter, the other case of this starting letter was not recognized
979        (#976).
980    
981    12. If a pattern that was studied started with a repeated Unicode property
982        test, for example, \p{Nd}+, there was the theoretical possibility of
983        setting up an incorrect bitmap of starting bytes, but fortunately it could
984        not have actually happened in practice until change 8 above was made (it
985        added property types that matched character-matching opcodes).
986    
987    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
988        possible starting bytes for non-anchored patterns.
989    
990    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
991        \R, and also a number of cases that involve Unicode properties, both
992        explicit and implicit when PCRE_UCP is set.
993    
994    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
995        input, it could crash or give wrong results if characters with values
996        greater than 0xc0 were present in the subject string. (Detail: it assumed
997        UTF-8 input when processing these items.)
998    
999    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1000        size_t is 64-bit (#991).
1001    
1002    17. Added a check for running out of memory when PCRE is compiled with
1003        --disable-stack-for-recursion (#990).
1004    
1005    18. If the last data line in a file for pcretest does not have a newline on
1006        the end, a newline was missing in the output.
1007    
1008    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1009        less than 128) in its various bitmaps. However, there is a facility for
1010        generating tables according to the current locale when PCRE is compiled. It
1011        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1012        characters, are recognized by isspace() and therefore were getting set in
1013        these tables, and indeed these tables seem to approximate to ISO 8859. This
1014        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1015        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1016        which of course cannot start UTF-8 characters. I have changed the code so
1017        that only real ASCII characters (less than 128) and the correct starting
1018        bytes for UTF-8 encodings are set for characters greater than 127 when in
1019        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1020        altogether.)
1021    
1022    20. Added the /T option to pcretest so as to be able to run tests with non-
1023        standard character tables, thus making it possible to include the tests
1024        used for 19 above in the standard set of tests.
1025    
1026    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1027        reference to a subpattern the other side of a comment that contains an
1028        opening parenthesis caused either an internal compiling error, or a
1029        reference to the wrong subpattern.
1030    
1031    
1032    Version 8.02 19-Mar-2010
1033    ------------------------
1034    
1035    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1036    
1037    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1038        configured.
1039    
1040    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1041        original author of that file, following a query about its status.
1042    
1043    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1044        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1045    
1046    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1047        quantifier applied to a forward-referencing subroutine call, could compile
1048        incorrect code or give the error "internal error: previously-checked
1049        referenced subpattern not found".
1050    
1051    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1052        variables to point to external functions. For these systems, therefore,
1053        pcre_malloc etc. are now initialized to local functions that call the
1054        relevant global functions.
1055    
1056    7.  There were two entries missing in the vectors called coptable and poptable
1057        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1058        I've fixed the data, and added a kludgy way of testing at compile time that
1059        the lengths are correct (equal to the number of opcodes).
1060    
1061    8.  Following on from 7, I added a similar kludge to check the length of the
1062        eint vector in pcreposix.c.
1063    
1064    9.  Error texts for pcre_compile() are held as one long string to avoid too
1065        much relocation at load time. To find a text, the string is searched,
1066        counting zeros. There was no check for running off the end of the string,
1067        which could happen if a new error number was added without updating the
1068        string.
1069    
1070    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1071    
1072    11. \K was not working if it appeared in an atomic group or in a group that
1073        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1074        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1075        the assertion is positive, but not if it is negative.
1076    
1077    12. Change 11 fortuitously reduced the size of the stack frame used in the
1078        "match()" function of pcre_exec.c by one pointer. Forthcoming
1079        implementation of support for (*MARK) will need an extra pointer on the
1080        stack; I have reserved it now, so that the stack frame size does not
1081        decrease.
1082    
1083    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1084        item in branch that calls a recursion is a subroutine call - as in the
1085        second branch in the above example - was incorrectly given the compile-
1086        time error "recursive call could loop indefinitely" because pcre_compile()
1087        was not correctly checking the subroutine for matching a non-empty string.
1088    
1089    14. The checks for overrunning compiling workspace could trigger after an
1090        overrun had occurred. This is a "should never occur" error, but it can be
1091        triggered by pathological patterns such as hundreds of nested parentheses.
1092        The checks now trigger 100 bytes before the end of the workspace.
1093    
1094    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1095    
1096    
1097    Version 8.01 19-Jan-2010
1098    ------------------------
1099    
1100    1.  If a pattern contained a conditional subpattern with only one branch (in
1101        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1102        computed the wrong minimum data length (which is of course zero for such
1103        subpatterns). This could cause incorrect "no match" results.
1104    
1105    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1106        the pattern is reset in the first branch, pcre_compile() failed with
1107        "internal error: code overflow at offset...". This happened only when
1108        the reset was to the original external option setting. (An optimization
1109        abstracts leading options settings into an external setting, which was the
1110        cause of this.)
1111    
1112    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1113        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1114        assertion pattern did not match (meaning that the assertion was true), it
1115        was incorrectly treated as false if the SKIP had been reached during the
1116        matching. This also applied to assertions used as conditions.
1117    
1118    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1119        assertion subpattern, including such a pattern used as a condition,
1120        unpredictable results occurred, instead of the error return
1121        PCRE_ERROR_DFA_UITEM.
1122    
1123    5.  The C++ GlobalReplace function was not working like Perl for the special
1124        situation when an empty string is matched. It now does the fancy magic
1125        stuff that is necessary.
1126    
1127    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1128        removed. (These were left over from very, very early versions of PCRE.)
1129    
1130    7.  Some cosmetic changes to the code to make life easier when compiling it
1131        as part of something else:
1132    
1133        (a) Change DEBUG to PCRE_DEBUG.
1134    
1135        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1136            called "current" as "current_branch", to prevent a collision with the
1137            Linux macro when compiled as a kernel module.
1138    
1139        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1140            prevent a collision with the Linux macro when compiled as a kernel
1141            module.
1142    
1143    8.  In pcre_compile() there are some checks for integer overflows that used to
1144        cast potentially large values to (double). This has been changed to that
1145        when building, a check for int64_t is made, and if it is found, it is used
1146        instead, thus avoiding the use of floating point arithmetic. (There is no
1147        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1148        double.
1149    
1150    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1151        2005 (difference between two addresses compared to an unsigned value).
1152    
1153    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1154        custom one, because of the following reported problem in Windows:
1155    
1156          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1157              under Win32.
1158          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1159              therefore missing the function definition.
1160          - The compiler thus generates a "C" signature for the test function.
1161          - The linker fails to find the "C" function.
1162          - PCRE fails to configure if asked to do so against libbz2.
1163    
1164    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1165        messages were output:
1166    
1167          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1168          rerunning libtoolize, to keep the correct libtool macros in-tree.
1169          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1170    
1171        I have done both of these things.
1172    
1173    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1174        most of the time, it *can* run out if it is given a pattern that contains a
1175        runaway infinite recursion. I updated the discussion in the pcrestack man
1176        page.
1177    
1178    13. Now that we have gone to the x.xx style of version numbers, the minor
1179        version may start with zero. Using 08 or 09 is a bad idea because users
1180        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1181        interpreted as invalid octal numbers. I've updated the previous comment in
1182        configure.ac, and also added a check that gives an error if 08 or 09 are
1183        used.
1184    
1185    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1186        causing partial matching to fail when the end of the subject matched \W
1187        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1188    
1189    15. There were some discrepancies between the declarations in pcre_internal.h
1190        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1191        their definitions. The declarations used "const uschar *" and the
1192        definitions used USPTR. Even though USPTR is normally defined as "const
1193        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1194        reported that: "This difference in casting confuses some C++ compilers, for
1195        example, SunCC recognizes above declarations as different functions and
1196        generates broken code for hbpcre." I have changed the declarations to use
1197        USPTR.
1198    
1199    16. GNU libtool is named differently on some systems. The autogen.sh script now
1200        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1201        (FreeBSD).
1202    
1203    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1204        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1205        comment: "Figure out how to create a longlong from a string: strtoll and
1206        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1207        instance, but it only takes 2 args instead of 3!"
1208    
1209    18. A subtle bug concerned with back references has been fixed by a change of
1210        specification, with a corresponding code fix. A pattern such as
1211        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1212        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1213        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1214        same bug. Such groups have to be quantified to be useful, or contained
1215        inside another quantified group. (If there's no repetition, the reference
1216        can never match.) The problem arises because, having left the group and
1217        moved on to the rest of the pattern, a later failure that backtracks into
1218        the group uses the captured value from the final iteration of the group
1219        rather than the correct earlier one. I have fixed this in PCRE by forcing
1220        any group that contains a reference to itself to be an atomic group; that
1221        is, there cannot be any backtracking into it once it has completed. This is
1222        similar to recursive and subroutine calls.
1223    
1224    
1225    Version 8.00 19-Oct-09
1226    ----------------------
1227    
1228    1.  The table for translating pcre_compile() error codes into POSIX error codes
1229        was out-of-date, and there was no check on the pcre_compile() error code
1230        being within the table. This could lead to an OK return being given in
1231        error.
1232    
1233    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1234        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1235        in a Windows environment.
1236    
1237    3.  The pcregrep --count option prints the count for each file even when it is
1238        zero, as does GNU grep. However, pcregrep was also printing all files when
1239        --files-with-matches was added. Now, when both options are given, it prints
1240        counts only for those files that have at least one match. (GNU grep just
1241        prints the file name in this circumstance, but including the count seems
1242        more useful - otherwise, why use --count?) Also ensured that the
1243        combination -clh just lists non-zero counts, with no names.
1244    
1245    4.  The long form of the pcregrep -F option was incorrectly implemented as
1246        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1247        but it seems right to fix it, and I didn't think it was worth preserving
1248        the old behaviour.
1249    
1250    5.  The command line items --regex=pattern and --regexp=pattern were not
1251        recognized by pcregrep, which required --regex pattern or --regexp pattern
1252        (with a space rather than an '='). The man page documented the '=' forms,
1253        which are compatible with GNU grep; these now work.
1254    
1255    6.  No libpcreposix.pc file was created for pkg-config; there was just
1256        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1257    
1258    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1259        when UCP support is not needed, by modifying the Python script that
1260        generates it from Unicode data files. This should not matter if the module
1261        is correctly used as a library, but I received one complaint about 50K of
1262        unwanted data. My guess is that the person linked everything into his
1263        program rather than using a library. Anyway, it does no harm.
1264    
1265    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1266        was a minimum greater than 1 for a wide character in a possessive
1267        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1268        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1269        character. Chaos in the form of incorrect output or a compiling loop could
1270        result.
1271    
1272    9.  The restrictions on what a pattern can contain when partial matching is
1273        requested for pcre_exec() have been removed. All patterns can now be
1274        partially matched by this function. In addition, if there are at least two
1275        slots in the offset vector, the offset of the earliest inspected character
1276        for the match and the offset of the end of the subject are set in them when
1277        PCRE_ERROR_PARTIAL is returned.
1278    
1279    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1280        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1281        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1282        and may be more useful for multi-segment matching.
1283    
1284    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1285        used to be given if ever the end of the subject was reached; now it is
1286        given only if matching could not proceed because another character was
1287        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1288        string "Z", which now yields "no match" instead of "partial match". In the
1289        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1290        final character ended with (*FAIL).
1291    
1292    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1293        if the pattern had a "must contain" character that was already found in the
1294        earlier partial match, unless partial matching was again requested. For
1295        example, with the pattern /dog.(body)?/, the "must contain" character is
1296        "g". If the first part-match was for the string "dog", restarting with
1297        "sbody" failed. This bug has been fixed.
1298    
1299    13. The string returned by pcre_dfa_exec() after a partial match has been
1300        changed so that it starts at the first inspected character rather than the
1301        first character of the match. This makes a difference only if the pattern
1302        starts with a lookbehind assertion or \b or \B (\K is not supported by
1303        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1304        matching functions compatible, and I think it's the right thing to do.
1305    
1306    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1307        so that the demonstration program is easily available in environments where
1308        PCRE has not been installed from source.
1309    
1310    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1311        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1312        library.
1313    
1314    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1315        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1316        is not the first non-POSIX option to be added. Clearly some people find
1317        these options useful.
1318    
1319    17. If a caller to the POSIX matching function regexec() passes a non-zero
1320        value for nmatch with a NULL value for pmatch, the value of
1321        nmatch is forced to zero.
1322    
1323    18. RunGrepTest did not have a test for the availability of the -u option of
1324        the diff command, as RunTest does. It now checks in the same way as
1325        RunTest, and also checks for the -b option.
1326    
1327    19. If an odd number of negated classes containing just a single character
1328        interposed, within parentheses, between a forward reference to a named
1329        subpattern and the definition of the subpattern, compilation crashed with
1330        an internal error, complaining that it could not find the referenced
1331        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1332        [The bug was that it was starting one character too far in when skipping
1333        over the character class, thus treating the ] as data rather than
1334        terminating the class. This meant it could skip too much.]
1335    
1336    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1337        /g option in pcretest when the pattern contains \K, which makes it possible
1338        to have an empty string match not at the start, even when the pattern is
1339        anchored. Updated pcretest and pcredemo to use this option.
1340    
1341    21. If the maximum number of capturing subpatterns in a recursion was greater
1342        than the maximum at the outer level, the higher number was returned, but
1343        with unset values at the outer level. The correct (outer level) value is
1344        now given.
1345    
1346    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1347        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1348        make it do so. The string so far is captured, making this feature
1349        compatible with Perl.
1350    
1351    23. The tests have been re-organized, adding tests 11 and 12, to make it
1352        possible to check the Perl 5.10 features against Perl 5.10.
1353    
1354    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1355        pattern matches a fixed length string. PCRE did not allow this; now it
1356        does. Neither allows recursion.
1357    
1358    25. I finally figured out how to implement a request to provide the minimum
1359        length of subject string that was needed in order to match a given pattern.
1360        (It was back references and recursion that I had previously got hung up
1361        on.) This code has now been added to pcre_study(); it finds a lower bound
1362        to the length of subject needed. It is not necessarily the greatest lower
1363        bound, but using it to avoid searching strings that are too short does give
1364        some useful speed-ups. The value is available to calling programs via
1365        pcre_fullinfo().
1366    
1367    26. While implementing 25, I discovered to my embarrassment that pcretest had
1368        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1369        study optimizations had never been tested with that matching function.
1370        Oops. What is worse, even when it was passed study data, there was a bug in
1371        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1372        were also very few tests of studied patterns with pcre_dfa_exec().
1373    
1374    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1375        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1376        on the other side of the coin, they are no longer allowed to have different
1377        names, because these cannot be distinguished in PCRE, and this has caused
1378        confusion. (This is a difference from Perl.)
1379    
1380    28. When duplicate subpattern names are present (necessarily with different
1381        numbers, as required by 27 above), and a test is made by name in a
1382        conditional pattern, either for a subpattern having been matched, or for
1383        recursion in such a pattern, all the associated numbered subpatterns are
1384        tested, and the overall condition is true if the condition is true for any
1385        one of them. This is the way Perl works, and is also more like the way
1386        testing by number works.
1387    
1388    
1389    Version 7.9 11-Apr-09
1390    ---------------------
1391    
1392    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1393        (pcretest), all targets were linked against these libraries. This included
1394        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1395        libraries. This caused unwanted dependencies to be created. This problem
1396        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1397        pcretest is linked with readline.
1398    
1399    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1400        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1401        moved outside it again, because FALSE and TRUE are already defined in AIX,
1402        but BOOL is not.
1403    
1404    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1405        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1406    
1407    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1408        hyphens) following file names and line numbers when outputting matching
1409        lines. This is not true; no spaces are inserted. I have also clarified the
1410        wording for the --colour (or --color) option.
1411    
1412    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1413        was not coloured; this is different to GNU grep, so I have changed it to be
1414        the same.
1415    
1416    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1417        each matching line was coloured. Now it goes on to look for further matches
1418        of any of the test patterns, which is the same behaviour as GNU grep.
1419    
1420    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1421        doesn't make sense to accept an empty string match in pcregrep, so I have
1422        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1423        seems to be how GNU grep behaves.
1424    
1425    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1426        start or after a newline", because the conditional assertion was not being
1427        correctly handled. The rule now is that both the assertion and what follows
1428        in the first alternative must satisfy the test.
1429    
1430    9.  If auto-callout was enabled in a pattern with a conditional group whose
1431        condition was an assertion, PCRE could crash during matching, both with
1432        pcre_exec() and pcre_dfa_exec().
1433    
1434    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1435        used for matching.
1436    
1437    11. Unicode property support in character classes was not working for
1438        characters (bytes) greater than 127 when not in UTF-8 mode.
1439    
1440    12. Added the -M command line option to pcretest.
1441    
1442    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1443    
1444    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1445    
1446    16. Added comments and documentation about mis-use of no_arg in the C++
1447        wrapper.
1448    
1449    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1450        from Martin Jerabek that uses macro names for all relevant character and
1451        string constants.
1452    
1453    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1454        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1455        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1456        these, but not everybody uses configure.
1457    
1458    19. A conditional group that had only one branch was not being correctly
1459        recognized as an item that could match an empty string. This meant that an
1460        enclosing group might also not be so recognized, causing infinite looping
1461        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1462        with the subject "ab", where knowledge that the repeated group can match
1463        nothing is needed in order to break the loop.
1464    
1465    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1466        exec(), but without supplying a callout function, matching went wrong.
1467    
1468    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1469        leak if the size of the offset vector was greater than 30. When the vector
1470        is smaller, the saved offsets during recursion go onto a local stack
1471        vector, but for larger vectors malloc() is used. It was failing to free
1472        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1473        error, in fact).
1474    
1475    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1476        heapframe that is used only when UTF-8 support is enabled. This caused no
1477        problem, but was untidy.
1478    
1479    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1480        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1481        included within another project.
1482    
1483    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1484        slightly modified by me:
1485    
1486          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1487              not building pcregrep.
1488    
1489          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1490              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1491    
1492    25. Forward references, both numeric and by name, in patterns that made use of
1493        duplicate group numbers, could behave incorrectly or give incorrect errors,
1494        because when scanning forward to find the reference group, PCRE was not
1495        taking into account the duplicate group numbers. A pattern such as
1496        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1497    
1498    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1499        the feature of a custom pointer more persuasive (as requested by a user).
1500    
1501    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1502        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1503        pre-defined.
1504    
1505    28. Added support for (*UTF8) at the start of a pattern.
1506    
1507    29. Arrange for flags added by the "release type" setting in CMake to be shown
1508        in the configuration summary.
1509    
1510    
1511    Version 7.8 05-Sep-08
1512    ---------------------
1513    
1514    1.  Replaced UCP searching code with optimized version as implemented for Ad
1515        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1516        stage table and inline lookup instead of a function, giving speed ups of 2
1517        to 5 times on some simple patterns that I tested. Permission was given to
1518        distribute the MultiStage2.py script that generates the tables (it's not in
1519        the tarball, but is in the Subversion repository).
1520    
1521    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1522        scripts.
1523    
1524    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1525        a group with a zero qualifier. The result of the study could be incorrect,
1526        or the function might crash, depending on the pattern.
1527    
1528    4.  Caseless matching was not working for non-ASCII characters in back
1529        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1530        It now works when Unicode Property Support is available.
1531    
1532    5.  In pcretest, an escape such as \x{de} in the data was always generating
1533        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1534        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1535        truncation.
1536    
1537    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1538    
1539    7.  Added two (int) casts to pcregrep when printing the difference of two
1540        pointers, in case they are 64-bit values.
1541    
1542    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1543        test 2 if it fails.
1544    
1545    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1546        and a #define of that name to empty if it is not externally set. This is to
1547        allow users of MSVC to set it if necessary.
1548    
1549    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1550        the convenience functions in the pcre_get.c source file.
1551    
1552    11. An option change at the start of a pattern that had top-level alternatives
1553        could cause overwriting and/or a crash. This command provoked a crash in
1554        some environments:
1555    
1556          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1557    
1558        This potential security problem was recorded as CVE-2008-2371.
1559    
1560    12. For a pattern where the match had to start at the beginning or immediately
1561        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1562        pcre_dfa_exec() could read past the end of the passed subject if there was
1563        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1564        pcretest so that it places the subject at the end of its malloc-ed buffer.
1565    
1566    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1567        exec() might read past the end of the data buffer in UTF-8 mode.
1568    
1569    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1570        the data contained the byte 0x85 as part of a UTF-8 character within its
1571        first line. This applied both to normal and DFA matching.
1572    
1573    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1574        /^[^d]*?$/8 failed to match "abc".
1575    
1576    16. Added a missing copyright notice to pcrecpp_internal.h.
1577    
1578    17. Make it more clear in the documentation that values returned from
1579        pcre_exec() in ovector are byte offsets, not character counts.
1580    
1581    18. Tidied a few places to stop certain compilers from issuing warnings.
1582    
1583    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1584        supplied by Stefan Weber. I made a further small update for 7.8 because
1585        there is a change of source arrangements: the pcre_searchfuncs.c module is
1586        replaced by pcre_ucd.c.
1587    
1588    
1589    Version 7.7 07-May-08
1590    ---------------------
1591    
1592    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1593        a string to a long long, pretend we don't even have a long long." This is
1594        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1595    
1596    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1597        pre-7.6 versions, which defined a global no_arg variable instead of putting
1598        it in the RE class. (See also #8 below.)
1599    
1600    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1601        Lopes.
1602    
1603    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1604    
1605        (1) The include/exclude patterns were being applied to the whole pathnames
1606            of files, instead of just to the final components.
1607    
1608        (2) If there was more than one level of directory, the subdirectories were
1609            skipped unless they satisfied the include/exclude conditions. This is
1610            inconsistent with GNU grep (and could even be seen as contrary to the
1611            pcregrep specification - which I improved to make it absolutely clear).
1612            The action now is always to scan all levels of directory, and just
1613            apply the include/exclude patterns to regular files.
1614    
1615    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1616        --exclude_dir in the tests to avoid scanning .svn directories.
1617    
1618    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1619        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1620        doesn't support NULs in patterns.
1621    
1622    7.  Added some missing "const"s to declarations of static tables in
1623        pcre_compile.c and pcre_dfa_exec.c.
1624    
1625    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1626        caused by fix #2  above. (Subsequently also a second patch to fix the
1627        first patch. And a third patch - this was a messy problem.)
1628    
1629    9.  Applied Craig's patch to remove the use of push_back().
1630    
1631    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1632        matching function regexec().
1633    
1634    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1635        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1636        references. PCRE supports relative numbers with this syntax (I don't think
1637        Oniguruma does).
1638    
1639    12. Previously, a group with a zero repeat such as (...){0} was completely
1640        omitted from the compiled regex. However, this means that if the group
1641        was called as a subroutine from elsewhere in the pattern, things went wrong
1642        (an internal error was given). Such groups are now left in the compiled
1643        pattern, with a new opcode that causes them to be skipped at execution
1644        time.
1645    
1646    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1647        to the way PCRE behaves:
1648    
1649        (a) A lone ] character is dis-allowed (Perl treats it as data).
1650    
1651        (b) A back reference to an unmatched subpattern matches an empty string
1652            (Perl fails the current match path).
1653    
1654        (c) A data ] in a character class must be notated as \] because if the
1655            first data character in a class is ], it defines an empty class. (In
1656            Perl it is not possible to have an empty class.) The empty class []
1657            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1658            The negative empty class [^] matches any one character, independently
1659            of the DOTALL setting.
1660    
1661    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1662        non-existent subpattern following a character class starting with ']' and
1663        containing () gave an internal compiling error instead of "reference to
1664        non-existent subpattern". Fortunately, when the pattern did exist, the
1665        compiled code was correct. (When scanning forwards to check for the
1666        existencd of the subpattern, it was treating the data ']' as terminating
1667        the class, so got the count wrong. When actually compiling, the reference
1668        was subsequently set up correctly.)
1669    
1670    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1671        it was being rejected as not supported by pcre_dfa_exec(), even though
1672        other assertions are supported. I have made pcre_dfa_exec() support
1673        (*FAIL).
1674    
1675    16. The implementation of 13c above involved the invention of a new opcode,
1676        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1677        cannot be changed at match time, I realized I could make a small
1678        improvement to matching performance by compiling OP_ALLANY instead of
1679        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1680        on the OP_ANY path.
1681    
1682    17. Compiling pcretest on Windows with readline support failed without the
1683        following two fixes: (1) Make the unistd.h include conditional on
1684        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1685    
1686    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1687        ncurses library to be included for pcretest when ReadLine support is
1688        requested, but also to allow for it to be overridden. This patch came from
1689        Daniel Bergström.
1690    
1691    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1692        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1693        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1694        spotting this.
1695    
1696    
1697    Version 7.6 28-Jan-08
1698    ---------------------
1699    
1700    1.  A character class containing a very large number of characters with
1701        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1702        overflow.
1703    
1704    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1705        HAVE_LONG_LONG is not defined.
1706    
1707    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1708        bring them up to date and include new features. This patch includes:
1709    
1710        - Fixed PH's badly added libz and libbz2 support.
1711        - Fixed a problem with static linking.
1712        - Added pcredemo. [But later removed - see 7 below.]
1713        - Fixed dftables problem and added an option.
1714        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1715            HAVE_LONG_LONG.
1716        - Added readline support for pcretest.
1717        - Added an listing of the option settings after cmake has run.
1718    
1719    4.  A user submitted a patch to Makefile that makes it easy to create
1720        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1721        Makefile.am that cause it to include this special target, without
1722        affecting anything else. Note that the same mingw target plus all
1723        the other distribution libraries and programs are now supported
1724        when configuring with CMake (see 6 below) instead of with
1725        Configure/Make.
1726    
1727    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1728        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1729        exported in the Windows port". It has not yet been confirmed that the patch
1730        solves the problem, but it does no harm.
1731    
1732    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1733        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1734        with CMake, and also correct the comment about stack recursion.
1735    
1736    7.  Remove the automatic building of pcredemo from the ./configure system and
1737        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1738        of a program that users should build themselves after PCRE is installed, so
1739        building it automatically is not really right. What is more, it gave
1740        trouble in some build environments.
1741    
1742    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1743    
1744    
1745    Version 7.5 10-Jan-08
1746  ---------------------  ---------------------
1747    
1748  1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'  1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1749      values in parens when parsing an RE using the C++ wrapper."      values in parens when parsing an RE using the C++ wrapper."
1750    
1751  2.  Negative specials like \S did not work in character classes in UTF-8 mode.  2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1752      Characters greater than 255 were excluded from the class instead of being      Characters greater than 255 were excluded from the class instead of being
1753      included.      included.
1754    
1755  3.  The same bug as (2) above applied to negated POSIX classes such as  3.  The same bug as (2) above applied to negated POSIX classes such as
1756      [:^space:].      [:^space:].
1757    
1758  4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it  4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1759      defined or documented. It seems to have been a typo for PCRE_STATIC, so      defined or documented. It seems to have been a typo for PCRE_STATIC, so
1760      I have changed it.      I have changed it.
1761    
1762  5.  The construct (?&) was not diagnosed as a syntax error (it referenced the  5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1763      first named subpattern) and a construct such as (?&a) would reference the      first named subpattern) and a construct such as (?&a) would reference the
1764      first named subpattern whose name started with "a" (in other words, the      first named subpattern whose name started with "a" (in other words, the
1765      length check was missing). Both these problems are fixed. "Subpattern name      length check was missing). Both these problems are fixed. "Subpattern name
1766      expected" is now given for (?&) (a zero-length name), and this patch also      expected" is now given for (?&) (a zero-length name), and this patch also
1767      makes it give the same error for \k'' (previously it complained that that      makes it give the same error for \k'' (previously it complained that that
1768      was a reference to a non-existent subpattern).      was a reference to a non-existent subpattern).
1769    
1770  6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;  6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1771      this is right because (?- can be followed by option settings as well as by      this is right because (?- can be followed by option settings as well as by
1772      digits. I have, however, made the messages clearer.      digits. I have, however, made the messages clearer.
1773    
1774  7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns  7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1775      than the number used in the conditional) now cause a compile-time error.      than the number used in the conditional) now cause a compile-time error.
1776      This is actually not compatible with Perl, which accepts such patterns, but      This is actually not compatible with Perl, which accepts such patterns, but
1777      treats the conditional as always being FALSE (as PCRE used to), but it      treats the conditional as always being FALSE (as PCRE used to), but it
1778      seems to me that giving a diagnostic is better.      seems to me that giving a diagnostic is better.
1779    
1780  8.  Change "alphameric" to the more common word "alphanumeric" in comments  8.  Change "alphameric" to the more common word "alphanumeric" in comments
1781      and messages.      and messages.
1782    
1783  9.  Fix two occurrences of "backslash" in comments that should have been  9.  Fix two occurrences of "backslash" in comments that should have been
1784      "backspace".      "backspace".
1785    
1786  10. Remove two redundant lines of code that can never be obeyed (their function  10. Remove two redundant lines of code that can never be obeyed (their function
1787      was moved elsewhere).      was moved elsewhere).
1788    
1789  11. The program that makes PCRE's Unicode character property table had a bug  11. The program that makes PCRE's Unicode character property table had a bug
1790      which caused it to generate incorrect table entries for sequences of      which caused it to generate incorrect table entries for sequences of
1791      characters that have the same character type, but are in different scripts.      characters that have the same character type, but are in different scripts.
1792      It amalgamated them into a single range, with the script of the first of      It amalgamated them into a single range, with the script of the first of
1793      them. In other words, some characters were in the wrong script. There were      them. In other words, some characters were in the wrong script. There were
1794      thirteen such cases, affecting characters in the following ranges:      thirteen such cases, affecting characters in the following ranges:
1795    
1796        U+002b0 - U+002c1        U+002b0 - U+002c1
1797        U+0060c - U+0060d        U+0060c - U+0060d
1798        U+0061e - U+00612        U+0061e - U+00612
1799        U+0064b - U+0065e        U+0064b - U+0065e
1800        U+0074d - U+0076d        U+0074d - U+0076d
1801        U+01800 - U+01805        U+01800 - U+01805
# Line 66  Version 7.5 12-Nov-07 Line 1807  Version 7.5 12-Nov-07
1807        U+0fb46 - U+0fbb1        U+0fb46 - U+0fbb1
1808        U+10450 - U+1049d        U+10450 - U+1049d
1809    
1810    12. The -o option (show only the matching part of a line) for pcregrep was not
1811        compatible with GNU grep in that, if there was more than one match in a
1812        line, it showed only the first of them. It now behaves in the same way as
1813        GNU grep.
1814    
1815    13. If the -o and -v options were combined for pcregrep, it printed a blank
1816        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1817        does the same. The return code can be used to tell if there were any
1818        non-matching lines.
1819    
1820    14. Added --file-offsets and --line-offsets to pcregrep.
1821    
1822    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1823        infinitely looping recursion. The bug was that positive lookaheads were not
1824        being skipped when checking for a possible empty match (negative lookaheads
1825        and both kinds of lookbehind were skipped).
1826    
1827    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1828        inclusion of <windows.h> to before rather than after the definition of
1829        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1830    
1831    17. Specifying a possessive quantifier with a specific limit for a Unicode
1832        character property caused pcre_compile() to compile bad code, which led at
1833        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1834        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1835        caused the error; without that there was no problem.
1836    
1837    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1838    
1839    19. Added --enable-pcretest-libreadline.
1840    
1841    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1842        RE::GlobalReplace(). As a result, the number of replacements returned was
1843        double what it should be. I removed one of the increments, but Craig sent a
1844        later patch that removed the other one (the right fix) and added unit tests
1845        that check the return values (which was not done before).
1846    
1847    21. Several CMake things:
1848    
1849        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1850            the names libpcre and libpcreposix, not just pcre and pcreposix.
1851    
1852        (2) The above change means that pcretest and pcregrep are now correctly
1853            linked with the newly-built libraries, not previously installed ones.
1854    
1855        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1856    
1857    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1858        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1859        UTF-8 newline character). The key issue is that the pattern starts .*;
1860        this means that the match must be either at the beginning, or after a
1861        newline. The bug was in the code for advancing after a failed match and
1862        checking that the new position followed a newline. It was not taking
1863        account of UTF-8 characters correctly.
1864    
1865    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1866        character classes. PCRE was not treating the sequence [:...:] as a
1867        character class unless the ... were all letters. Perl, however, seems to
1868        allow any characters between [: and :], though of course it rejects as
1869        unknown any "names" that contain non-letters, because all the known class
1870        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1871        for example, whereas PCRE did not - it did not recognize a POSIX character
1872        class. This seemed a bit dangerous, so the code has been changed to be
1873        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1874        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1875        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1876        Perl does, and where it didn't before.
1877    
1878    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1879        Windows environments %n is disabled by default.
1880    
1881    
1882  Version 7.4 21-Sep-07  Version 7.4 21-Sep-07
1883  ---------------------  ---------------------

Legend:
Removed from v.277  
changed lines
  Added in v.1038

  ViewVC Help
Powered by ViewVC 1.1.5