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

Diff of /code/trunk/ChangeLog

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

revision 131 by ph10, Mon Mar 26 15:10:12 2007 UTC revision 1190 by ph10, Tue Oct 30 16:49:19 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 12-Mar-07  Version 8.32
5    ------------
6    
7    1.  Improved JIT compiler optimizations for first character search and single
8        character iterators.
9    
10    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
11        Patch by Daniel Richard G.
12    
13    3.  Single character iterator optimizations in the JIT compiler.
14    
15    4.  Improved JIT compiler optimizations for character ranges.
16    
17    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
18        Reported by Giuseppe D'Angelo.
19    
20    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
21        of a line, was being set incorrectly in cases where .* appeared inside
22        atomic brackets at the start of a pattern, or where there was a subsequent
23        *PRUNE or *SKIP.
24    
25    7.  Improved instruction cache flush for POWER/PowerPC.
26        Patch by Daniel Richard G.
27    
28    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
29        grep:
30    
31        (a) There is now no limit to the number of patterns to be matched.
32    
33        (b) An error is given if a pattern is too long.
34    
35        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
36            are now supported.
37    
38        (d) --exclude-from and --include-from (multiple use) have been added.
39    
40        (e) Exclusions and inclusions now apply to all files and directories, not
41            just to those obtained from scanning a directory recursively.
42    
43        (f) Multiple uses of -f and --file-list are now supported.
44    
45        (g) In a Windows environment, the default for -d has been changed from
46            "read" (the GNU grep default) to "skip", because otherwise the presence
47            of a directory in the file list provokes an error.
48    
49        (h) The documentation has been revised and clarified in places.
50    
51    9.  Improve the matching speed of capturing brackets.
52    
53    10. Changed the meaning of \X so that it now matches a Unicode extended
54        grapheme cluster.
55    
56    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
57        out POSIX threads when JIT support is configured.
58    
59    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
60    
61    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
62        structure could go wrong in environments where size_t is not the same size
63        as int.
64    
65    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
66        set.
67    
68    15. The EBCDIC support had decayed; later updates to the code had included
69        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
70        general tidy up of EBCDIC-related issues, and the documentation was also
71        not quite right. There is now a test that can be run on ASCII systems to
72        check some of the EBCDIC-related things (but is it not a full test).
73    
74    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
75        in a small tidy to the code.
76    
77    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
78    
79    18. If the --only-matching (-o) option in pcregrep is specified multiple
80        times, each one causes appropriate output. For example, -o1 -o2 outputs the
81        substrings matched by the 1st and 2nd capturing parentheses. A separating
82        string can be specified by --om-separator (default empty).
83    
84    19. Improving the first n character searches.
85    
86    20. Turn case lists for horizontal and vertical white space into macros so that
87        they are defined only once.
88    
89    21. This set of changes together give more compatible Unicode case-folding
90        behaviour for characters that have more than one other case when UCP
91        support is available.
92    
93        (a) The Unicode property table now has offsets into a new table of sets of
94            three or more characters that are case-equivalent. The MultiStage2.py
95            script that generates these tables (the pcre_ucd.c file) now scans
96            CaseFolding.txt instead of UnicodeData.txt for character case
97            information.
98    
99        (b) The code for adding characters or ranges of characters to a character
100            class has been abstracted into a generalized function that also handles
101            case-independence. In UTF-mode with UCP support, this uses the new data
102            to handle characters with more than one other case.
103    
104        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
105            whose other case is greater than 256 are now correctly matched
106            caselessly. Previously, the high codepoint matched the low one, but not
107            vice versa.
108    
109        (d) The processing of \h, \H, \v, and \ in character classes now makes use
110            of the new class addition function, using character lists defined as
111            macros alongside the case definitions of 20 above.
112    
113        (e) Caseless back references now work with characters that have more than
114            one other case.
115    
116        (f) General caseless matching of characters with more than one other case
117            is supported.
118    
119    22. Unicode character properties were updated from Unicode 6.2.0
120    
121    23. Improved CMake support under Windows. Patch by Daniel Richard G.
122    
123    24. Add support for 32-bit character strings, and UTF-32
124    
125    25. Major JIT compiler update (code refactoring and bugfixing).
126        Experimental Sparc 32 support is added.
127    
128    26. Applied a modified version of Daniel Richard G's patch to create
129        pcre.h.generic and config.h.generic by "make" instead of in the
130        PrepareRelease script.
131    
132    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
133        pcre_compile.c when checking for a zero character.
134    
135    28. Introducing a native interface for JIT. Through this interface, the compiled
136        machine code can be directly executed. The purpose of this interface is to
137        provide fast pattern matching, so several sanity checks are not performed.
138        However, feature tests are still performed. The new interface provides
139        1.4x speedup compared to the old one.
140    
141    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
142        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
143        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
144    
145    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
146        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
147        the "old" RFC 2279). Instead, it ended up passing a negative length to
148        pcre_exec().
149    
150    
151    Version 8.31 06-July-2012
152    -------------------------
153    
154    1.  Fixing a wrong JIT test case and some compiler warnings.
155    
156    2.  Removed a bashism from the RunTest script.
157    
158    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
159        to unsigned type, result still unsigned" that was given by an MS compiler
160        on encountering the code "-sizeof(xxx)".
161    
162    4.  Partial matching support is added to the JIT compiler.
163    
164    5.  Fixed several bugs concerned with partial matching of items that consist
165        of more than one character:
166    
167        (a) /^(..)\1/ did not partially match "aba" because checking references was
168            done on an "all or nothing" basis. This also applied to repeated
169            references.
170    
171        (b) \R did not give a hard partial match if \r was found at the end of the
172            subject.
173    
174        (c) \X did not give a hard partial match after matching one or more
175            characters at the end of the subject.
176    
177        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
178            a partial match for the string "\r".
179    
180        (e) When newline was set to CRLF, the metacharacter "." did not recognize
181            a partial match for a CR character at the end of the subject string.
182    
183    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
184        running pcretest, the text "(JIT)" added to the output whenever JIT is
185        actually used to run the match.
186    
187    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
188        or /S+[+] with a digit between 1 and 7.
189    
190    8.  OP_NOT now supports any UTF character not just single-byte ones.
191    
192    9.  (*MARK) control verb is now supported by the JIT compiler.
193    
194    10. The command "./RunTest list" lists the available tests without actually
195        running any of them. (Because I keep forgetting what they all are.)
196    
197    11. Add PCRE_INFO_MAXLOOKBEHIND.
198    
199    12. Applied a (slightly modified) user-supplied patch that improves performance
200        when the heap is used for recursion (compiled with --disable-stack-for-
201        recursion). Instead of malloc and free for each heap frame each time a
202        logical recursion happens, frames are retained on a chain and re-used where
203        possible. This sometimes gives as much as 30% improvement.
204    
205    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
206        call.
207    
208    14. As documented, (*COMMIT) is now confined to within a positive assertion.
209    
210    15. It is now possible to link pcretest with libedit as an alternative to
211        libreadline.
212    
213    16. (*COMMIT) control verb is now supported by the JIT compiler.
214    
215    17. The Unicode data tables have been updated to Unicode 6.1.0.
216    
217    18. Added --file-list option to pcregrep.
218    
219    19. Added binary file support to pcregrep, including the -a, --binary-files,
220        -I, and --text options.
221    
222    20. The madvise function is renamed for posix_madvise for QNX compatibility
223        reasons. Fixed by Giuseppe D'Angelo.
224    
225    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
226    
227    22. Changed the option for creating symbolic links for 16-bit man pages from
228        -s to -sf so that re-installing does not cause issues.
229    
230    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
231    
232    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
233        matches in certain environments (the workspace was not being correctly
234        retained). Also added to pcre_dfa_exec() a simple plausibility check on
235        some of the workspace data at the beginning of a restart.
236    
237    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
238        was not doing so when it should - probably a typo introduced by SVN 528
239        (change 8.10/14).
240    
241    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
242        \w+ when the character tables indicated that \x{c4} was a word character.
243        There were several related cases, all because the tests for doing a table
244        lookup were testing for characters less than 127 instead of 255.
245    
246    27. If a pattern contains capturing parentheses that are not used in a match,
247        their slots in the ovector are set to -1. For those that are higher than
248        any matched groups, this happens at the end of processing. In the case when
249        there were back references that the ovector was too small to contain
250        (causing temporary malloc'd memory to be used during matching), and the
251        highest capturing number was not used, memory off the end of the ovector
252        was incorrectly being set to -1. (It was using the size of the temporary
253        memory instead of the true size.)
254    
255    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
256        ovector size, it uses memory at the end of the block that it has got.
257    
258    29. Check for an overlong MARK name and give an error at compile time. The
259        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
260    
261    30. JIT compiler update.
262    
263    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
264        Rill for the patch.
265    
266    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
267    
268    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
269    
270    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
271        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
272        was enabled.
273    
274    35. Improve JIT code generation for greedy plus quantifier.
275    
276    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
277        1 to "aa" instead of to an empty string. The bug affected repeated groups
278        that could potentially match an empty string.
279    
280    37. Optimizing single character iterators in JIT.
281    
282    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
283        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
284        codepoints that are too big for the mode are faulted, and in a UTF mode,
285        disallowed codepoints are also faulted.
286    
287    39. If PCRE was compiled with UTF support, in three places in the DFA
288        matcher there was code that should only have been obeyed in UTF mode, but
289        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
290        processing when bytes with values greater than 127 were present. In 16-bit
291        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
292        both cases the values are those that cannot be the first data item in a UTF
293        character. The three items that might have provoked this were recursions,
294        possessively repeated groups, and atomic groups.
295    
296    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
297        and pcregrep, because some OS require shared objects to be explicitly
298        passed to ld, causing the link step to fail if they are not.
299    
300    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
301        mode, patterns that started with \h* or \R* might be incorrectly matched.
302    
303    
304    Version 8.30 04-February-2012
305    -----------------------------
306    
307    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
308        name is defined in ctype.h.
309    
310    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
311        only in quite long subpatterns.
312    
313    3.  Removed the function pcre_info(), which has been obsolete and deprecated
314        since it was replaced by pcre_fullinfo() in February 2000.
315    
316    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
317        match a (*MARK), and the match failed at the start of the subject, a
318        reference to memory before the start of the subject could occur. This bug
319        was introduced by fix 17 of release 8.21.
320    
321    5.  A reference to an unset group with zero minimum repetition was giving
322        totally wrong answers (in non-JavaScript-compatibility mode). For example,
323        /(another)?(\1?)test/ matched against "hello world test". This bug was
324        introduced in release 8.13.
325    
326    6.  Add support for 16-bit character strings (a large amount of work involving
327        many changes and refactorings).
328    
329    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
330        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
331        from a file.
332    
333    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
334        rounding is not applied in this particular case).
335    
336    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
337        if they appear, or are escaped, in patterns.
338    
339    10. Get rid of a number of -Wunused-but-set-variable warnings.
340    
341    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
342        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
343        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
344        also returns the mark "x". This bug applied to capturing parentheses,
345        non-capturing parentheses, and atomic parentheses. It also applied to some
346        assertions.
347    
348    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
349        information out of configure.ac instead of relying on pcre.h.generic, which
350        is not stored in the repository.
351    
352    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
353        -lreadline.
354    
355    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
356    
357    15. Applied Graycode's patch to put the top-level frame on the stack rather
358        than the heap when not using the stack for recursion. This gives a
359        performance improvement in many cases when recursion is not deep.
360    
361    16. Experimental code added to "pcretest -C" to output the stack frame size.
362    
363    
364    Version 8.21 12-Dec-2011
365    ------------------------
366    
367    1.  Updating the JIT compiler.
368    
369    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
370        are added as well.
371    
372    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
373        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
374        calling _pcre_jit_exec. Some extra comments are added.
375    
376    4.  (*MARK) settings inside atomic groups that do not contain any capturing
377        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
378        was introduced by change 18 for 8.20.
379    
380    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
381        ECMA-262 standard.
382    
383    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
384        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
385        This bug was probably introduced by change 9 of 8.13.
386    
387    7.  While fixing 6 above, I noticed that a number of other items were being
388        incorrectly rejected as "not fixed length". This arose partly because newer
389        opcodes had not been added to the fixed-length checking code. I have (a)
390        corrected the bug and added tests for these items, and (b) arranged for an
391        error to occur if an unknown opcode is encountered while checking for fixed
392        length instead of just assuming "not fixed length". The items that were
393        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
394        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
395        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
396    
397    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
398        being incorrectly compiled and would have given unpredicatble results.
399    
400    9.  A possessively repeated subpattern with minimum repeat count greater than
401        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
402        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
403        the first (A) could occur when it should not.
404    
405    10. Add a cast and remove a redundant test from the code.
406    
407    11. JIT should use pcre_malloc/pcre_free for allocation.
408    
409    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
410        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
411        is anything other than /usr, -L is still shown).
412    
413    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
414    
415    14. Perl does not support \N without a following name in a [] class; PCRE now
416        also gives an error.
417    
418    15. If a forward reference was repeated with an upper limit of around 2000,
419        it caused the error "internal error: overran compiling workspace". The
420        maximum number of forward references (including repeats) was limited by the
421        internal workspace, and dependent on the LINK_SIZE. The code has been
422        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
423        the default depends on LINK_SIZE. There is a new upper limit (for safety)
424        of around 200,000 forward references. While doing this, I also speeded up
425        the filling in of repeated forward references.
426    
427    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
428        incorrectly expecting the subject to contain another "a" after the start.
429    
430    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
431        in the match, the SKIP should be ignored. This was not happening; instead
432        the SKIP was being treated as NOMATCH. For patterns such as
433        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
434        tested.
435    
436    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
437        now much more compatible with Perl, in particular in cases where the result
438        is a non-match for a non-anchored pattern. For example, if
439        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
440        "m", where previously it did not return a name. A side effect of this
441        change is that for partial matches, the last encountered mark name is
442        returned, as for non matches. A number of tests that were previously not
443        Perl-compatible have been moved into the Perl-compatible test files. The
444        refactoring has had the pleasing side effect of removing one argument from
445        the match() function, thus reducing its stack requirements.
446    
447    19. If the /S+ option was used in pcretest to study a pattern using JIT,
448        subsequent uses of /S (without +) incorrectly behaved like /S+.
449    
450    21. Retrieve executable code size support for the JIT compiler and fixing
451        some warnings.
452    
453    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
454        not work when the shorter character appeared right at the end of the
455        subject string.
456    
457    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
458        systems.
459    
460    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
461        output it when the /M option is used in pcretest.
462    
463    25. The CheckMan script was not being included in the distribution. Also, added
464        an explicit "perl" to run Perl scripts from the PrepareRelease script
465        because this is reportedly needed in Windows.
466    
467    26. If study data was being save in a file and studying had not found a set of
468        "starts with" bytes for the pattern, the data written to the file (though
469        never used) was taken from uninitialized memory and so caused valgrind to
470        complain.
471    
472    27. Updated RunTest.bat as provided by Sheri Pierce.
473    
474    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
475    
476    29. Computation of memory usage for the table of capturing group names was
477        giving an unnecessarily large value.
478    
479    
480    Version 8.20 21-Oct-2011
481    ------------------------
482    
483    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
484        a POSIX class. After further experiments with Perl, which convinced me that
485        Perl has bugs and confusions, a closing square bracket is no longer allowed
486        in a POSIX name. This bug also affected patterns with classes that started
487        with full stops.
488    
489    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
490        captured substring, but while checking the failing first alternative,
491        substring 1 is temporarily captured. If the output vector supplied to
492        pcre_exec() was not big enough for this capture, the yield of the function
493        was still zero ("insufficient space for captured substrings"). This cannot
494        be totally fixed without adding another stack variable, which seems a lot
495        of expense for a edge case. However, I have improved the situation in cases
496        such as /(a)(b)x|abc/ matched against "abc", where the return code
497        indicates that fewer than the maximum number of slots in the ovector have
498        been set.
499    
500    3.  Related to (2) above: when there are more back references in a pattern than
501        slots in the output vector, pcre_exec() uses temporary memory during
502        matching, and copies in the captures as far as possible afterwards. It was
503        using the entire output vector, but this conflicts with the specification
504        that only 2/3 is used for passing back captured substrings. Now it uses
505        only the first 2/3, for compatibility. This is, of course, another edge
506        case.
507    
508    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
509        main code base, and can be used by building with --enable-jit. When this is
510        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
511        runtime --no-jit option is given.
512    
513    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
514        ovector, the return from the function was zero, implying that there were
515        other matches that did not fit. The correct "exactly full" value is now
516        returned.
517    
518    6.  If a subpattern that was called recursively or as a subroutine contained
519        (*PRUNE) or any other control that caused it to give a non-standard return,
520        invalid errors such as "Error -26 (nested recursion at the same subject
521        position)" or even infinite loops could occur.
522    
523    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
524        computing the minimum length on reaching *ACCEPT, and so ended up with the
525        wrong value of 1 rather than 0. Further investigation indicates that
526        computing a minimum subject length in the presence of *ACCEPT is difficult
527        (think back references, subroutine calls), and so I have changed the code
528        so that no minimum is registered for a pattern that contains *ACCEPT.
529    
530    8.  If (*THEN) was present in the first (true) branch of a conditional group,
531        it was not handled as intended. [But see 16 below.]
532    
533    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
534        Sheri Pierce.
535    
536    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
537        the first byte in a match must be "a".
538    
539    11. Change 17 for 8.13 increased the recursion depth for patterns like
540        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
541        pattern contains any instances of (*THEN). If it does not, the old
542        optimizations are restored. It would be nice to do this on a per-group
543        basis, but at the moment that is not feasible.
544    
545    12. In some environments, the output of pcretest -C is CRLF terminated. This
546        broke RunTest's code that checks for the link size. A single white space
547        character after the value is now allowed for.
548    
549    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
550        For "fr", it uses the Windows-specific input and output files.
551    
552    14. If (*THEN) appeared in a group that was called recursively or as a
553        subroutine, it did not work as intended. [But see next item.]
554    
555    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
556        pattern fragments (but not containing any | characters). If A and B are
557        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
558        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
559        D. In other words, Perl considers parentheses that do not contain any |
560        characters to be part of a surrounding alternative, whereas PCRE was
561        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
562        differently. PCRE now behaves in the same way as Perl, except in the case
563        of subroutine/recursion calls such as (?1) which have in any case always
564        been different (but PCRE had them first :-).
565    
566    16. Related to 15 above: Perl does not treat the | in a conditional group as
567        creating alternatives. Such a group is treated in the same way as an
568        ordinary group without any | characters when processing (*THEN). PCRE has
569        been changed to match Perl's behaviour.
570    
571    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
572        RunGrepTest script failed.
573    
574    18. Change 22 for version 13 caused atomic groups to use more stack. This is
575        inevitable for groups that contain captures, but it can lead to a lot of
576        stack use in large patterns. The old behaviour has been restored for atomic
577        groups that do not contain any capturing parentheses.
578    
579    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
580        suppress the check for a minimum subject length at run time. (If it was
581        given to pcre_exec() or pcre_dfa_exec() it did work.)
582    
583    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
584        fail to work when decoding hex characters in data strings in EBCDIC
585        environments.
586    
587    21. It appears that in at least one Mac OS environment, the isxdigit() function
588        is implemented as a macro that evaluates to its argument more than once,
589        contravening the C 90 Standard (I haven't checked a later standard). There
590        was an instance in pcretest which caused it to go wrong when processing
591        \x{...} escapes in subject strings. The has been rewritten to avoid using
592        things like p++ in the argument of isxdigit().
593    
594    
595    Version 8.13 16-Aug-2011
596    ------------------------
597    
598    1.  The Unicode data tables have been updated to Unicode 6.0.0.
599    
600    2.  Two minor typos in pcre_internal.h have been fixed.
601    
602    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
603        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
604        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
605    
606    4.  There were a number of related bugs in the code for matching backrefences
607        caselessly in UTF-8 mode when codes for the characters concerned were
608        different numbers of bytes. For example, U+023A and U+2C65 are an upper
609        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
610        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
611        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
612        2-byte code at the end of the subject (it thought there wasn't enough data
613        left).
614    
615    5.  Comprehensive information about what went wrong is now returned by
616        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
617        as the output vector has at least 2 elements. The offset of the start of
618        the failing character and a reason code are placed in the vector.
619    
620    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
621        now returned is for the first byte of the failing character, instead of the
622        last byte inspected. This is an incompatible change, but I hope it is small
623        enough not to be a problem. It makes the returned offset consistent with
624        pcre_exec() and pcre_dfa_exec().
625    
626    7.  pcretest now gives a text phrase as well as the error number when
627        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
628        failure, the offset and reason code are output.
629    
630    8.  When \R was used with a maximizing quantifier it failed to skip backwards
631        over a \r\n pair if the subsequent match failed. Instead, it just skipped
632        back over a single character (\n). This seems wrong (because it treated the
633        two characters as a single entity when going forwards), conflicts with the
634        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
635        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
636        has been changed.
637    
638    9.  Some internal refactoring has changed the processing so that the handling
639        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
640        time (the PCRE_DOTALL option was changed this way some time ago: version
641        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
642        which was always a bit of a fudge. It also means that there is one less
643        argument for the match() function, which reduces its stack requirements
644        slightly. This change also fixes an incompatibility with Perl: the pattern
645        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
646    
647    10. More internal refactoring has drastically reduced the number of recursive
648        calls to match() for possessively repeated groups such as (abc)++ when
649        using pcre_exec().
650    
651    11. While implementing 10, a number of bugs in the handling of groups were
652        discovered and fixed:
653    
654        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
655        (a|)*(?1) gave a compile-time internal error.
656        ((a|)+)+  did not notice that the outer group could match an empty string.
657        (^a|^)+   was not marked as anchored.
658        (.*a|.*)+ was not marked as matching at start or after a newline.
659    
660    12. Yet more internal refactoring has removed another argument from the match()
661        function. Special calls to this function are now indicated by setting a
662        value in a variable in the "match data" data block.
663    
664    13. Be more explicit in pcre_study() instead of relying on "default" for
665        opcodes that mean there is no starting character; this means that when new
666        ones are added and accidentally left out of pcre_study(), testing should
667        pick them up.
668    
669    14. The -s option of pcretest has been documented for ages as being an old
670        synonym of -m (show memory usage). I have changed it to mean "force study
671        for every regex", that is, assume /S for every regex. This is similar to -i
672        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
673        using it. It makes it easier to run collections of tests with and without
674        study enabled, and thereby test pcre_study() more easily. All the standard
675        tests are now run with and without -s (but some patterns can be marked as
676        "never study" - see 20 below).
677    
678    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
679        restoration of the capturing data to the outer values was not happening
680        correctly.
681    
682    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
683        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
684        pattern had matched an empty string, and so incorrectly returned a no
685        match.
686    
687    17. There was optimizing code for the last branch of non-capturing parentheses,
688        and also for the obeyed branch of a conditional subexpression, which used
689        tail recursion to cut down on stack usage. Unfortunately, now that there is
690        the possibility of (*THEN) occurring in these branches, tail recursion is
691        no longer possible because the return has to be checked for (*THEN). These
692        two optimizations have therefore been removed. [But see 8.20/11 above.]
693    
694    18. If a pattern containing \R was studied, it was assumed that \R always
695        matched two bytes, thus causing the minimum subject length to be
696        incorrectly computed because \R can also match just one byte.
697    
698    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
699        was incorrectly computed.
700    
701    20. If /S is present twice on a test pattern in pcretest input, it now
702        *disables* studying, thereby overriding the use of -s on the command line
703        (see 14 above). This is necessary for one or two tests to keep the output
704        identical in both cases.
705    
706    21. When (*ACCEPT) was used in an assertion that matched an empty string and
707        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
708    
709    22. When an atomic group that contained a capturing parenthesis was
710        successfully matched, but the branch in which it appeared failed, the
711        capturing was not being forgotten if a higher numbered group was later
712        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
713        group 1 to "a", when in fact it should be unset. This applied to multi-
714        branched capturing and non-capturing groups, repeated or not, and also to
715        positive assertions (capturing in negative assertions does not happen
716        in PCRE) and also to nested atomic groups.
717    
718    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
719        subject after a captured substring, to make it easier to tell which of a
720        number of identical substrings has been captured.
721    
722    24. The way atomic groups are processed by pcre_exec() has been changed so that
723        if they are repeated, backtracking one repetition now resets captured
724        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
725        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
726        "aaa". Previously, it would have been "a". As part of this code
727        refactoring, the way recursive calls are handled has also been changed.
728    
729    25. If an assertion condition captured any substrings, they were not passed
730        back unless some other capturing happened later. For example, if
731        (?(?=(a))a) was matched against "a", no capturing was returned.
732    
733    26. When studying a pattern that contained subroutine calls or assertions,
734        the code for finding the minimum length of a possible match was handling
735        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
736        group 1 called group 2 while simultaneously a separate group 2 called group
737        1). A stack overflow occurred in this case. I have fixed this by limiting
738        the recursion depth to 10.
739    
740    27. Updated RunTest.bat in the distribution to the version supplied by Tom
741        Fortmann. This supports explicit test numbers on the command line, and has
742        argument validation and error reporting.
743    
744    28. An instance of \X with an unlimited repeat could fail if at any point the
745        first character it looked at was a mark character.
746    
747    29. Some minor code refactoring concerning Unicode properties and scripts
748        should reduce the stack requirement of match() slightly.
749    
750    30. Added the '=' option to pcretest to check the setting of unused capturing
751        slots at the end of the pattern, which are documented as being -1, but are
752        not included in the return count.
753    
754    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
755        compiled something random. Now it gives a compile-time error (as does
756        Perl).
757    
758    32. A *MARK encountered during the processing of a positive assertion is now
759        recorded and passed back (compatible with Perl).
760    
761    33. If --only-matching or --colour was set on a pcregrep call whose pattern
762        had alternative anchored branches, the search for a second match in a line
763        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
764        matched the line "0102" twice. The same bug affected patterns that started
765        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
766        twice.
767    
768    34. Previously, PCRE did not allow quantification of assertions. However, Perl
769        does, and because of capturing effects, quantifying parenthesized
770        assertions may at times be useful. Quantifiers are now allowed for
771        parenthesized assertions.
772    
773    35. A minor code tidy in pcre_compile() when checking options for \R usage.
774    
775    36. \g was being checked for fancy things in a character class, when it should
776        just be a literal "g".
777    
778    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
779        appearance of a nested POSIX class supersedes an apparent external class.
780        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
781        unescaped square brackets may also appear as part of class names. For
782        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
783        more like Perl. (But see 8.20/1 above.)
784    
785    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
786        was because it thought it was \N{name}, which is not supported).
787    
788    39. Add minix to OS list not supporting the -S option in pcretest.
789    
790    40. PCRE tries to detect cases of infinite recursion at compile time, but it
791        cannot analyze patterns in sufficient detail to catch mutual recursions
792        such as ((?1))((?2)). There is now a runtime test that gives an error if a
793        subgroup is called recursively as a subpattern for a second time at the
794        same position in the subject string. In previous releases this might have
795        been caught by the recursion limit, or it might have run out of stack.
796    
797    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
798        happen only once. PCRE was, however incorrectly giving a compile time error
799        "recursive call could loop indefinitely" because it cannot analyze the
800        pattern in sufficient detail. The compile time test no longer happens when
801        PCRE is compiling a conditional subpattern, but actual runaway loops are
802        now caught at runtime (see 40 above).
803    
804    42. It seems that Perl allows any characters other than a closing parenthesis
805        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
806        has been changed to be the same.
807    
808    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
809        as not to get warnings when autogen.sh is called. Also changed
810        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
811    
812    44. To help people who use pcregrep to scan files containing exceedingly long
813        lines, the following changes have been made:
814    
815        (a) The default value of the buffer size parameter has been increased from
816            8K to 20K. (The actual buffer used is three times this size.)
817    
818        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
819            PCRE is built.
820    
821        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
822            to be set at run time.
823    
824        (d) Numerical values in pcregrep options can be followed by K or M, for
825            example --buffer-size=50K.
826    
827        (e) If a line being scanned overflows pcregrep's buffer, an error is now
828            given and the return code is set to 2.
829    
830    45. Add a pointer to the latest mark to the callout data block.
831    
832    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
833        partial match of an empty string instead of no match. This was specific to
834        the use of ".".
835    
836    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
837        complete match instead of a partial match. This bug was dependent on both
838        the PCRE_UTF8 and PCRE_DOTALL options being set.
839    
840    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
841        starting byte set, because \b was not being ignored.
842    
843    
844    Version 8.12 15-Jan-2011
845    ------------------------
846    
847    1.  Fixed some typos in the markup of the man pages, and wrote a script that
848        checks for such things as part of the documentation building process.
849    
850    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
851        --match-limit and --recursion-limit options (added for 8.11). In
852        particular, this made one of the standard tests fail. (The integer value
853        went into the wrong half of a long int.)
854    
855    3.  If the --colour option was given to pcregrep with -v (invert match), it
856        did strange things, either producing crazy output, or crashing. It should,
857        of course, ignore a request for colour when reporting lines that do not
858        match.
859    
860    4.  Another pcregrep bug caused similar problems if --colour was specified with
861        -M (multiline) and the pattern match finished with a line ending.
862    
863    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
864        matched in multiline mode, the following line was shown as part of the
865        match. This seems wrong, so I have changed it.
866    
867    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
868        the check for further matches in the same line (so they could be coloured)
869        to overrun the end of the current line. If another match was found, it was
870        incorrectly shown (and then shown again when found in the next line).
871    
872    7.  If pcregrep was compiled under Windows, there was a reference to the
873        function pcregrep_exit() before it was defined. I am assuming this was
874        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
875        reported by a user. I've moved the definition above the reference.
876    
877    
878    Version 8.11 10-Dec-2010
879    ------------------------
880    
881    1.  (*THEN) was not working properly if there were untried alternatives prior
882        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
883        backtracked to try for "b" instead of moving to the next alternative branch
884        at the same level (in this case, to look for "c"). The Perl documentation
885        is clear that when (*THEN) is backtracked onto, it goes to the "next
886        alternative in the innermost enclosing group".
887    
888    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
889        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
890        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
891        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
892        (*THEN).
893    
894    3.  If \s appeared in a character class, it removed the VT character from
895        the class, even if it had been included by some previous item, for example
896        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
897        of \s, but is part of the POSIX "space" class.)
898    
899    4.  A partial match never returns an empty string (because you can always
900        match an empty string at the end of the subject); however the checking for
901        an empty string was starting at the "start of match" point. This has been
902        changed to the "earliest inspected character" point, because the returned
903        data for a partial match starts at this character. This means that, for
904        example, /(?<=abc)def/ gives a partial match for the subject "abc"
905        (previously it gave "no match").
906    
907    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
908        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
909        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
910        has an implication that the given string is incomplete (because a partial
911        match is preferred over a full match). For this reason, these items now
912        give a partial match in this situation. [Aside: previously, the one case
913        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
914        match rather than a full match, which was wrong by the old rules, but is
915        now correct.]
916    
917    6.  There was a bug in the handling of #-introduced comments, recognized when
918        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
919        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
920        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
921        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
922        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
923        places in pcre_compile().
924    
925    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
926        comments when looking ahead for named forward references to subpatterns,
927        the only newline sequence it recognized was NL. It now handles newlines
928        according to the set newline convention.
929    
930    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
931        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
932        cater for a lack of strerror(). These oversights have been fixed.
933    
934    9.  Added --match-limit and --recursion-limit to pcregrep.
935    
936    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
937    
938    11. When the -o option was used, pcregrep was setting a return code of 1, even
939        when matches were found, and --line-buffered was not being honoured.
940    
941    12. Added an optional parentheses number to the -o and --only-matching options
942        of pcregrep.
943    
944    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
945        can match an empty string. The code to do it in pcretest and pcredemo
946        needed fixing:
947    
948        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
949            only one byte after an empty string match just before CRLF (this case
950            just got forgotten; "any" and "anycrlf" were OK).
951    
952        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
953            mode when an empty string match preceded an ASCII character followed by
954            a non-ASCII character. (The code for advancing by one character rather
955            than one byte was nonsense.)
956    
957        (c) The pcredemo.c sample program did not have any code at all to handle
958            the cases when CRLF is a valid newline sequence.
959    
960    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
961        as a starting offset was within the subject string. There is now a new
962        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
963        negative or greater than the length of the string. In order to test this,
964        pcretest is extended to allow the setting of negative starting offsets.
965    
966    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
967        starting offset points to the beginning of a UTF-8 character was
968        unnecessarily clumsy. I tidied it up.
969    
970    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
971        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
972    
973    17. Nobody had reported that the --include_dir option, which was added in
974        release 7.7 should have been called --include-dir (hyphen, not underscore)
975        for compatibility with GNU grep. I have changed it to --include-dir, but
976        left --include_dir as an undocumented synonym, and the same for
977        --exclude-dir, though that is not available in GNU grep, at least as of
978        release 2.5.4.
979    
980    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
981        characters from a string of bytes) have been redefined so as not to use
982        loops, in order to improve performance in some environments. At the same
983        time, I abstracted some of the common code into auxiliary macros to save
984        repetition (this should not affect the compiled code).
985    
986    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
987        compile-time error is now given if \c is not followed by an ASCII
988        character, that is, a byte less than 128. (In EBCDIC mode, the code is
989        different, and any byte value is allowed.)
990    
991    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
992        START_OPTIMIZE option, which is now allowed at compile time - but just
993        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
994        to pcregrep and other applications that have no direct access to PCRE
995        options. The new /Y option in pcretest sets this option when calling
996        pcre_compile().
997    
998    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
999        back references. Groups containing recursive back references were forced to
1000        be atomic by that change, but in the case of named groups, the amount of
1001        memory required was incorrectly computed, leading to "Failed: internal
1002        error: code overflow". This has been fixed.
1003    
1004    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1005        pcretest.c, to avoid build problems in some Borland environments.
1006    
1007    
1008    Version 8.10 25-Jun-2010
1009    ------------------------
1010    
1011    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1012        THEN.
1013    
1014    2.  (*ACCEPT) was not working when inside an atomic group.
1015    
1016    3.  Inside a character class, \B is treated as a literal by default, but
1017        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1018        causes the error). The code is unchanged, but I tidied the documentation.
1019    
1020    4.  Inside a character class, PCRE always treated \R and \X as literals,
1021        whereas Perl faults them if its -w option is set. I have changed PCRE so
1022        that it faults them when PCRE_EXTRA is set.
1023    
1024    5.  Added support for \N, which always matches any character other than
1025        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1026    
1027    6.  When compiling pcregrep with newer versions of gcc which may have
1028        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1029        declared with attribute warn_unused_result" were given. Just casting the
1030        result to (void) does not stop the warnings; a more elaborate fudge is
1031        needed. I've used a macro to implement this.
1032    
1033    7.  Minor change to pcretest.c to avoid a compiler warning.
1034    
1035    8.  Added four artifical Unicode properties to help with an option to make
1036        \s etc use properties (see next item). The new properties are: Xan
1037        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1038    
1039    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1040        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1041        this option. Modified pcretest to add /W to test this facility. Added
1042        REG_UCP to make it available via the POSIX interface.
1043    
1044    10. Added --line-buffered to pcregrep.
1045    
1046    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1047        studied, and the match started with a letter with a code point greater than
1048        127 whose first byte was different to the first byte of the other case of
1049        the letter, the other case of this starting letter was not recognized
1050        (#976).
1051    
1052    12. If a pattern that was studied started with a repeated Unicode property
1053        test, for example, \p{Nd}+, there was the theoretical possibility of
1054        setting up an incorrect bitmap of starting bytes, but fortunately it could
1055        not have actually happened in practice until change 8 above was made (it
1056        added property types that matched character-matching opcodes).
1057    
1058    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1059        possible starting bytes for non-anchored patterns.
1060    
1061    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1062        \R, and also a number of cases that involve Unicode properties, both
1063        explicit and implicit when PCRE_UCP is set.
1064    
1065    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1066        input, it could crash or give wrong results if characters with values
1067        greater than 0xc0 were present in the subject string. (Detail: it assumed
1068        UTF-8 input when processing these items.)
1069    
1070    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1071        size_t is 64-bit (#991).
1072    
1073    17. Added a check for running out of memory when PCRE is compiled with
1074        --disable-stack-for-recursion (#990).
1075    
1076    18. If the last data line in a file for pcretest does not have a newline on
1077        the end, a newline was missing in the output.
1078    
1079    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1080        less than 128) in its various bitmaps. However, there is a facility for
1081        generating tables according to the current locale when PCRE is compiled. It
1082        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1083        characters, are recognized by isspace() and therefore were getting set in
1084        these tables, and indeed these tables seem to approximate to ISO 8859. This
1085        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1086        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1087        which of course cannot start UTF-8 characters. I have changed the code so
1088        that only real ASCII characters (less than 128) and the correct starting
1089        bytes for UTF-8 encodings are set for characters greater than 127 when in
1090        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1091        altogether.)
1092    
1093    20. Added the /T option to pcretest so as to be able to run tests with non-
1094        standard character tables, thus making it possible to include the tests
1095        used for 19 above in the standard set of tests.
1096    
1097    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1098        reference to a subpattern the other side of a comment that contains an
1099        opening parenthesis caused either an internal compiling error, or a
1100        reference to the wrong subpattern.
1101    
1102    
1103    Version 8.02 19-Mar-2010
1104    ------------------------
1105    
1106    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1107    
1108    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1109        configured.
1110    
1111    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1112        original author of that file, following a query about its status.
1113    
1114    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1115        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1116    
1117    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1118        quantifier applied to a forward-referencing subroutine call, could compile
1119        incorrect code or give the error "internal error: previously-checked
1120        referenced subpattern not found".
1121    
1122    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1123        variables to point to external functions. For these systems, therefore,
1124        pcre_malloc etc. are now initialized to local functions that call the
1125        relevant global functions.
1126    
1127    7.  There were two entries missing in the vectors called coptable and poptable
1128        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1129        I've fixed the data, and added a kludgy way of testing at compile time that
1130        the lengths are correct (equal to the number of opcodes).
1131    
1132    8.  Following on from 7, I added a similar kludge to check the length of the
1133        eint vector in pcreposix.c.
1134    
1135    9.  Error texts for pcre_compile() are held as one long string to avoid too
1136        much relocation at load time. To find a text, the string is searched,
1137        counting zeros. There was no check for running off the end of the string,
1138        which could happen if a new error number was added without updating the
1139        string.
1140    
1141    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1142    
1143    11. \K was not working if it appeared in an atomic group or in a group that
1144        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1145        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1146        the assertion is positive, but not if it is negative.
1147    
1148    12. Change 11 fortuitously reduced the size of the stack frame used in the
1149        "match()" function of pcre_exec.c by one pointer. Forthcoming
1150        implementation of support for (*MARK) will need an extra pointer on the
1151        stack; I have reserved it now, so that the stack frame size does not
1152        decrease.
1153    
1154    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1155        item in branch that calls a recursion is a subroutine call - as in the
1156        second branch in the above example - was incorrectly given the compile-
1157        time error "recursive call could loop indefinitely" because pcre_compile()
1158        was not correctly checking the subroutine for matching a non-empty string.
1159    
1160    14. The checks for overrunning compiling workspace could trigger after an
1161        overrun had occurred. This is a "should never occur" error, but it can be
1162        triggered by pathological patterns such as hundreds of nested parentheses.
1163        The checks now trigger 100 bytes before the end of the workspace.
1164    
1165    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1166    
1167    
1168    Version 8.01 19-Jan-2010
1169    ------------------------
1170    
1171    1.  If a pattern contained a conditional subpattern with only one branch (in
1172        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1173        computed the wrong minimum data length (which is of course zero for such
1174        subpatterns). This could cause incorrect "no match" results.
1175    
1176    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1177        the pattern is reset in the first branch, pcre_compile() failed with
1178        "internal error: code overflow at offset...". This happened only when
1179        the reset was to the original external option setting. (An optimization
1180        abstracts leading options settings into an external setting, which was the
1181        cause of this.)
1182    
1183    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1184        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1185        assertion pattern did not match (meaning that the assertion was true), it
1186        was incorrectly treated as false if the SKIP had been reached during the
1187        matching. This also applied to assertions used as conditions.
1188    
1189    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1190        assertion subpattern, including such a pattern used as a condition,
1191        unpredictable results occurred, instead of the error return
1192        PCRE_ERROR_DFA_UITEM.
1193    
1194    5.  The C++ GlobalReplace function was not working like Perl for the special
1195        situation when an empty string is matched. It now does the fancy magic
1196        stuff that is necessary.
1197    
1198    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1199        removed. (These were left over from very, very early versions of PCRE.)
1200    
1201    7.  Some cosmetic changes to the code to make life easier when compiling it
1202        as part of something else:
1203    
1204        (a) Change DEBUG to PCRE_DEBUG.
1205    
1206        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1207            called "current" as "current_branch", to prevent a collision with the
1208            Linux macro when compiled as a kernel module.
1209    
1210        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1211            prevent a collision with the Linux macro when compiled as a kernel
1212            module.
1213    
1214    8.  In pcre_compile() there are some checks for integer overflows that used to
1215        cast potentially large values to (double). This has been changed to that
1216        when building, a check for int64_t is made, and if it is found, it is used
1217        instead, thus avoiding the use of floating point arithmetic. (There is no
1218        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1219        double.
1220    
1221    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1222        2005 (difference between two addresses compared to an unsigned value).
1223    
1224    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1225        custom one, because of the following reported problem in Windows:
1226    
1227          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1228              under Win32.
1229          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1230              therefore missing the function definition.
1231          - The compiler thus generates a "C" signature for the test function.
1232          - The linker fails to find the "C" function.
1233          - PCRE fails to configure if asked to do so against libbz2.
1234    
1235    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1236        messages were output:
1237    
1238          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1239          rerunning libtoolize, to keep the correct libtool macros in-tree.
1240          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1241    
1242        I have done both of these things.
1243    
1244    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1245        most of the time, it *can* run out if it is given a pattern that contains a
1246        runaway infinite recursion. I updated the discussion in the pcrestack man
1247        page.
1248    
1249    13. Now that we have gone to the x.xx style of version numbers, the minor
1250        version may start with zero. Using 08 or 09 is a bad idea because users
1251        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1252        interpreted as invalid octal numbers. I've updated the previous comment in
1253        configure.ac, and also added a check that gives an error if 08 or 09 are
1254        used.
1255    
1256    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1257        causing partial matching to fail when the end of the subject matched \W
1258        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1259    
1260    15. There were some discrepancies between the declarations in pcre_internal.h
1261        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1262        their definitions. The declarations used "const uschar *" and the
1263        definitions used USPTR. Even though USPTR is normally defined as "const
1264        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1265        reported that: "This difference in casting confuses some C++ compilers, for
1266        example, SunCC recognizes above declarations as different functions and
1267        generates broken code for hbpcre." I have changed the declarations to use
1268        USPTR.
1269    
1270    16. GNU libtool is named differently on some systems. The autogen.sh script now
1271        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1272        (FreeBSD).
1273    
1274    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1275        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1276        comment: "Figure out how to create a longlong from a string: strtoll and
1277        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1278        instance, but it only takes 2 args instead of 3!"
1279    
1280    18. A subtle bug concerned with back references has been fixed by a change of
1281        specification, with a corresponding code fix. A pattern such as
1282        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1283        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1284        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1285        same bug. Such groups have to be quantified to be useful, or contained
1286        inside another quantified group. (If there's no repetition, the reference
1287        can never match.) The problem arises because, having left the group and
1288        moved on to the rest of the pattern, a later failure that backtracks into
1289        the group uses the captured value from the final iteration of the group
1290        rather than the correct earlier one. I have fixed this in PCRE by forcing
1291        any group that contains a reference to itself to be an atomic group; that
1292        is, there cannot be any backtracking into it once it has completed. This is
1293        similar to recursive and subroutine calls.
1294    
1295    
1296    Version 8.00 19-Oct-09
1297    ----------------------
1298    
1299    1.  The table for translating pcre_compile() error codes into POSIX error codes
1300        was out-of-date, and there was no check on the pcre_compile() error code
1301        being within the table. This could lead to an OK return being given in
1302        error.
1303    
1304    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1305        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1306        in a Windows environment.
1307    
1308    3.  The pcregrep --count option prints the count for each file even when it is
1309        zero, as does GNU grep. However, pcregrep was also printing all files when
1310        --files-with-matches was added. Now, when both options are given, it prints
1311        counts only for those files that have at least one match. (GNU grep just
1312        prints the file name in this circumstance, but including the count seems
1313        more useful - otherwise, why use --count?) Also ensured that the
1314        combination -clh just lists non-zero counts, with no names.
1315    
1316    4.  The long form of the pcregrep -F option was incorrectly implemented as
1317        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1318        but it seems right to fix it, and I didn't think it was worth preserving
1319        the old behaviour.
1320    
1321    5.  The command line items --regex=pattern and --regexp=pattern were not
1322        recognized by pcregrep, which required --regex pattern or --regexp pattern
1323        (with a space rather than an '='). The man page documented the '=' forms,
1324        which are compatible with GNU grep; these now work.
1325    
1326    6.  No libpcreposix.pc file was created for pkg-config; there was just
1327        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1328    
1329    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1330        when UCP support is not needed, by modifying the Python script that
1331        generates it from Unicode data files. This should not matter if the module
1332        is correctly used as a library, but I received one complaint about 50K of
1333        unwanted data. My guess is that the person linked everything into his
1334        program rather than using a library. Anyway, it does no harm.
1335    
1336    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1337        was a minimum greater than 1 for a wide character in a possessive
1338        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1339        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1340        character. Chaos in the form of incorrect output or a compiling loop could
1341        result.
1342    
1343    9.  The restrictions on what a pattern can contain when partial matching is
1344        requested for pcre_exec() have been removed. All patterns can now be
1345        partially matched by this function. In addition, if there are at least two
1346        slots in the offset vector, the offset of the earliest inspected character
1347        for the match and the offset of the end of the subject are set in them when
1348        PCRE_ERROR_PARTIAL is returned.
1349    
1350    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1351        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1352        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1353        and may be more useful for multi-segment matching.
1354    
1355    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1356        used to be given if ever the end of the subject was reached; now it is
1357        given only if matching could not proceed because another character was
1358        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1359        string "Z", which now yields "no match" instead of "partial match". In the
1360        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1361        final character ended with (*FAIL).
1362    
1363    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1364        if the pattern had a "must contain" character that was already found in the
1365        earlier partial match, unless partial matching was again requested. For
1366        example, with the pattern /dog.(body)?/, the "must contain" character is
1367        "g". If the first part-match was for the string "dog", restarting with
1368        "sbody" failed. This bug has been fixed.
1369    
1370    13. The string returned by pcre_dfa_exec() after a partial match has been
1371        changed so that it starts at the first inspected character rather than the
1372        first character of the match. This makes a difference only if the pattern
1373        starts with a lookbehind assertion or \b or \B (\K is not supported by
1374        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1375        matching functions compatible, and I think it's the right thing to do.
1376    
1377    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1378        so that the demonstration program is easily available in environments where
1379        PCRE has not been installed from source.
1380    
1381    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1382        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1383        library.
1384    
1385    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1386        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1387        is not the first non-POSIX option to be added. Clearly some people find
1388        these options useful.
1389    
1390    17. If a caller to the POSIX matching function regexec() passes a non-zero
1391        value for nmatch with a NULL value for pmatch, the value of
1392        nmatch is forced to zero.
1393    
1394    18. RunGrepTest did not have a test for the availability of the -u option of
1395        the diff command, as RunTest does. It now checks in the same way as
1396        RunTest, and also checks for the -b option.
1397    
1398    19. If an odd number of negated classes containing just a single character
1399        interposed, within parentheses, between a forward reference to a named
1400        subpattern and the definition of the subpattern, compilation crashed with
1401        an internal error, complaining that it could not find the referenced
1402        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1403        [The bug was that it was starting one character too far in when skipping
1404        over the character class, thus treating the ] as data rather than
1405        terminating the class. This meant it could skip too much.]
1406    
1407    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1408        /g option in pcretest when the pattern contains \K, which makes it possible
1409        to have an empty string match not at the start, even when the pattern is
1410        anchored. Updated pcretest and pcredemo to use this option.
1411    
1412    21. If the maximum number of capturing subpatterns in a recursion was greater
1413        than the maximum at the outer level, the higher number was returned, but
1414        with unset values at the outer level. The correct (outer level) value is
1415        now given.
1416    
1417    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1418        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1419        make it do so. The string so far is captured, making this feature
1420        compatible with Perl.
1421    
1422    23. The tests have been re-organized, adding tests 11 and 12, to make it
1423        possible to check the Perl 5.10 features against Perl 5.10.
1424    
1425    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1426        pattern matches a fixed length string. PCRE did not allow this; now it
1427        does. Neither allows recursion.
1428    
1429    25. I finally figured out how to implement a request to provide the minimum
1430        length of subject string that was needed in order to match a given pattern.
1431        (It was back references and recursion that I had previously got hung up
1432        on.) This code has now been added to pcre_study(); it finds a lower bound
1433        to the length of subject needed. It is not necessarily the greatest lower
1434        bound, but using it to avoid searching strings that are too short does give
1435        some useful speed-ups. The value is available to calling programs via
1436        pcre_fullinfo().
1437    
1438    26. While implementing 25, I discovered to my embarrassment that pcretest had
1439        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1440        study optimizations had never been tested with that matching function.
1441        Oops. What is worse, even when it was passed study data, there was a bug in
1442        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1443        were also very few tests of studied patterns with pcre_dfa_exec().
1444    
1445    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1446        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1447        on the other side of the coin, they are no longer allowed to have different
1448        names, because these cannot be distinguished in PCRE, and this has caused
1449        confusion. (This is a difference from Perl.)
1450    
1451    28. When duplicate subpattern names are present (necessarily with different
1452        numbers, as required by 27 above), and a test is made by name in a
1453        conditional pattern, either for a subpattern having been matched, or for
1454        recursion in such a pattern, all the associated numbered subpatterns are
1455        tested, and the overall condition is true if the condition is true for any
1456        one of them. This is the way Perl works, and is also more like the way
1457        testing by number works.
1458    
1459    
1460    Version 7.9 11-Apr-09
1461    ---------------------
1462    
1463    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1464        (pcretest), all targets were linked against these libraries. This included
1465        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1466        libraries. This caused unwanted dependencies to be created. This problem
1467        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1468        pcretest is linked with readline.
1469    
1470    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1471        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1472        moved outside it again, because FALSE and TRUE are already defined in AIX,
1473        but BOOL is not.
1474    
1475    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1476        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1477    
1478    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1479        hyphens) following file names and line numbers when outputting matching
1480        lines. This is not true; no spaces are inserted. I have also clarified the
1481        wording for the --colour (or --color) option.
1482    
1483    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1484        was not coloured; this is different to GNU grep, so I have changed it to be
1485        the same.
1486    
1487    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1488        each matching line was coloured. Now it goes on to look for further matches
1489        of any of the test patterns, which is the same behaviour as GNU grep.
1490    
1491    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1492        doesn't make sense to accept an empty string match in pcregrep, so I have
1493        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1494        seems to be how GNU grep behaves.
1495    
1496    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1497        start or after a newline", because the conditional assertion was not being
1498        correctly handled. The rule now is that both the assertion and what follows
1499        in the first alternative must satisfy the test.
1500    
1501    9.  If auto-callout was enabled in a pattern with a conditional group whose
1502        condition was an assertion, PCRE could crash during matching, both with
1503        pcre_exec() and pcre_dfa_exec().
1504    
1505    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1506        used for matching.
1507    
1508    11. Unicode property support in character classes was not working for
1509        characters (bytes) greater than 127 when not in UTF-8 mode.
1510    
1511    12. Added the -M command line option to pcretest.
1512    
1513    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1514    
1515    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1516    
1517    16. Added comments and documentation about mis-use of no_arg in the C++
1518        wrapper.
1519    
1520    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1521        from Martin Jerabek that uses macro names for all relevant character and
1522        string constants.
1523    
1524    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1525        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1526        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1527        these, but not everybody uses configure.
1528    
1529    19. A conditional group that had only one branch was not being correctly
1530        recognized as an item that could match an empty string. This meant that an
1531        enclosing group might also not be so recognized, causing infinite looping
1532        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1533        with the subject "ab", where knowledge that the repeated group can match
1534        nothing is needed in order to break the loop.
1535    
1536    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1537        exec(), but without supplying a callout function, matching went wrong.
1538    
1539    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1540        leak if the size of the offset vector was greater than 30. When the vector
1541        is smaller, the saved offsets during recursion go onto a local stack
1542        vector, but for larger vectors malloc() is used. It was failing to free
1543        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1544        error, in fact).
1545    
1546    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1547        heapframe that is used only when UTF-8 support is enabled. This caused no
1548        problem, but was untidy.
1549    
1550    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1551        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1552        included within another project.
1553    
1554    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1555        slightly modified by me:
1556    
1557          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1558              not building pcregrep.
1559    
1560          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1561              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1562    
1563    25. Forward references, both numeric and by name, in patterns that made use of
1564        duplicate group numbers, could behave incorrectly or give incorrect errors,
1565        because when scanning forward to find the reference group, PCRE was not
1566        taking into account the duplicate group numbers. A pattern such as
1567        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1568    
1569    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1570        the feature of a custom pointer more persuasive (as requested by a user).
1571    
1572    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1573        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1574        pre-defined.
1575    
1576    28. Added support for (*UTF8) at the start of a pattern.
1577    
1578    29. Arrange for flags added by the "release type" setting in CMake to be shown
1579        in the configuration summary.
1580    
1581    
1582    Version 7.8 05-Sep-08
1583    ---------------------
1584    
1585    1.  Replaced UCP searching code with optimized version as implemented for Ad
1586        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1587        stage table and inline lookup instead of a function, giving speed ups of 2
1588        to 5 times on some simple patterns that I tested. Permission was given to
1589        distribute the MultiStage2.py script that generates the tables (it's not in
1590        the tarball, but is in the Subversion repository).
1591    
1592    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1593        scripts.
1594    
1595    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1596        a group with a zero qualifier. The result of the study could be incorrect,
1597        or the function might crash, depending on the pattern.
1598    
1599    4.  Caseless matching was not working for non-ASCII characters in back
1600        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1601        It now works when Unicode Property Support is available.
1602    
1603    5.  In pcretest, an escape such as \x{de} in the data was always generating
1604        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1605        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1606        truncation.
1607    
1608    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1609    
1610    7.  Added two (int) casts to pcregrep when printing the difference of two
1611        pointers, in case they are 64-bit values.
1612    
1613    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1614        test 2 if it fails.
1615    
1616    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1617        and a #define of that name to empty if it is not externally set. This is to
1618        allow users of MSVC to set it if necessary.
1619    
1620    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1621        the convenience functions in the pcre_get.c source file.
1622    
1623    11. An option change at the start of a pattern that had top-level alternatives
1624        could cause overwriting and/or a crash. This command provoked a crash in
1625        some environments:
1626    
1627          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1628    
1629        This potential security problem was recorded as CVE-2008-2371.
1630    
1631    12. For a pattern where the match had to start at the beginning or immediately
1632        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1633        pcre_dfa_exec() could read past the end of the passed subject if there was
1634        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1635        pcretest so that it places the subject at the end of its malloc-ed buffer.
1636    
1637    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1638        exec() might read past the end of the data buffer in UTF-8 mode.
1639    
1640    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1641        the data contained the byte 0x85 as part of a UTF-8 character within its
1642        first line. This applied both to normal and DFA matching.
1643    
1644    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1645        /^[^d]*?$/8 failed to match "abc".
1646    
1647    16. Added a missing copyright notice to pcrecpp_internal.h.
1648    
1649    17. Make it more clear in the documentation that values returned from
1650        pcre_exec() in ovector are byte offsets, not character counts.
1651    
1652    18. Tidied a few places to stop certain compilers from issuing warnings.
1653    
1654    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1655        supplied by Stefan Weber. I made a further small update for 7.8 because
1656        there is a change of source arrangements: the pcre_searchfuncs.c module is
1657        replaced by pcre_ucd.c.
1658    
1659    
1660    Version 7.7 07-May-08
1661    ---------------------
1662    
1663    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1664        a string to a long long, pretend we don't even have a long long." This is
1665        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1666    
1667    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1668        pre-7.6 versions, which defined a global no_arg variable instead of putting
1669        it in the RE class. (See also #8 below.)
1670    
1671    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1672        Lopes.
1673    
1674    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1675    
1676        (1) The include/exclude patterns were being applied to the whole pathnames
1677            of files, instead of just to the final components.
1678    
1679        (2) If there was more than one level of directory, the subdirectories were
1680            skipped unless they satisfied the include/exclude conditions. This is
1681            inconsistent with GNU grep (and could even be seen as contrary to the
1682            pcregrep specification - which I improved to make it absolutely clear).
1683            The action now is always to scan all levels of directory, and just
1684            apply the include/exclude patterns to regular files.
1685    
1686    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1687        --exclude_dir in the tests to avoid scanning .svn directories.
1688    
1689    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1690        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1691        doesn't support NULs in patterns.
1692    
1693    7.  Added some missing "const"s to declarations of static tables in
1694        pcre_compile.c and pcre_dfa_exec.c.
1695    
1696    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1697        caused by fix #2  above. (Subsequently also a second patch to fix the
1698        first patch. And a third patch - this was a messy problem.)
1699    
1700    9.  Applied Craig's patch to remove the use of push_back().
1701    
1702    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1703        matching function regexec().
1704    
1705    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1706        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1707        references. PCRE supports relative numbers with this syntax (I don't think
1708        Oniguruma does).
1709    
1710    12. Previously, a group with a zero repeat such as (...){0} was completely
1711        omitted from the compiled regex. However, this means that if the group
1712        was called as a subroutine from elsewhere in the pattern, things went wrong
1713        (an internal error was given). Such groups are now left in the compiled
1714        pattern, with a new opcode that causes them to be skipped at execution
1715        time.
1716    
1717    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1718        to the way PCRE behaves:
1719    
1720        (a) A lone ] character is dis-allowed (Perl treats it as data).
1721    
1722        (b) A back reference to an unmatched subpattern matches an empty string
1723            (Perl fails the current match path).
1724    
1725        (c) A data ] in a character class must be notated as \] because if the
1726            first data character in a class is ], it defines an empty class. (In
1727            Perl it is not possible to have an empty class.) The empty class []
1728            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1729            The negative empty class [^] matches any one character, independently
1730            of the DOTALL setting.
1731    
1732    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1733        non-existent subpattern following a character class starting with ']' and
1734        containing () gave an internal compiling error instead of "reference to
1735        non-existent subpattern". Fortunately, when the pattern did exist, the
1736        compiled code was correct. (When scanning forwards to check for the
1737        existencd of the subpattern, it was treating the data ']' as terminating
1738        the class, so got the count wrong. When actually compiling, the reference
1739        was subsequently set up correctly.)
1740    
1741    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1742        it was being rejected as not supported by pcre_dfa_exec(), even though
1743        other assertions are supported. I have made pcre_dfa_exec() support
1744        (*FAIL).
1745    
1746    16. The implementation of 13c above involved the invention of a new opcode,
1747        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1748        cannot be changed at match time, I realized I could make a small
1749        improvement to matching performance by compiling OP_ALLANY instead of
1750        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1751        on the OP_ANY path.
1752    
1753    17. Compiling pcretest on Windows with readline support failed without the
1754        following two fixes: (1) Make the unistd.h include conditional on
1755        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1756    
1757    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1758        ncurses library to be included for pcretest when ReadLine support is
1759        requested, but also to allow for it to be overridden. This patch came from
1760        Daniel Bergström.
1761    
1762    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1763        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1764        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1765        spotting this.
1766    
1767    
1768    Version 7.6 28-Jan-08
1769    ---------------------
1770    
1771    1.  A character class containing a very large number of characters with
1772        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1773        overflow.
1774    
1775    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1776        HAVE_LONG_LONG is not defined.
1777    
1778    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1779        bring them up to date and include new features. This patch includes:
1780    
1781        - Fixed PH's badly added libz and libbz2 support.
1782        - Fixed a problem with static linking.
1783        - Added pcredemo. [But later removed - see 7 below.]
1784        - Fixed dftables problem and added an option.
1785        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1786            HAVE_LONG_LONG.
1787        - Added readline support for pcretest.
1788        - Added an listing of the option settings after cmake has run.
1789    
1790    4.  A user submitted a patch to Makefile that makes it easy to create
1791        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1792        Makefile.am that cause it to include this special target, without
1793        affecting anything else. Note that the same mingw target plus all
1794        the other distribution libraries and programs are now supported
1795        when configuring with CMake (see 6 below) instead of with
1796        Configure/Make.
1797    
1798    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1799        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1800        exported in the Windows port". It has not yet been confirmed that the patch
1801        solves the problem, but it does no harm.
1802    
1803    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1804        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1805        with CMake, and also correct the comment about stack recursion.
1806    
1807    7.  Remove the automatic building of pcredemo from the ./configure system and
1808        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1809        of a program that users should build themselves after PCRE is installed, so
1810        building it automatically is not really right. What is more, it gave
1811        trouble in some build environments.
1812    
1813    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1814    
1815    
1816    Version 7.5 10-Jan-08
1817    ---------------------
1818    
1819    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1820        values in parens when parsing an RE using the C++ wrapper."
1821    
1822    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1823        Characters greater than 255 were excluded from the class instead of being
1824        included.
1825    
1826    3.  The same bug as (2) above applied to negated POSIX classes such as
1827        [:^space:].
1828    
1829    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1830        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1831        I have changed it.
1832    
1833    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1834        first named subpattern) and a construct such as (?&a) would reference the
1835        first named subpattern whose name started with "a" (in other words, the
1836        length check was missing). Both these problems are fixed. "Subpattern name
1837        expected" is now given for (?&) (a zero-length name), and this patch also
1838        makes it give the same error for \k'' (previously it complained that that
1839        was a reference to a non-existent subpattern).
1840    
1841    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1842        this is right because (?- can be followed by option settings as well as by
1843        digits. I have, however, made the messages clearer.
1844    
1845    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1846        than the number used in the conditional) now cause a compile-time error.
1847        This is actually not compatible with Perl, which accepts such patterns, but
1848        treats the conditional as always being FALSE (as PCRE used to), but it
1849        seems to me that giving a diagnostic is better.
1850    
1851    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1852        and messages.
1853    
1854    9.  Fix two occurrences of "backslash" in comments that should have been
1855        "backspace".
1856    
1857    10. Remove two redundant lines of code that can never be obeyed (their function
1858        was moved elsewhere).
1859    
1860    11. The program that makes PCRE's Unicode character property table had a bug
1861        which caused it to generate incorrect table entries for sequences of
1862        characters that have the same character type, but are in different scripts.
1863        It amalgamated them into a single range, with the script of the first of
1864        them. In other words, some characters were in the wrong script. There were
1865        thirteen such cases, affecting characters in the following ranges:
1866    
1867          U+002b0 - U+002c1
1868          U+0060c - U+0060d
1869          U+0061e - U+00612
1870          U+0064b - U+0065e
1871          U+0074d - U+0076d
1872          U+01800 - U+01805
1873          U+01d00 - U+01d77
1874          U+01d9b - U+01dbf
1875          U+0200b - U+0200f
1876          U+030fc - U+030fe
1877          U+03260 - U+0327f
1878          U+0fb46 - U+0fbb1
1879          U+10450 - U+1049d
1880    
1881    12. The -o option (show only the matching part of a line) for pcregrep was not
1882        compatible with GNU grep in that, if there was more than one match in a
1883        line, it showed only the first of them. It now behaves in the same way as
1884        GNU grep.
1885    
1886    13. If the -o and -v options were combined for pcregrep, it printed a blank
1887        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1888        does the same. The return code can be used to tell if there were any
1889        non-matching lines.
1890    
1891    14. Added --file-offsets and --line-offsets to pcregrep.
1892    
1893    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1894        infinitely looping recursion. The bug was that positive lookaheads were not
1895        being skipped when checking for a possible empty match (negative lookaheads
1896        and both kinds of lookbehind were skipped).
1897    
1898    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1899        inclusion of <windows.h> to before rather than after the definition of
1900        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1901    
1902    17. Specifying a possessive quantifier with a specific limit for a Unicode
1903        character property caused pcre_compile() to compile bad code, which led at
1904        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1905        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1906        caused the error; without that there was no problem.
1907    
1908    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1909    
1910    19. Added --enable-pcretest-libreadline.
1911    
1912    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1913        RE::GlobalReplace(). As a result, the number of replacements returned was
1914        double what it should be. I removed one of the increments, but Craig sent a
1915        later patch that removed the other one (the right fix) and added unit tests
1916        that check the return values (which was not done before).
1917    
1918    21. Several CMake things:
1919    
1920        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1921            the names libpcre and libpcreposix, not just pcre and pcreposix.
1922    
1923        (2) The above change means that pcretest and pcregrep are now correctly
1924            linked with the newly-built libraries, not previously installed ones.
1925    
1926        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1927    
1928    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1929        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1930        UTF-8 newline character). The key issue is that the pattern starts .*;
1931        this means that the match must be either at the beginning, or after a
1932        newline. The bug was in the code for advancing after a failed match and
1933        checking that the new position followed a newline. It was not taking
1934        account of UTF-8 characters correctly.
1935    
1936    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1937        character classes. PCRE was not treating the sequence [:...:] as a
1938        character class unless the ... were all letters. Perl, however, seems to
1939        allow any characters between [: and :], though of course it rejects as
1940        unknown any "names" that contain non-letters, because all the known class
1941        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1942        for example, whereas PCRE did not - it did not recognize a POSIX character
1943        class. This seemed a bit dangerous, so the code has been changed to be
1944        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1945        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1946        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1947        Perl does, and where it didn't before.
1948    
1949    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1950        Windows environments %n is disabled by default.
1951    
1952    
1953    Version 7.4 21-Sep-07
1954    ---------------------
1955    
1956    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1957        means that a class such as [\s] counted as "explicit reference to CR or
1958        LF". That isn't really right - the whole point of the change was to try to
1959        help when there was an actual mention of one of the two characters. So now
1960        the change happens only if \r or \n (or a literal CR or LF) character is
1961        encountered.
1962    
1963    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1964        of both had grown to the point where there were only 3 bits left.
1965        Fortunately, there was spare space in the data structure, and so I have
1966        moved the internal flags into a new 16-bit field to free up more option
1967        bits.
1968    
1969    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1970        but did not set the internal JCHANGED flag - either of these is enough to
1971        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1972        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1973        start sets both bits.
1974    
1975    4.  Added options (at build time, compile time, exec time) to change \R from
1976        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1977    
1978    5.  doc/pcresyntax.html was missing from the distribution.
1979    
1980    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1981        compatibility, even though it is no longer used.
1982    
1983    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1984        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1985        windows.h file is present (where different names are used). [This was
1986        reversed later after testing - see 16 below.]
1987    
1988    8.  Changed all #include <config.h> to #include "config.h". There were also
1989        some further <pcre.h> cases that I changed to "pcre.h".
1990    
1991    9.  When pcregrep was used with the --colour option, it missed the line ending
1992        sequence off the lines that it output.
1993    
1994    10. It was pointed out to me that arrays of string pointers cause lots of
1995        relocations when a shared library is dynamically loaded. A technique of
1996        using a single long string with a table of offsets can drastically reduce
1997        these. I have refactored PCRE in four places to do this. The result is
1998        dramatic:
1999    
2000          Originally:                          290
2001          After changing UCP table:            187
2002          After changing error message table:   43
2003          After changing table of "verbs"       36
2004          After changing table of Posix names   22
2005    
2006        Thanks to the folks working on Gregex for glib for this insight.
2007    
2008    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2009        unicode-properties was also set.
2010    
2011    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2012    
2013    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2014        checked only for CRLF.
2015    
2016    14. Added casts to pcretest.c to avoid compiler warnings.
2017    
2018    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2019    
2020    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2021        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2022        entirely. This removes changes made in 7 above.
2023    
2024    17. The CMake files have been updated, and there is now more information about
2025        building with CMake in the NON-UNIX-USE document.
2026    
2027    
2028    Version 7.3 28-Aug-07
2029    ---------------------
2030    
2031     1. In the rejigging of the build system that eventually resulted in 7.1, the
2032        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2033        brackets there is not right, since it causes compilers to look for an
2034        installed pcre.h, not the version that is in the source that is being
2035        compiled (which of course may be different). I have changed it back to:
2036    
2037          #include "pcre.h"
2038    
2039        I have a vague recollection that the change was concerned with compiling in
2040        different directories, but in the new build system, that is taken care of
2041        by the VPATH setting the Makefile.
2042    
2043     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2044        when the subject happened to end in the byte 0x85 (e.g. if the last
2045        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2046        characters but of course it shouldn't be taken as a newline when it is part
2047        of another character. The bug was that, for an unlimited repeat of . in
2048        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2049        characters when looking for a newline.
2050    
2051     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2052    
2053     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2054        in debug output.
2055    
2056     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2057        long printing in the pcrecpp unittest when running under MinGW.
2058    
2059     6. ESC_K was left out of the EBCDIC table.
2060    
2061     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2062        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2063        limit also applies to "virtual nesting" when a pattern is recursive, and in
2064        this case 1000 isn't so big. I have been able to remove this limit at the
2065        expense of backing off one optimization in certain circumstances. Normally,
2066        when pcre_exec() would call its internal match() function recursively and
2067        immediately return the result unconditionally, it uses a "tail recursion"
2068        feature to save stack. However, when a subpattern that can match an empty
2069        string has an unlimited repetition quantifier, it no longer makes this
2070        optimization. That gives it a stack frame in which to save the data for
2071        checking that an empty string has been matched. Previously this was taken
2072        from the 1000-entry workspace that had been reserved. So now there is no
2073        explicit limit, but more stack is used.
2074    
2075     8. Applied Daniel's patches to solve problems with the import/export magic
2076        syntax that is required for Windows, and which was going wrong for the
2077        pcreposix and pcrecpp parts of the library. These were overlooked when this
2078        problem was solved for the main library.
2079    
2080     9. There were some crude static tests to avoid integer overflow when computing
2081        the size of patterns that contain repeated groups with explicit upper
2082        limits. As the maximum quantifier is 65535, the maximum group length was
2083        set at 30,000 so that the product of these two numbers did not overflow a
2084        32-bit integer. However, it turns out that people want to use groups that
2085        are longer than 30,000 bytes (though not repeat them that many times).
2086        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2087        made it possible to implement the integer overflow checks in a much more
2088        dynamic way, which I have now done. The artificial limitation on group
2089        length has been removed - we now have only the limit on the total length of
2090        the compiled pattern, which depends on the LINK_SIZE setting.
2091    
2092    10. Fixed a bug in the documentation for get/copy named substring when
2093        duplicate names are permitted. If none of the named substrings are set, the
2094        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2095        empty string.
2096    
2097    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2098        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2099        because the ] is interpreted as the first data character and the
2100        terminating ] is not found. PCRE has been made compatible with Perl in this
2101        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2102        cause memory overwriting.
2103    
2104    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2105        string has been matched (to stop an infinite loop). It was not recognizing
2106        a conditional subpattern that could match an empty string if that
2107        subpattern was within another subpattern. For example, it looped when
2108        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2109        condition was not nested. This bug has been fixed.
2110    
2111    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2112        past the start of the subject in the presence of bytes with the top bit
2113        set, for example "\x8aBCD".
2114    
2115    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2116        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2117    
2118    14. Optimized (?!) to (*FAIL).
2119    
2120    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2121        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2122        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2123        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2124        does: it's just the validity check that is more restrictive.
2125    
2126    16. Inserted checks for integer overflows during escape sequence (backslash)
2127        processing, and also fixed erroneous offset values for syntax errors during
2128        backslash processing.
2129    
2130    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2131        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2132    
2133    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2134        caused an overrun.
2135    
2136    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2137        something other than just ASCII characters) inside a group that had an
2138        unlimited repeat caused a loop at compile time (while checking to see
2139        whether the group could match an empty string).
2140    
2141    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2142        [\P{Any}] did so. (Error in the code for printing property names.)
2143    
2144    21. An orphan \E inside a character class could cause a crash.
2145    
2146    22. A repeated capturing bracket such as (A)? could cause a wild memory
2147        reference during compilation.
2148    
2149    23. There are several functions in pcre_compile() that scan along a compiled
2150        expression for various reasons (e.g. to see if it's fixed length for look
2151        behind). There were bugs in these functions when a repeated \p or \P was
2152        present in the pattern. These operators have additional parameters compared
2153        with \d, etc, and these were not being taken into account when moving along
2154        the compiled data. Specifically:
2155    
2156        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2157            length.
2158    
2159        (b) An item such as \pL+ within a repeated group could cause crashes or
2160            loops.
2161    
2162        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2163            "reference to non-existent subpattern" error.
2164    
2165        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2166    
2167    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2168        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2169    
2170    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2171    
2172    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2173        character were causing crashes (broken optimization).
2174    
2175    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2176        \p or \P) caused a compile-time loop.
2177    
2178    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2179        break. For example, the unstudied pattern [\r\n]A does not match the string
2180        "\r\nA" because change 7.0/46 below moves the current point on by two
2181        characters after failing to match at the start. However, the pattern \nA
2182        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2183        the same is true. There doesn't seem any very clean way out of this, but
2184        what I have chosen to do makes the common cases work: PCRE now takes note
2185        of whether there can be an explicit match for \r or \n anywhere in the
2186        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2187        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2188        pattern has explicit CR or LF references.
2189    
2190    29. Added (*CR) etc for changing newline setting at start of pattern.
2191    
2192    
2193    Version 7.2 19-Jun-07
2194    ---------------------
2195    
2196     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2197        which is apparently normally available under Windows.
2198    
2199     2. Re-jig the pcregrep tests with different newline settings in an attempt
2200        to make them independent of the local environment's newline setting.
2201    
2202     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2203    
2204     4. Some of the "internals" tests were previously cut out when the link size
2205        was not 2, because the output contained actual offsets. The recent new
2206        "Z" feature of pcretest means that these can be cut out, making the tests
2207        usable with all link sizes.
2208    
2209     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2210        stack recursion. This gives a massive performance boost under BSD, but just
2211        a small improvement under Linux. However, it saves one field in the frame
2212        in all cases.
2213    
2214     6. Added more features from the forthcoming Perl 5.10:
2215    
2216        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2217            recursion call. It refers to the nth most recently opened parentheses.
2218    
2219        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2220            to be opened parentheses.
2221    
2222        (c) Conditions that refer to capturing parentheses can be specified
2223            relatively, for example, (?(-2)... or (?(+3)...
2224    
2225        (d) \K resets the start of the current match so that everything before
2226            is not part of it.
2227    
2228        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2229    
2230        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2231            reference syntax.
2232    
2233        (g) (?| introduces a group in which the numbering of parentheses in each
2234            alternative starts with the same number.
2235    
2236        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2237    
2238     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2239        PCRE_INFO_JCHANGED.
2240    
2241     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2242        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2243        for detecting groups that can match an empty string.
2244    
2245     9. A pattern with a very large number of alternatives (more than several
2246        hundred) was running out of internal workspace during the pre-compile
2247        phase, where pcre_compile() figures out how much memory will be needed. A
2248        bit of new cunning has reduced the workspace needed for groups with
2249        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2250        workspace instead of running out of the 4096 that are available.
2251    
2252    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2253    
2254    11. Applied patch from Google to remove an optimization that didn't quite work.
2255        The report of the bug said:
2256    
2257          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2258          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2259          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2260    
2261    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2262        it matched the wrong number of bytes.
2263    
2264    
2265    Version 7.1 24-Apr-07
2266  ---------------------  ---------------------
2267    
2268   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
# Line 33  Version 7.1 12-Mar-07 Line 2294  Version 7.1 12-Mar-07
2294   5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan   5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2295      Weber: (1) pcre_internal.h was missing some function renames; (2) updated      Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2296      makevp.bat for the current PCRE, using the additional files      makevp.bat for the current PCRE, using the additional files
2297      makevp-c.txt, makevp-l.txt, and pcregexp.pas.      makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2298    
2299   6. A Windows user reported a minor discrepancy with test 2, which turned out   6. A Windows user reported a minor discrepancy with test 2, which turned out
2300      to be caused by a trailing space on an input line that had got lost in his      to be caused by a trailing space on an input line that had got lost in his
# Line 115  Version 7.1 12-Mar-07 Line 2376  Version 7.1 12-Mar-07
2376    
2377  14. If erroroffset was passed as NULL to pcre_compile, it provoked a  14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2378      segmentation fault instead of returning the appropriate error message.      segmentation fault instead of returning the appropriate error message.
2379    
2380  15. In multiline mode when the newline sequence was set to "any", the pattern  15. In multiline mode when the newline sequence was set to "any", the pattern
2381      ^$ would give a match between the \r and \n of a subject such as "A\r\nB".      ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2382      This doesn't seem right; it now treats the CRLF combination as the line      This doesn't seem right; it now treats the CRLF combination as the line
2383      ending, and so does not match in that case. It's only a pattern such as ^$      ending, and so does not match in that case. It's only a pattern such as ^$
2384      that would hit this one: something like ^ABC$ would have failed after \r      that would hit this one: something like ^ABC$ would have failed after \r
2385      and then tried again after \r\n.      and then tried again after \r\n.
2386    
2387  16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"  16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2388      in an attempt to make files that differ only in their line terminators      in an attempt to make files that differ only in their line terminators
2389      compare equal. This works on Linux.      compare equal. This works on Linux.
2390    
2391    17. Under certain error circumstances pcregrep might try to free random memory
2392        as it exited. This is now fixed, thanks to valgrind.
2393    
2394    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2395        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2396        was because its rules for how to advance for /g after matching an empty
2397        string at the end of a line did not allow for this case. They now check for
2398        it specially.
2399    
2400    20. pcretest is supposed to handle patterns and data of any length, by
2401        extending its buffers when necessary. It was getting this wrong when the
2402        buffer for a data line had to be extended.
2403    
2404    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2405        CRLF as a newline sequence.
2406    
2407    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2408        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2409        I have nevertheless tidied it up.
2410    
2411    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2412    
2413    24. Added a man page for pcre-config.
2414    
2415    
2416  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06

Legend:
Removed from v.131  
changed lines
  Added in v.1190

  ViewVC Help
Powered by ViewVC 1.1.5