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

Diff of /code/trunk/ChangeLog

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

revision 325 by ph10, Sat Mar 8 17:13:02 2008 UTC revision 1240 by ph10, Wed Jan 23 16:36:41 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.7 05-Mar-08  Version 8.33 xx-xxxx-201x
5    -------------------------
6    
7    1.  Added 'U' to some constants that are compared to unsigned integers, to
8        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
9        variables that are added to signed variables, to ensure the result is
10        signed and can be negated.
11    
12    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
13        CMake config files.
14    
15    3.  Revise the creation of config.h.generic so that all boolean macros are
16        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
17        overriding via -D on the command line possible.
18    
19    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
20        to unsigned int is reported to make a quite noticeable speed difference in
21        a specific Windows environment. Testing on Linux did also appear to show
22        some benefit (and it is clearly not harmful). Also fixed the definition of
23        Xop which should be unsigned.
24    
25    5.  Related to (4), changing the definition of the intermediate variable cc
26        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
27        improvements.
28    
29    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
30        unnecessary spaces.
31    
32    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
33        and later.
34    
35    
36    Version 8.32 30-November-2012
37    -----------------------------
38    
39    1.  Improved JIT compiler optimizations for first character search and single
40        character iterators.
41    
42    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
43        Patch by Daniel Richard G.
44    
45    3.  Single character iterator optimizations in the JIT compiler.
46    
47    4.  Improved JIT compiler optimizations for character ranges.
48    
49    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
50        Reported by Giuseppe D'Angelo.
51    
52    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
53        of a line, was being set incorrectly in cases where .* appeared inside
54        atomic brackets at the start of a pattern, or where there was a subsequent
55        *PRUNE or *SKIP.
56    
57    7.  Improved instruction cache flush for POWER/PowerPC.
58        Patch by Daniel Richard G.
59    
60    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
61        grep:
62    
63        (a) There is now no limit to the number of patterns to be matched.
64    
65        (b) An error is given if a pattern is too long.
66    
67        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
68            are now supported.
69    
70        (d) --exclude-from and --include-from (multiple use) have been added.
71    
72        (e) Exclusions and inclusions now apply to all files and directories, not
73            just to those obtained from scanning a directory recursively.
74    
75        (f) Multiple uses of -f and --file-list are now supported.
76    
77        (g) In a Windows environment, the default for -d has been changed from
78            "read" (the GNU grep default) to "skip", because otherwise the presence
79            of a directory in the file list provokes an error.
80    
81        (h) The documentation has been revised and clarified in places.
82    
83    9.  Improve the matching speed of capturing brackets.
84    
85    10. Changed the meaning of \X so that it now matches a Unicode extended
86        grapheme cluster.
87    
88    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
89        out POSIX threads when JIT support is configured.
90    
91    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
92    
93    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
94        structure could go wrong in environments where size_t is not the same size
95        as int.
96    
97    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
98        set.
99    
100    15. The EBCDIC support had decayed; later updates to the code had included
101        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
102        general tidy up of EBCDIC-related issues, and the documentation was also
103        not quite right. There is now a test that can be run on ASCII systems to
104        check some of the EBCDIC-related things (but is it not a full test).
105    
106    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
107        in a small tidy to the code.
108    
109    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
110    
111    18. If the --only-matching (-o) option in pcregrep is specified multiple
112        times, each one causes appropriate output. For example, -o1 -o2 outputs the
113        substrings matched by the 1st and 2nd capturing parentheses. A separating
114        string can be specified by --om-separator (default empty).
115    
116    19. Improving the first n character searches.
117    
118    20. Turn case lists for horizontal and vertical white space into macros so that
119        they are defined only once.
120    
121    21. This set of changes together give more compatible Unicode case-folding
122        behaviour for characters that have more than one other case when UCP
123        support is available.
124    
125        (a) The Unicode property table now has offsets into a new table of sets of
126            three or more characters that are case-equivalent. The MultiStage2.py
127            script that generates these tables (the pcre_ucd.c file) now scans
128            CaseFolding.txt instead of UnicodeData.txt for character case
129            information.
130    
131        (b) The code for adding characters or ranges of characters to a character
132            class has been abstracted into a generalized function that also handles
133            case-independence. In UTF-mode with UCP support, this uses the new data
134            to handle characters with more than one other case.
135    
136        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
137            whose other case is greater than 256 are now correctly matched
138            caselessly. Previously, the high codepoint matched the low one, but not
139            vice versa.
140    
141        (d) The processing of \h, \H, \v, and \ in character classes now makes use
142            of the new class addition function, using character lists defined as
143            macros alongside the case definitions of 20 above.
144    
145        (e) Caseless back references now work with characters that have more than
146            one other case.
147    
148        (f) General caseless matching of characters with more than one other case
149            is supported.
150    
151    22. Unicode character properties were updated from Unicode 6.2.0
152    
153    23. Improved CMake support under Windows. Patch by Daniel Richard G.
154    
155    24. Add support for 32-bit character strings, and UTF-32
156    
157    25. Major JIT compiler update (code refactoring and bugfixing).
158        Experimental Sparc 32 support is added.
159    
160    26. Applied a modified version of Daniel Richard G's patch to create
161        pcre.h.generic and config.h.generic by "make" instead of in the
162        PrepareRelease script.
163    
164    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
165        pcre_compile.c when checking for a zero character.
166    
167    28. Introducing a native interface for JIT. Through this interface, the compiled
168        machine code can be directly executed. The purpose of this interface is to
169        provide fast pattern matching, so several sanity checks are not performed.
170        However, feature tests are still performed. The new interface provides
171        1.4x speedup compared to the old one.
172    
173    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
174        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
175        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
176    
177    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
178        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
179        the "old" RFC 2279). Instead, it ended up passing a negative length to
180        pcre_exec().
181    
182    31. Add support for GCC's visibility feature to hide internal functions.
183    
184    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
185        "unknown -C option" after outputting 0 or 1.
186    
187    33. There is now support for generating a code coverage report for the test
188        suite in environments where gcc is the compiler and lcov is installed. This
189        is mainly for the benefit of the developers.
190    
191    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
192        unaddressable using valgrind annotations, allowing valgrind to detect
193        invalid memory accesses. This is mainly for the benefit of the developers.
194    
195    25. (*UTF) can now be used to start a pattern in any of the three libraries.
196    
197    26. Give configure error if --enable-cpp but no C++ compiler found.
198    
199    
200    Version 8.31 06-July-2012
201    -------------------------
202    
203    1.  Fixing a wrong JIT test case and some compiler warnings.
204    
205    2.  Removed a bashism from the RunTest script.
206    
207    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
208        to unsigned type, result still unsigned" that was given by an MS compiler
209        on encountering the code "-sizeof(xxx)".
210    
211    4.  Partial matching support is added to the JIT compiler.
212    
213    5.  Fixed several bugs concerned with partial matching of items that consist
214        of more than one character:
215    
216        (a) /^(..)\1/ did not partially match "aba" because checking references was
217            done on an "all or nothing" basis. This also applied to repeated
218            references.
219    
220        (b) \R did not give a hard partial match if \r was found at the end of the
221            subject.
222    
223        (c) \X did not give a hard partial match after matching one or more
224            characters at the end of the subject.
225    
226        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
227            a partial match for the string "\r".
228    
229        (e) When newline was set to CRLF, the metacharacter "." did not recognize
230            a partial match for a CR character at the end of the subject string.
231    
232    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
233        running pcretest, the text "(JIT)" added to the output whenever JIT is
234        actually used to run the match.
235    
236    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
237        or /S+[+] with a digit between 1 and 7.
238    
239    8.  OP_NOT now supports any UTF character not just single-byte ones.
240    
241    9.  (*MARK) control verb is now supported by the JIT compiler.
242    
243    10. The command "./RunTest list" lists the available tests without actually
244        running any of them. (Because I keep forgetting what they all are.)
245    
246    11. Add PCRE_INFO_MAXLOOKBEHIND.
247    
248    12. Applied a (slightly modified) user-supplied patch that improves performance
249        when the heap is used for recursion (compiled with --disable-stack-for-
250        recursion). Instead of malloc and free for each heap frame each time a
251        logical recursion happens, frames are retained on a chain and re-used where
252        possible. This sometimes gives as much as 30% improvement.
253    
254    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
255        call.
256    
257    14. As documented, (*COMMIT) is now confined to within a positive assertion.
258    
259    15. It is now possible to link pcretest with libedit as an alternative to
260        libreadline.
261    
262    16. (*COMMIT) control verb is now supported by the JIT compiler.
263    
264    17. The Unicode data tables have been updated to Unicode 6.1.0.
265    
266    18. Added --file-list option to pcregrep.
267    
268    19. Added binary file support to pcregrep, including the -a, --binary-files,
269        -I, and --text options.
270    
271    20. The madvise function is renamed for posix_madvise for QNX compatibility
272        reasons. Fixed by Giuseppe D'Angelo.
273    
274    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
275    
276    22. Changed the option for creating symbolic links for 16-bit man pages from
277        -s to -sf so that re-installing does not cause issues.
278    
279    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
280    
281    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
282        matches in certain environments (the workspace was not being correctly
283        retained). Also added to pcre_dfa_exec() a simple plausibility check on
284        some of the workspace data at the beginning of a restart.
285    
286    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
287        was not doing so when it should - probably a typo introduced by SVN 528
288        (change 8.10/14).
289    
290    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
291        \w+ when the character tables indicated that \x{c4} was a word character.
292        There were several related cases, all because the tests for doing a table
293        lookup were testing for characters less than 127 instead of 255.
294    
295    27. If a pattern contains capturing parentheses that are not used in a match,
296        their slots in the ovector are set to -1. For those that are higher than
297        any matched groups, this happens at the end of processing. In the case when
298        there were back references that the ovector was too small to contain
299        (causing temporary malloc'd memory to be used during matching), and the
300        highest capturing number was not used, memory off the end of the ovector
301        was incorrectly being set to -1. (It was using the size of the temporary
302        memory instead of the true size.)
303    
304    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
305        ovector size, it uses memory at the end of the block that it has got.
306    
307    29. Check for an overlong MARK name and give an error at compile time. The
308        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
309    
310    30. JIT compiler update.
311    
312    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
313        Rill for the patch.
314    
315    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
316    
317    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
318    
319    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
320        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
321        was enabled.
322    
323    35. Improve JIT code generation for greedy plus quantifier.
324    
325    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
326        1 to "aa" instead of to an empty string. The bug affected repeated groups
327        that could potentially match an empty string.
328    
329    37. Optimizing single character iterators in JIT.
330    
331    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
332        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
333        codepoints that are too big for the mode are faulted, and in a UTF mode,
334        disallowed codepoints are also faulted.
335    
336    39. If PCRE was compiled with UTF support, in three places in the DFA
337        matcher there was code that should only have been obeyed in UTF mode, but
338        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
339        processing when bytes with values greater than 127 were present. In 16-bit
340        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
341        both cases the values are those that cannot be the first data item in a UTF
342        character. The three items that might have provoked this were recursions,
343        possessively repeated groups, and atomic groups.
344    
345    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
346        and pcregrep, because some OS require shared objects to be explicitly
347        passed to ld, causing the link step to fail if they are not.
348    
349    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
350        mode, patterns that started with \h* or \R* might be incorrectly matched.
351    
352    
353    Version 8.30 04-February-2012
354    -----------------------------
355    
356    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
357        name is defined in ctype.h.
358    
359    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
360        only in quite long subpatterns.
361    
362    3.  Removed the function pcre_info(), which has been obsolete and deprecated
363        since it was replaced by pcre_fullinfo() in February 2000.
364    
365    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
366        match a (*MARK), and the match failed at the start of the subject, a
367        reference to memory before the start of the subject could occur. This bug
368        was introduced by fix 17 of release 8.21.
369    
370    5.  A reference to an unset group with zero minimum repetition was giving
371        totally wrong answers (in non-JavaScript-compatibility mode). For example,
372        /(another)?(\1?)test/ matched against "hello world test". This bug was
373        introduced in release 8.13.
374    
375    6.  Add support for 16-bit character strings (a large amount of work involving
376        many changes and refactorings).
377    
378    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
379        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
380        from a file.
381    
382    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
383        rounding is not applied in this particular case).
384    
385    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
386        if they appear, or are escaped, in patterns.
387    
388    10. Get rid of a number of -Wunused-but-set-variable warnings.
389    
390    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
391        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
392        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
393        also returns the mark "x". This bug applied to capturing parentheses,
394        non-capturing parentheses, and atomic parentheses. It also applied to some
395        assertions.
396    
397    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
398        information out of configure.ac instead of relying on pcre.h.generic, which
399        is not stored in the repository.
400    
401    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
402        -lreadline.
403    
404    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
405    
406    15. Applied Graycode's patch to put the top-level frame on the stack rather
407        than the heap when not using the stack for recursion. This gives a
408        performance improvement in many cases when recursion is not deep.
409    
410    16. Experimental code added to "pcretest -C" to output the stack frame size.
411    
412    
413    Version 8.21 12-Dec-2011
414    ------------------------
415    
416    1.  Updating the JIT compiler.
417    
418    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
419        are added as well.
420    
421    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
422        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
423        calling _pcre_jit_exec. Some extra comments are added.
424    
425    4.  (*MARK) settings inside atomic groups that do not contain any capturing
426        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
427        was introduced by change 18 for 8.20.
428    
429    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
430        ECMA-262 standard.
431    
432    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
433        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
434        This bug was probably introduced by change 9 of 8.13.
435    
436    7.  While fixing 6 above, I noticed that a number of other items were being
437        incorrectly rejected as "not fixed length". This arose partly because newer
438        opcodes had not been added to the fixed-length checking code. I have (a)
439        corrected the bug and added tests for these items, and (b) arranged for an
440        error to occur if an unknown opcode is encountered while checking for fixed
441        length instead of just assuming "not fixed length". The items that were
442        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
443        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
444        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
445    
446    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
447        being incorrectly compiled and would have given unpredicatble results.
448    
449    9.  A possessively repeated subpattern with minimum repeat count greater than
450        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
451        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
452        the first (A) could occur when it should not.
453    
454    10. Add a cast and remove a redundant test from the code.
455    
456    11. JIT should use pcre_malloc/pcre_free for allocation.
457    
458    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
459        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
460        is anything other than /usr, -L is still shown).
461    
462    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
463    
464    14. Perl does not support \N without a following name in a [] class; PCRE now
465        also gives an error.
466    
467    15. If a forward reference was repeated with an upper limit of around 2000,
468        it caused the error "internal error: overran compiling workspace". The
469        maximum number of forward references (including repeats) was limited by the
470        internal workspace, and dependent on the LINK_SIZE. The code has been
471        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
472        the default depends on LINK_SIZE. There is a new upper limit (for safety)
473        of around 200,000 forward references. While doing this, I also speeded up
474        the filling in of repeated forward references.
475    
476    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
477        incorrectly expecting the subject to contain another "a" after the start.
478    
479    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
480        in the match, the SKIP should be ignored. This was not happening; instead
481        the SKIP was being treated as NOMATCH. For patterns such as
482        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
483        tested.
484    
485    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
486        now much more compatible with Perl, in particular in cases where the result
487        is a non-match for a non-anchored pattern. For example, if
488        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
489        "m", where previously it did not return a name. A side effect of this
490        change is that for partial matches, the last encountered mark name is
491        returned, as for non matches. A number of tests that were previously not
492        Perl-compatible have been moved into the Perl-compatible test files. The
493        refactoring has had the pleasing side effect of removing one argument from
494        the match() function, thus reducing its stack requirements.
495    
496    19. If the /S+ option was used in pcretest to study a pattern using JIT,
497        subsequent uses of /S (without +) incorrectly behaved like /S+.
498    
499    21. Retrieve executable code size support for the JIT compiler and fixing
500        some warnings.
501    
502    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
503        not work when the shorter character appeared right at the end of the
504        subject string.
505    
506    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
507        systems.
508    
509    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
510        output it when the /M option is used in pcretest.
511    
512    25. The CheckMan script was not being included in the distribution. Also, added
513        an explicit "perl" to run Perl scripts from the PrepareRelease script
514        because this is reportedly needed in Windows.
515    
516    26. If study data was being save in a file and studying had not found a set of
517        "starts with" bytes for the pattern, the data written to the file (though
518        never used) was taken from uninitialized memory and so caused valgrind to
519        complain.
520    
521    27. Updated RunTest.bat as provided by Sheri Pierce.
522    
523    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
524    
525    29. Computation of memory usage for the table of capturing group names was
526        giving an unnecessarily large value.
527    
528    
529    Version 8.20 21-Oct-2011
530    ------------------------
531    
532    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
533        a POSIX class. After further experiments with Perl, which convinced me that
534        Perl has bugs and confusions, a closing square bracket is no longer allowed
535        in a POSIX name. This bug also affected patterns with classes that started
536        with full stops.
537    
538    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
539        captured substring, but while checking the failing first alternative,
540        substring 1 is temporarily captured. If the output vector supplied to
541        pcre_exec() was not big enough for this capture, the yield of the function
542        was still zero ("insufficient space for captured substrings"). This cannot
543        be totally fixed without adding another stack variable, which seems a lot
544        of expense for a edge case. However, I have improved the situation in cases
545        such as /(a)(b)x|abc/ matched against "abc", where the return code
546        indicates that fewer than the maximum number of slots in the ovector have
547        been set.
548    
549    3.  Related to (2) above: when there are more back references in a pattern than
550        slots in the output vector, pcre_exec() uses temporary memory during
551        matching, and copies in the captures as far as possible afterwards. It was
552        using the entire output vector, but this conflicts with the specification
553        that only 2/3 is used for passing back captured substrings. Now it uses
554        only the first 2/3, for compatibility. This is, of course, another edge
555        case.
556    
557    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
558        main code base, and can be used by building with --enable-jit. When this is
559        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
560        runtime --no-jit option is given.
561    
562    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
563        ovector, the return from the function was zero, implying that there were
564        other matches that did not fit. The correct "exactly full" value is now
565        returned.
566    
567    6.  If a subpattern that was called recursively or as a subroutine contained
568        (*PRUNE) or any other control that caused it to give a non-standard return,
569        invalid errors such as "Error -26 (nested recursion at the same subject
570        position)" or even infinite loops could occur.
571    
572    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
573        computing the minimum length on reaching *ACCEPT, and so ended up with the
574        wrong value of 1 rather than 0. Further investigation indicates that
575        computing a minimum subject length in the presence of *ACCEPT is difficult
576        (think back references, subroutine calls), and so I have changed the code
577        so that no minimum is registered for a pattern that contains *ACCEPT.
578    
579    8.  If (*THEN) was present in the first (true) branch of a conditional group,
580        it was not handled as intended. [But see 16 below.]
581    
582    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
583        Sheri Pierce.
584    
585    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
586        the first byte in a match must be "a".
587    
588    11. Change 17 for 8.13 increased the recursion depth for patterns like
589        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
590        pattern contains any instances of (*THEN). If it does not, the old
591        optimizations are restored. It would be nice to do this on a per-group
592        basis, but at the moment that is not feasible.
593    
594    12. In some environments, the output of pcretest -C is CRLF terminated. This
595        broke RunTest's code that checks for the link size. A single white space
596        character after the value is now allowed for.
597    
598    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
599        For "fr", it uses the Windows-specific input and output files.
600    
601    14. If (*THEN) appeared in a group that was called recursively or as a
602        subroutine, it did not work as intended. [But see next item.]
603    
604    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
605        pattern fragments (but not containing any | characters). If A and B are
606        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
607        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
608        D. In other words, Perl considers parentheses that do not contain any |
609        characters to be part of a surrounding alternative, whereas PCRE was
610        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
611        differently. PCRE now behaves in the same way as Perl, except in the case
612        of subroutine/recursion calls such as (?1) which have in any case always
613        been different (but PCRE had them first :-).
614    
615    16. Related to 15 above: Perl does not treat the | in a conditional group as
616        creating alternatives. Such a group is treated in the same way as an
617        ordinary group without any | characters when processing (*THEN). PCRE has
618        been changed to match Perl's behaviour.
619    
620    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
621        RunGrepTest script failed.
622    
623    18. Change 22 for version 13 caused atomic groups to use more stack. This is
624        inevitable for groups that contain captures, but it can lead to a lot of
625        stack use in large patterns. The old behaviour has been restored for atomic
626        groups that do not contain any capturing parentheses.
627    
628    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
629        suppress the check for a minimum subject length at run time. (If it was
630        given to pcre_exec() or pcre_dfa_exec() it did work.)
631    
632    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
633        fail to work when decoding hex characters in data strings in EBCDIC
634        environments.
635    
636    21. It appears that in at least one Mac OS environment, the isxdigit() function
637        is implemented as a macro that evaluates to its argument more than once,
638        contravening the C 90 Standard (I haven't checked a later standard). There
639        was an instance in pcretest which caused it to go wrong when processing
640        \x{...} escapes in subject strings. The has been rewritten to avoid using
641        things like p++ in the argument of isxdigit().
642    
643    
644    Version 8.13 16-Aug-2011
645    ------------------------
646    
647    1.  The Unicode data tables have been updated to Unicode 6.0.0.
648    
649    2.  Two minor typos in pcre_internal.h have been fixed.
650    
651    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
652        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
653        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
654    
655    4.  There were a number of related bugs in the code for matching backrefences
656        caselessly in UTF-8 mode when codes for the characters concerned were
657        different numbers of bytes. For example, U+023A and U+2C65 are an upper
658        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
659        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
660        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
661        2-byte code at the end of the subject (it thought there wasn't enough data
662        left).
663    
664    5.  Comprehensive information about what went wrong is now returned by
665        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
666        as the output vector has at least 2 elements. The offset of the start of
667        the failing character and a reason code are placed in the vector.
668    
669    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
670        now returned is for the first byte of the failing character, instead of the
671        last byte inspected. This is an incompatible change, but I hope it is small
672        enough not to be a problem. It makes the returned offset consistent with
673        pcre_exec() and pcre_dfa_exec().
674    
675    7.  pcretest now gives a text phrase as well as the error number when
676        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
677        failure, the offset and reason code are output.
678    
679    8.  When \R was used with a maximizing quantifier it failed to skip backwards
680        over a \r\n pair if the subsequent match failed. Instead, it just skipped
681        back over a single character (\n). This seems wrong (because it treated the
682        two characters as a single entity when going forwards), conflicts with the
683        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
684        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
685        has been changed.
686    
687    9.  Some internal refactoring has changed the processing so that the handling
688        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
689        time (the PCRE_DOTALL option was changed this way some time ago: version
690        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
691        which was always a bit of a fudge. It also means that there is one less
692        argument for the match() function, which reduces its stack requirements
693        slightly. This change also fixes an incompatibility with Perl: the pattern
694        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
695    
696    10. More internal refactoring has drastically reduced the number of recursive
697        calls to match() for possessively repeated groups such as (abc)++ when
698        using pcre_exec().
699    
700    11. While implementing 10, a number of bugs in the handling of groups were
701        discovered and fixed:
702    
703        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
704        (a|)*(?1) gave a compile-time internal error.
705        ((a|)+)+  did not notice that the outer group could match an empty string.
706        (^a|^)+   was not marked as anchored.
707        (.*a|.*)+ was not marked as matching at start or after a newline.
708    
709    12. Yet more internal refactoring has removed another argument from the match()
710        function. Special calls to this function are now indicated by setting a
711        value in a variable in the "match data" data block.
712    
713    13. Be more explicit in pcre_study() instead of relying on "default" for
714        opcodes that mean there is no starting character; this means that when new
715        ones are added and accidentally left out of pcre_study(), testing should
716        pick them up.
717    
718    14. The -s option of pcretest has been documented for ages as being an old
719        synonym of -m (show memory usage). I have changed it to mean "force study
720        for every regex", that is, assume /S for every regex. This is similar to -i
721        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
722        using it. It makes it easier to run collections of tests with and without
723        study enabled, and thereby test pcre_study() more easily. All the standard
724        tests are now run with and without -s (but some patterns can be marked as
725        "never study" - see 20 below).
726    
727    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
728        restoration of the capturing data to the outer values was not happening
729        correctly.
730    
731    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
732        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
733        pattern had matched an empty string, and so incorrectly returned a no
734        match.
735    
736    17. There was optimizing code for the last branch of non-capturing parentheses,
737        and also for the obeyed branch of a conditional subexpression, which used
738        tail recursion to cut down on stack usage. Unfortunately, now that there is
739        the possibility of (*THEN) occurring in these branches, tail recursion is
740        no longer possible because the return has to be checked for (*THEN). These
741        two optimizations have therefore been removed. [But see 8.20/11 above.]
742    
743    18. If a pattern containing \R was studied, it was assumed that \R always
744        matched two bytes, thus causing the minimum subject length to be
745        incorrectly computed because \R can also match just one byte.
746    
747    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
748        was incorrectly computed.
749    
750    20. If /S is present twice on a test pattern in pcretest input, it now
751        *disables* studying, thereby overriding the use of -s on the command line
752        (see 14 above). This is necessary for one or two tests to keep the output
753        identical in both cases.
754    
755    21. When (*ACCEPT) was used in an assertion that matched an empty string and
756        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
757    
758    22. When an atomic group that contained a capturing parenthesis was
759        successfully matched, but the branch in which it appeared failed, the
760        capturing was not being forgotten if a higher numbered group was later
761        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
762        group 1 to "a", when in fact it should be unset. This applied to multi-
763        branched capturing and non-capturing groups, repeated or not, and also to
764        positive assertions (capturing in negative assertions does not happen
765        in PCRE) and also to nested atomic groups.
766    
767    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
768        subject after a captured substring, to make it easier to tell which of a
769        number of identical substrings has been captured.
770    
771    24. The way atomic groups are processed by pcre_exec() has been changed so that
772        if they are repeated, backtracking one repetition now resets captured
773        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
774        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
775        "aaa". Previously, it would have been "a". As part of this code
776        refactoring, the way recursive calls are handled has also been changed.
777    
778    25. If an assertion condition captured any substrings, they were not passed
779        back unless some other capturing happened later. For example, if
780        (?(?=(a))a) was matched against "a", no capturing was returned.
781    
782    26. When studying a pattern that contained subroutine calls or assertions,
783        the code for finding the minimum length of a possible match was handling
784        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
785        group 1 called group 2 while simultaneously a separate group 2 called group
786        1). A stack overflow occurred in this case. I have fixed this by limiting
787        the recursion depth to 10.
788    
789    27. Updated RunTest.bat in the distribution to the version supplied by Tom
790        Fortmann. This supports explicit test numbers on the command line, and has
791        argument validation and error reporting.
792    
793    28. An instance of \X with an unlimited repeat could fail if at any point the
794        first character it looked at was a mark character.
795    
796    29. Some minor code refactoring concerning Unicode properties and scripts
797        should reduce the stack requirement of match() slightly.
798    
799    30. Added the '=' option to pcretest to check the setting of unused capturing
800        slots at the end of the pattern, which are documented as being -1, but are
801        not included in the return count.
802    
803    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
804        compiled something random. Now it gives a compile-time error (as does
805        Perl).
806    
807    32. A *MARK encountered during the processing of a positive assertion is now
808        recorded and passed back (compatible with Perl).
809    
810    33. If --only-matching or --colour was set on a pcregrep call whose pattern
811        had alternative anchored branches, the search for a second match in a line
812        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
813        matched the line "0102" twice. The same bug affected patterns that started
814        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
815        twice.
816    
817    34. Previously, PCRE did not allow quantification of assertions. However, Perl
818        does, and because of capturing effects, quantifying parenthesized
819        assertions may at times be useful. Quantifiers are now allowed for
820        parenthesized assertions.
821    
822    35. A minor code tidy in pcre_compile() when checking options for \R usage.
823    
824    36. \g was being checked for fancy things in a character class, when it should
825        just be a literal "g".
826    
827    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
828        appearance of a nested POSIX class supersedes an apparent external class.
829        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
830        unescaped square brackets may also appear as part of class names. For
831        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
832        more like Perl. (But see 8.20/1 above.)
833    
834    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
835        was because it thought it was \N{name}, which is not supported).
836    
837    39. Add minix to OS list not supporting the -S option in pcretest.
838    
839    40. PCRE tries to detect cases of infinite recursion at compile time, but it
840        cannot analyze patterns in sufficient detail to catch mutual recursions
841        such as ((?1))((?2)). There is now a runtime test that gives an error if a
842        subgroup is called recursively as a subpattern for a second time at the
843        same position in the subject string. In previous releases this might have
844        been caught by the recursion limit, or it might have run out of stack.
845    
846    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
847        happen only once. PCRE was, however incorrectly giving a compile time error
848        "recursive call could loop indefinitely" because it cannot analyze the
849        pattern in sufficient detail. The compile time test no longer happens when
850        PCRE is compiling a conditional subpattern, but actual runaway loops are
851        now caught at runtime (see 40 above).
852    
853    42. It seems that Perl allows any characters other than a closing parenthesis
854        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
855        has been changed to be the same.
856    
857    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
858        as not to get warnings when autogen.sh is called. Also changed
859        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
860    
861    44. To help people who use pcregrep to scan files containing exceedingly long
862        lines, the following changes have been made:
863    
864        (a) The default value of the buffer size parameter has been increased from
865            8K to 20K. (The actual buffer used is three times this size.)
866    
867        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
868            PCRE is built.
869    
870        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
871            to be set at run time.
872    
873        (d) Numerical values in pcregrep options can be followed by K or M, for
874            example --buffer-size=50K.
875    
876        (e) If a line being scanned overflows pcregrep's buffer, an error is now
877            given and the return code is set to 2.
878    
879    45. Add a pointer to the latest mark to the callout data block.
880    
881    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
882        partial match of an empty string instead of no match. This was specific to
883        the use of ".".
884    
885    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
886        complete match instead of a partial match. This bug was dependent on both
887        the PCRE_UTF8 and PCRE_DOTALL options being set.
888    
889    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
890        starting byte set, because \b was not being ignored.
891    
892    
893    Version 8.12 15-Jan-2011
894    ------------------------
895    
896    1.  Fixed some typos in the markup of the man pages, and wrote a script that
897        checks for such things as part of the documentation building process.
898    
899    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
900        --match-limit and --recursion-limit options (added for 8.11). In
901        particular, this made one of the standard tests fail. (The integer value
902        went into the wrong half of a long int.)
903    
904    3.  If the --colour option was given to pcregrep with -v (invert match), it
905        did strange things, either producing crazy output, or crashing. It should,
906        of course, ignore a request for colour when reporting lines that do not
907        match.
908    
909    4.  Another pcregrep bug caused similar problems if --colour was specified with
910        -M (multiline) and the pattern match finished with a line ending.
911    
912    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
913        matched in multiline mode, the following line was shown as part of the
914        match. This seems wrong, so I have changed it.
915    
916    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
917        the check for further matches in the same line (so they could be coloured)
918        to overrun the end of the current line. If another match was found, it was
919        incorrectly shown (and then shown again when found in the next line).
920    
921    7.  If pcregrep was compiled under Windows, there was a reference to the
922        function pcregrep_exit() before it was defined. I am assuming this was
923        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
924        reported by a user. I've moved the definition above the reference.
925    
926    
927    Version 8.11 10-Dec-2010
928    ------------------------
929    
930    1.  (*THEN) was not working properly if there were untried alternatives prior
931        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
932        backtracked to try for "b" instead of moving to the next alternative branch
933        at the same level (in this case, to look for "c"). The Perl documentation
934        is clear that when (*THEN) is backtracked onto, it goes to the "next
935        alternative in the innermost enclosing group".
936    
937    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
938        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
939        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
940        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
941        (*THEN).
942    
943    3.  If \s appeared in a character class, it removed the VT character from
944        the class, even if it had been included by some previous item, for example
945        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
946        of \s, but is part of the POSIX "space" class.)
947    
948    4.  A partial match never returns an empty string (because you can always
949        match an empty string at the end of the subject); however the checking for
950        an empty string was starting at the "start of match" point. This has been
951        changed to the "earliest inspected character" point, because the returned
952        data for a partial match starts at this character. This means that, for
953        example, /(?<=abc)def/ gives a partial match for the subject "abc"
954        (previously it gave "no match").
955    
956    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
957        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
958        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
959        has an implication that the given string is incomplete (because a partial
960        match is preferred over a full match). For this reason, these items now
961        give a partial match in this situation. [Aside: previously, the one case
962        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
963        match rather than a full match, which was wrong by the old rules, but is
964        now correct.]
965    
966    6.  There was a bug in the handling of #-introduced comments, recognized when
967        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
968        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
969        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
970        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
971        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
972        places in pcre_compile().
973    
974    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
975        comments when looking ahead for named forward references to subpatterns,
976        the only newline sequence it recognized was NL. It now handles newlines
977        according to the set newline convention.
978    
979    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
980        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
981        cater for a lack of strerror(). These oversights have been fixed.
982    
983    9.  Added --match-limit and --recursion-limit to pcregrep.
984    
985    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
986    
987    11. When the -o option was used, pcregrep was setting a return code of 1, even
988        when matches were found, and --line-buffered was not being honoured.
989    
990    12. Added an optional parentheses number to the -o and --only-matching options
991        of pcregrep.
992    
993    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
994        can match an empty string. The code to do it in pcretest and pcredemo
995        needed fixing:
996    
997        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
998            only one byte after an empty string match just before CRLF (this case
999            just got forgotten; "any" and "anycrlf" were OK).
1000    
1001        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1002            mode when an empty string match preceded an ASCII character followed by
1003            a non-ASCII character. (The code for advancing by one character rather
1004            than one byte was nonsense.)
1005    
1006        (c) The pcredemo.c sample program did not have any code at all to handle
1007            the cases when CRLF is a valid newline sequence.
1008    
1009    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1010        as a starting offset was within the subject string. There is now a new
1011        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1012        negative or greater than the length of the string. In order to test this,
1013        pcretest is extended to allow the setting of negative starting offsets.
1014    
1015    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1016        starting offset points to the beginning of a UTF-8 character was
1017        unnecessarily clumsy. I tidied it up.
1018    
1019    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1020        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1021    
1022    17. Nobody had reported that the --include_dir option, which was added in
1023        release 7.7 should have been called --include-dir (hyphen, not underscore)
1024        for compatibility with GNU grep. I have changed it to --include-dir, but
1025        left --include_dir as an undocumented synonym, and the same for
1026        --exclude-dir, though that is not available in GNU grep, at least as of
1027        release 2.5.4.
1028    
1029    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1030        characters from a string of bytes) have been redefined so as not to use
1031        loops, in order to improve performance in some environments. At the same
1032        time, I abstracted some of the common code into auxiliary macros to save
1033        repetition (this should not affect the compiled code).
1034    
1035    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1036        compile-time error is now given if \c is not followed by an ASCII
1037        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1038        different, and any byte value is allowed.)
1039    
1040    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1041        START_OPTIMIZE option, which is now allowed at compile time - but just
1042        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1043        to pcregrep and other applications that have no direct access to PCRE
1044        options. The new /Y option in pcretest sets this option when calling
1045        pcre_compile().
1046    
1047    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1048        back references. Groups containing recursive back references were forced to
1049        be atomic by that change, but in the case of named groups, the amount of
1050        memory required was incorrectly computed, leading to "Failed: internal
1051        error: code overflow". This has been fixed.
1052    
1053    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1054        pcretest.c, to avoid build problems in some Borland environments.
1055    
1056    
1057    Version 8.10 25-Jun-2010
1058    ------------------------
1059    
1060    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1061        THEN.
1062    
1063    2.  (*ACCEPT) was not working when inside an atomic group.
1064    
1065    3.  Inside a character class, \B is treated as a literal by default, but
1066        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1067        causes the error). The code is unchanged, but I tidied the documentation.
1068    
1069    4.  Inside a character class, PCRE always treated \R and \X as literals,
1070        whereas Perl faults them if its -w option is set. I have changed PCRE so
1071        that it faults them when PCRE_EXTRA is set.
1072    
1073    5.  Added support for \N, which always matches any character other than
1074        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1075    
1076    6.  When compiling pcregrep with newer versions of gcc which may have
1077        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1078        declared with attribute warn_unused_result" were given. Just casting the
1079        result to (void) does not stop the warnings; a more elaborate fudge is
1080        needed. I've used a macro to implement this.
1081    
1082    7.  Minor change to pcretest.c to avoid a compiler warning.
1083    
1084    8.  Added four artifical Unicode properties to help with an option to make
1085        \s etc use properties (see next item). The new properties are: Xan
1086        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1087    
1088    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1089        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1090        this option. Modified pcretest to add /W to test this facility. Added
1091        REG_UCP to make it available via the POSIX interface.
1092    
1093    10. Added --line-buffered to pcregrep.
1094    
1095    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1096        studied, and the match started with a letter with a code point greater than
1097        127 whose first byte was different to the first byte of the other case of
1098        the letter, the other case of this starting letter was not recognized
1099        (#976).
1100    
1101    12. If a pattern that was studied started with a repeated Unicode property
1102        test, for example, \p{Nd}+, there was the theoretical possibility of
1103        setting up an incorrect bitmap of starting bytes, but fortunately it could
1104        not have actually happened in practice until change 8 above was made (it
1105        added property types that matched character-matching opcodes).
1106    
1107    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1108        possible starting bytes for non-anchored patterns.
1109    
1110    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1111        \R, and also a number of cases that involve Unicode properties, both
1112        explicit and implicit when PCRE_UCP is set.
1113    
1114    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1115        input, it could crash or give wrong results if characters with values
1116        greater than 0xc0 were present in the subject string. (Detail: it assumed
1117        UTF-8 input when processing these items.)
1118    
1119    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1120        size_t is 64-bit (#991).
1121    
1122    17. Added a check for running out of memory when PCRE is compiled with
1123        --disable-stack-for-recursion (#990).
1124    
1125    18. If the last data line in a file for pcretest does not have a newline on
1126        the end, a newline was missing in the output.
1127    
1128    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1129        less than 128) in its various bitmaps. However, there is a facility for
1130        generating tables according to the current locale when PCRE is compiled. It
1131        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1132        characters, are recognized by isspace() and therefore were getting set in
1133        these tables, and indeed these tables seem to approximate to ISO 8859. This
1134        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1135        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1136        which of course cannot start UTF-8 characters. I have changed the code so
1137        that only real ASCII characters (less than 128) and the correct starting
1138        bytes for UTF-8 encodings are set for characters greater than 127 when in
1139        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1140        altogether.)
1141    
1142    20. Added the /T option to pcretest so as to be able to run tests with non-
1143        standard character tables, thus making it possible to include the tests
1144        used for 19 above in the standard set of tests.
1145    
1146    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1147        reference to a subpattern the other side of a comment that contains an
1148        opening parenthesis caused either an internal compiling error, or a
1149        reference to the wrong subpattern.
1150    
1151    
1152    Version 8.02 19-Mar-2010
1153    ------------------------
1154    
1155    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1156    
1157    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1158        configured.
1159    
1160    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1161        original author of that file, following a query about its status.
1162    
1163    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1164        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1165    
1166    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1167        quantifier applied to a forward-referencing subroutine call, could compile
1168        incorrect code or give the error "internal error: previously-checked
1169        referenced subpattern not found".
1170    
1171    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1172        variables to point to external functions. For these systems, therefore,
1173        pcre_malloc etc. are now initialized to local functions that call the
1174        relevant global functions.
1175    
1176    7.  There were two entries missing in the vectors called coptable and poptable
1177        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1178        I've fixed the data, and added a kludgy way of testing at compile time that
1179        the lengths are correct (equal to the number of opcodes).
1180    
1181    8.  Following on from 7, I added a similar kludge to check the length of the
1182        eint vector in pcreposix.c.
1183    
1184    9.  Error texts for pcre_compile() are held as one long string to avoid too
1185        much relocation at load time. To find a text, the string is searched,
1186        counting zeros. There was no check for running off the end of the string,
1187        which could happen if a new error number was added without updating the
1188        string.
1189    
1190    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1191    
1192    11. \K was not working if it appeared in an atomic group or in a group that
1193        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1194        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1195        the assertion is positive, but not if it is negative.
1196    
1197    12. Change 11 fortuitously reduced the size of the stack frame used in the
1198        "match()" function of pcre_exec.c by one pointer. Forthcoming
1199        implementation of support for (*MARK) will need an extra pointer on the
1200        stack; I have reserved it now, so that the stack frame size does not
1201        decrease.
1202    
1203    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1204        item in branch that calls a recursion is a subroutine call - as in the
1205        second branch in the above example - was incorrectly given the compile-
1206        time error "recursive call could loop indefinitely" because pcre_compile()
1207        was not correctly checking the subroutine for matching a non-empty string.
1208    
1209    14. The checks for overrunning compiling workspace could trigger after an
1210        overrun had occurred. This is a "should never occur" error, but it can be
1211        triggered by pathological patterns such as hundreds of nested parentheses.
1212        The checks now trigger 100 bytes before the end of the workspace.
1213    
1214    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1215    
1216    
1217    Version 8.01 19-Jan-2010
1218    ------------------------
1219    
1220    1.  If a pattern contained a conditional subpattern with only one branch (in
1221        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1222        computed the wrong minimum data length (which is of course zero for such
1223        subpatterns). This could cause incorrect "no match" results.
1224    
1225    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1226        the pattern is reset in the first branch, pcre_compile() failed with
1227        "internal error: code overflow at offset...". This happened only when
1228        the reset was to the original external option setting. (An optimization
1229        abstracts leading options settings into an external setting, which was the
1230        cause of this.)
1231    
1232    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1233        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1234        assertion pattern did not match (meaning that the assertion was true), it
1235        was incorrectly treated as false if the SKIP had been reached during the
1236        matching. This also applied to assertions used as conditions.
1237    
1238    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1239        assertion subpattern, including such a pattern used as a condition,
1240        unpredictable results occurred, instead of the error return
1241        PCRE_ERROR_DFA_UITEM.
1242    
1243    5.  The C++ GlobalReplace function was not working like Perl for the special
1244        situation when an empty string is matched. It now does the fancy magic
1245        stuff that is necessary.
1246    
1247    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1248        removed. (These were left over from very, very early versions of PCRE.)
1249    
1250    7.  Some cosmetic changes to the code to make life easier when compiling it
1251        as part of something else:
1252    
1253        (a) Change DEBUG to PCRE_DEBUG.
1254    
1255        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1256            called "current" as "current_branch", to prevent a collision with the
1257            Linux macro when compiled as a kernel module.
1258    
1259        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1260            prevent a collision with the Linux macro when compiled as a kernel
1261            module.
1262    
1263    8.  In pcre_compile() there are some checks for integer overflows that used to
1264        cast potentially large values to (double). This has been changed to that
1265        when building, a check for int64_t is made, and if it is found, it is used
1266        instead, thus avoiding the use of floating point arithmetic. (There is no
1267        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1268        double.
1269    
1270    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1271        2005 (difference between two addresses compared to an unsigned value).
1272    
1273    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1274        custom one, because of the following reported problem in Windows:
1275    
1276          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1277              under Win32.
1278          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1279              therefore missing the function definition.
1280          - The compiler thus generates a "C" signature for the test function.
1281          - The linker fails to find the "C" function.
1282          - PCRE fails to configure if asked to do so against libbz2.
1283    
1284    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1285        messages were output:
1286    
1287          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1288          rerunning libtoolize, to keep the correct libtool macros in-tree.
1289          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1290    
1291        I have done both of these things.
1292    
1293    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1294        most of the time, it *can* run out if it is given a pattern that contains a
1295        runaway infinite recursion. I updated the discussion in the pcrestack man
1296        page.
1297    
1298    13. Now that we have gone to the x.xx style of version numbers, the minor
1299        version may start with zero. Using 08 or 09 is a bad idea because users
1300        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1301        interpreted as invalid octal numbers. I've updated the previous comment in
1302        configure.ac, and also added a check that gives an error if 08 or 09 are
1303        used.
1304    
1305    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1306        causing partial matching to fail when the end of the subject matched \W
1307        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1308    
1309    15. There were some discrepancies between the declarations in pcre_internal.h
1310        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1311        their definitions. The declarations used "const uschar *" and the
1312        definitions used USPTR. Even though USPTR is normally defined as "const
1313        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1314        reported that: "This difference in casting confuses some C++ compilers, for
1315        example, SunCC recognizes above declarations as different functions and
1316        generates broken code for hbpcre." I have changed the declarations to use
1317        USPTR.
1318    
1319    16. GNU libtool is named differently on some systems. The autogen.sh script now
1320        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1321        (FreeBSD).
1322    
1323    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1324        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1325        comment: "Figure out how to create a longlong from a string: strtoll and
1326        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1327        instance, but it only takes 2 args instead of 3!"
1328    
1329    18. A subtle bug concerned with back references has been fixed by a change of
1330        specification, with a corresponding code fix. A pattern such as
1331        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1332        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1333        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1334        same bug. Such groups have to be quantified to be useful, or contained
1335        inside another quantified group. (If there's no repetition, the reference
1336        can never match.) The problem arises because, having left the group and
1337        moved on to the rest of the pattern, a later failure that backtracks into
1338        the group uses the captured value from the final iteration of the group
1339        rather than the correct earlier one. I have fixed this in PCRE by forcing
1340        any group that contains a reference to itself to be an atomic group; that
1341        is, there cannot be any backtracking into it once it has completed. This is
1342        similar to recursive and subroutine calls.
1343    
1344    
1345    Version 8.00 19-Oct-09
1346    ----------------------
1347    
1348    1.  The table for translating pcre_compile() error codes into POSIX error codes
1349        was out-of-date, and there was no check on the pcre_compile() error code
1350        being within the table. This could lead to an OK return being given in
1351        error.
1352    
1353    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1354        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1355        in a Windows environment.
1356    
1357    3.  The pcregrep --count option prints the count for each file even when it is
1358        zero, as does GNU grep. However, pcregrep was also printing all files when
1359        --files-with-matches was added. Now, when both options are given, it prints
1360        counts only for those files that have at least one match. (GNU grep just
1361        prints the file name in this circumstance, but including the count seems
1362        more useful - otherwise, why use --count?) Also ensured that the
1363        combination -clh just lists non-zero counts, with no names.
1364    
1365    4.  The long form of the pcregrep -F option was incorrectly implemented as
1366        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1367        but it seems right to fix it, and I didn't think it was worth preserving
1368        the old behaviour.
1369    
1370    5.  The command line items --regex=pattern and --regexp=pattern were not
1371        recognized by pcregrep, which required --regex pattern or --regexp pattern
1372        (with a space rather than an '='). The man page documented the '=' forms,
1373        which are compatible with GNU grep; these now work.
1374    
1375    6.  No libpcreposix.pc file was created for pkg-config; there was just
1376        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1377    
1378    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1379        when UCP support is not needed, by modifying the Python script that
1380        generates it from Unicode data files. This should not matter if the module
1381        is correctly used as a library, but I received one complaint about 50K of
1382        unwanted data. My guess is that the person linked everything into his
1383        program rather than using a library. Anyway, it does no harm.
1384    
1385    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1386        was a minimum greater than 1 for a wide character in a possessive
1387        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1388        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1389        character. Chaos in the form of incorrect output or a compiling loop could
1390        result.
1391    
1392    9.  The restrictions on what a pattern can contain when partial matching is
1393        requested for pcre_exec() have been removed. All patterns can now be
1394        partially matched by this function. In addition, if there are at least two
1395        slots in the offset vector, the offset of the earliest inspected character
1396        for the match and the offset of the end of the subject are set in them when
1397        PCRE_ERROR_PARTIAL is returned.
1398    
1399    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1400        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1401        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1402        and may be more useful for multi-segment matching.
1403    
1404    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1405        used to be given if ever the end of the subject was reached; now it is
1406        given only if matching could not proceed because another character was
1407        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1408        string "Z", which now yields "no match" instead of "partial match". In the
1409        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1410        final character ended with (*FAIL).
1411    
1412    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1413        if the pattern had a "must contain" character that was already found in the
1414        earlier partial match, unless partial matching was again requested. For
1415        example, with the pattern /dog.(body)?/, the "must contain" character is
1416        "g". If the first part-match was for the string "dog", restarting with
1417        "sbody" failed. This bug has been fixed.
1418    
1419    13. The string returned by pcre_dfa_exec() after a partial match has been
1420        changed so that it starts at the first inspected character rather than the
1421        first character of the match. This makes a difference only if the pattern
1422        starts with a lookbehind assertion or \b or \B (\K is not supported by
1423        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1424        matching functions compatible, and I think it's the right thing to do.
1425    
1426    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1427        so that the demonstration program is easily available in environments where
1428        PCRE has not been installed from source.
1429    
1430    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1431        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1432        library.
1433    
1434    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1435        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1436        is not the first non-POSIX option to be added. Clearly some people find
1437        these options useful.
1438    
1439    17. If a caller to the POSIX matching function regexec() passes a non-zero
1440        value for nmatch with a NULL value for pmatch, the value of
1441        nmatch is forced to zero.
1442    
1443    18. RunGrepTest did not have a test for the availability of the -u option of
1444        the diff command, as RunTest does. It now checks in the same way as
1445        RunTest, and also checks for the -b option.
1446    
1447    19. If an odd number of negated classes containing just a single character
1448        interposed, within parentheses, between a forward reference to a named
1449        subpattern and the definition of the subpattern, compilation crashed with
1450        an internal error, complaining that it could not find the referenced
1451        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1452        [The bug was that it was starting one character too far in when skipping
1453        over the character class, thus treating the ] as data rather than
1454        terminating the class. This meant it could skip too much.]
1455    
1456    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1457        /g option in pcretest when the pattern contains \K, which makes it possible
1458        to have an empty string match not at the start, even when the pattern is
1459        anchored. Updated pcretest and pcredemo to use this option.
1460    
1461    21. If the maximum number of capturing subpatterns in a recursion was greater
1462        than the maximum at the outer level, the higher number was returned, but
1463        with unset values at the outer level. The correct (outer level) value is
1464        now given.
1465    
1466    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1467        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1468        make it do so. The string so far is captured, making this feature
1469        compatible with Perl.
1470    
1471    23. The tests have been re-organized, adding tests 11 and 12, to make it
1472        possible to check the Perl 5.10 features against Perl 5.10.
1473    
1474    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1475        pattern matches a fixed length string. PCRE did not allow this; now it
1476        does. Neither allows recursion.
1477    
1478    25. I finally figured out how to implement a request to provide the minimum
1479        length of subject string that was needed in order to match a given pattern.
1480        (It was back references and recursion that I had previously got hung up
1481        on.) This code has now been added to pcre_study(); it finds a lower bound
1482        to the length of subject needed. It is not necessarily the greatest lower
1483        bound, but using it to avoid searching strings that are too short does give
1484        some useful speed-ups. The value is available to calling programs via
1485        pcre_fullinfo().
1486    
1487    26. While implementing 25, I discovered to my embarrassment that pcretest had
1488        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1489        study optimizations had never been tested with that matching function.
1490        Oops. What is worse, even when it was passed study data, there was a bug in
1491        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1492        were also very few tests of studied patterns with pcre_dfa_exec().
1493    
1494    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1495        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1496        on the other side of the coin, they are no longer allowed to have different
1497        names, because these cannot be distinguished in PCRE, and this has caused
1498        confusion. (This is a difference from Perl.)
1499    
1500    28. When duplicate subpattern names are present (necessarily with different
1501        numbers, as required by 27 above), and a test is made by name in a
1502        conditional pattern, either for a subpattern having been matched, or for
1503        recursion in such a pattern, all the associated numbered subpatterns are
1504        tested, and the overall condition is true if the condition is true for any
1505        one of them. This is the way Perl works, and is also more like the way
1506        testing by number works.
1507    
1508    
1509    Version 7.9 11-Apr-09
1510    ---------------------
1511    
1512    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1513        (pcretest), all targets were linked against these libraries. This included
1514        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1515        libraries. This caused unwanted dependencies to be created. This problem
1516        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1517        pcretest is linked with readline.
1518    
1519    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1520        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1521        moved outside it again, because FALSE and TRUE are already defined in AIX,
1522        but BOOL is not.
1523    
1524    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1525        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1526    
1527    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1528        hyphens) following file names and line numbers when outputting matching
1529        lines. This is not true; no spaces are inserted. I have also clarified the
1530        wording for the --colour (or --color) option.
1531    
1532    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1533        was not coloured; this is different to GNU grep, so I have changed it to be
1534        the same.
1535    
1536    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1537        each matching line was coloured. Now it goes on to look for further matches
1538        of any of the test patterns, which is the same behaviour as GNU grep.
1539    
1540    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1541        doesn't make sense to accept an empty string match in pcregrep, so I have
1542        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1543        seems to be how GNU grep behaves.
1544    
1545    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1546        start or after a newline", because the conditional assertion was not being
1547        correctly handled. The rule now is that both the assertion and what follows
1548        in the first alternative must satisfy the test.
1549    
1550    9.  If auto-callout was enabled in a pattern with a conditional group whose
1551        condition was an assertion, PCRE could crash during matching, both with
1552        pcre_exec() and pcre_dfa_exec().
1553    
1554    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1555        used for matching.
1556    
1557    11. Unicode property support in character classes was not working for
1558        characters (bytes) greater than 127 when not in UTF-8 mode.
1559    
1560    12. Added the -M command line option to pcretest.
1561    
1562    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1563    
1564    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1565    
1566    16. Added comments and documentation about mis-use of no_arg in the C++
1567        wrapper.
1568    
1569    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1570        from Martin Jerabek that uses macro names for all relevant character and
1571        string constants.
1572    
1573    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1574        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1575        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1576        these, but not everybody uses configure.
1577    
1578    19. A conditional group that had only one branch was not being correctly
1579        recognized as an item that could match an empty string. This meant that an
1580        enclosing group might also not be so recognized, causing infinite looping
1581        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1582        with the subject "ab", where knowledge that the repeated group can match
1583        nothing is needed in order to break the loop.
1584    
1585    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1586        exec(), but without supplying a callout function, matching went wrong.
1587    
1588    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1589        leak if the size of the offset vector was greater than 30. When the vector
1590        is smaller, the saved offsets during recursion go onto a local stack
1591        vector, but for larger vectors malloc() is used. It was failing to free
1592        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1593        error, in fact).
1594    
1595    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1596        heapframe that is used only when UTF-8 support is enabled. This caused no
1597        problem, but was untidy.
1598    
1599    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1600        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1601        included within another project.
1602    
1603    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1604        slightly modified by me:
1605    
1606          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1607              not building pcregrep.
1608    
1609          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1610              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1611    
1612    25. Forward references, both numeric and by name, in patterns that made use of
1613        duplicate group numbers, could behave incorrectly or give incorrect errors,
1614        because when scanning forward to find the reference group, PCRE was not
1615        taking into account the duplicate group numbers. A pattern such as
1616        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1617    
1618    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1619        the feature of a custom pointer more persuasive (as requested by a user).
1620    
1621    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1622        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1623        pre-defined.
1624    
1625    28. Added support for (*UTF8) at the start of a pattern.
1626    
1627    29. Arrange for flags added by the "release type" setting in CMake to be shown
1628        in the configuration summary.
1629    
1630    
1631    Version 7.8 05-Sep-08
1632    ---------------------
1633    
1634    1.  Replaced UCP searching code with optimized version as implemented for Ad
1635        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1636        stage table and inline lookup instead of a function, giving speed ups of 2
1637        to 5 times on some simple patterns that I tested. Permission was given to
1638        distribute the MultiStage2.py script that generates the tables (it's not in
1639        the tarball, but is in the Subversion repository).
1640    
1641    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1642        scripts.
1643    
1644    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1645        a group with a zero qualifier. The result of the study could be incorrect,
1646        or the function might crash, depending on the pattern.
1647    
1648    4.  Caseless matching was not working for non-ASCII characters in back
1649        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1650        It now works when Unicode Property Support is available.
1651    
1652    5.  In pcretest, an escape such as \x{de} in the data was always generating
1653        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1654        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1655        truncation.
1656    
1657    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1658    
1659    7.  Added two (int) casts to pcregrep when printing the difference of two
1660        pointers, in case they are 64-bit values.
1661    
1662    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1663        test 2 if it fails.
1664    
1665    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1666        and a #define of that name to empty if it is not externally set. This is to
1667        allow users of MSVC to set it if necessary.
1668    
1669    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1670        the convenience functions in the pcre_get.c source file.
1671    
1672    11. An option change at the start of a pattern that had top-level alternatives
1673        could cause overwriting and/or a crash. This command provoked a crash in
1674        some environments:
1675    
1676          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1677    
1678        This potential security problem was recorded as CVE-2008-2371.
1679    
1680    12. For a pattern where the match had to start at the beginning or immediately
1681        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1682        pcre_dfa_exec() could read past the end of the passed subject if there was
1683        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1684        pcretest so that it places the subject at the end of its malloc-ed buffer.
1685    
1686    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1687        exec() might read past the end of the data buffer in UTF-8 mode.
1688    
1689    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1690        the data contained the byte 0x85 as part of a UTF-8 character within its
1691        first line. This applied both to normal and DFA matching.
1692    
1693    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1694        /^[^d]*?$/8 failed to match "abc".
1695    
1696    16. Added a missing copyright notice to pcrecpp_internal.h.
1697    
1698    17. Make it more clear in the documentation that values returned from
1699        pcre_exec() in ovector are byte offsets, not character counts.
1700    
1701    18. Tidied a few places to stop certain compilers from issuing warnings.
1702    
1703    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1704        supplied by Stefan Weber. I made a further small update for 7.8 because
1705        there is a change of source arrangements: the pcre_searchfuncs.c module is
1706        replaced by pcre_ucd.c.
1707    
1708    
1709    Version 7.7 07-May-08
1710  ---------------------  ---------------------
1711    
1712  1.  Applied Craig's patch to sort out a long long problem: "If we can't convert  1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1713      a string to a long long, pretend we don't even have a long long." This is      a string to a long long, pretend we don't even have a long long." This is
1714      done by checking for the strtoq, strtoll, and _strtoi64 functions.      done by checking for the strtoq, strtoll, and _strtoi64 functions.
1715    
1716  2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with  2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1717      pre-7.6 versions, which defined a global no_arg variable instead of putting      pre-7.6 versions, which defined a global no_arg variable instead of putting
1718      it in the RE class.      it in the RE class. (See also #8 below.)
1719    
1720  3.  Remove a line of dead code, identified by coverity and reported by Nuno  3.  Remove a line of dead code, identified by coverity and reported by Nuno
1721      Lopes.      Lopes.
1722    
1723  4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:  4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1724    
1725      (1) The include/exclude patterns were being applied to the whole pathnames      (1) The include/exclude patterns were being applied to the whole pathnames
1726          of files, instead of just to the final components.          of files, instead of just to the final components.
1727    
1728      (2) If there was more than one level of directory, the subdirectories were      (2) If there was more than one level of directory, the subdirectories were
1729          skipped unless they satisfied the include/exclude conditions. This is          skipped unless they satisfied the include/exclude conditions. This is
1730          inconsistent with GNU grep (and could even be seen as contrary to the          inconsistent with GNU grep (and could even be seen as contrary to the
1731          pcregrep specification - which I improved to make it absolutely clear).          pcregrep specification - which I improved to make it absolutely clear).
1732          The action now is always to scan all levels of directory, and just          The action now is always to scan all levels of directory, and just
1733          apply the include/exclude patterns to regular files.          apply the include/exclude patterns to regular files.
1734    
1735  5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used  5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1736      --exclude_dir in the tests to avoid scanning .svn directories.      --exclude_dir in the tests to avoid scanning .svn directories.
1737    
1738    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1739        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1740        doesn't support NULs in patterns.
1741    
1742    7.  Added some missing "const"s to declarations of static tables in
1743        pcre_compile.c and pcre_dfa_exec.c.
1744    
1745    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1746        caused by fix #2  above. (Subsequently also a second patch to fix the
1747        first patch. And a third patch - this was a messy problem.)
1748    
1749    9.  Applied Craig's patch to remove the use of push_back().
1750    
1751    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1752        matching function regexec().
1753    
1754    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1755        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1756        references. PCRE supports relative numbers with this syntax (I don't think
1757        Oniguruma does).
1758    
1759    12. Previously, a group with a zero repeat such as (...){0} was completely
1760        omitted from the compiled regex. However, this means that if the group
1761        was called as a subroutine from elsewhere in the pattern, things went wrong
1762        (an internal error was given). Such groups are now left in the compiled
1763        pattern, with a new opcode that causes them to be skipped at execution
1764        time.
1765    
1766    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1767        to the way PCRE behaves:
1768    
1769        (a) A lone ] character is dis-allowed (Perl treats it as data).
1770    
1771        (b) A back reference to an unmatched subpattern matches an empty string
1772            (Perl fails the current match path).
1773    
1774        (c) A data ] in a character class must be notated as \] because if the
1775            first data character in a class is ], it defines an empty class. (In
1776            Perl it is not possible to have an empty class.) The empty class []
1777            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1778            The negative empty class [^] matches any one character, independently
1779            of the DOTALL setting.
1780    
1781    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1782        non-existent subpattern following a character class starting with ']' and
1783        containing () gave an internal compiling error instead of "reference to
1784        non-existent subpattern". Fortunately, when the pattern did exist, the
1785        compiled code was correct. (When scanning forwards to check for the
1786        existencd of the subpattern, it was treating the data ']' as terminating
1787        the class, so got the count wrong. When actually compiling, the reference
1788        was subsequently set up correctly.)
1789    
1790    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1791        it was being rejected as not supported by pcre_dfa_exec(), even though
1792        other assertions are supported. I have made pcre_dfa_exec() support
1793        (*FAIL).
1794    
1795    16. The implementation of 13c above involved the invention of a new opcode,
1796        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1797        cannot be changed at match time, I realized I could make a small
1798        improvement to matching performance by compiling OP_ALLANY instead of
1799        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1800        on the OP_ANY path.
1801    
1802    17. Compiling pcretest on Windows with readline support failed without the
1803        following two fixes: (1) Make the unistd.h include conditional on
1804        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1805    
1806    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1807        ncurses library to be included for pcretest when ReadLine support is
1808        requested, but also to allow for it to be overridden. This patch came from
1809        Daniel Bergström.
1810    
1811    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1812        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1813        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1814        spotting this.
1815    
1816    
1817  Version 7.6 28-Jan-08  Version 7.6 28-Jan-08

Legend:
Removed from v.325  
changed lines
  Added in v.1240

  ViewVC Help
Powered by ViewVC 1.1.5