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

Diff of /code/trunk/ChangeLog

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

revision 195 by ph10, Mon Jul 30 13:23:28 2007 UTC revision 1045 by ph10, Sun Sep 23 16:50:00 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.3 05-Jul-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.
91    
92        (a) The Unicode property table now has offsets into a new table of sets of
93            three or more characters that are case-equivalent. The MultiStage2.py
94            script that generates these tables (the pcre_ucd.c file) now scans
95            CaseFolding.txt instead of UnicodeData.txt for character case
96            information.
97    
98        (b) The code for adding characters or ranges of characters to a character
99            class has been abstracted into a generalized function that also handles
100            case-independence. In UTF-mode with UCP support, this uses the new data
101            to handle characters with more than one other case.
102    
103        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
104            whose other case is greater than 256 are now correctly matched
105            caselessly. Previously, the high codepoint matched the low one, but not
106            vice versa.
107    
108        (d) The processing of \h, \H, \v, and \ in character classes now makes use
109            of the new class addition function, using character lists defined as
110            macros alongside the case definitions of 20 above.
111    
112    
113    Version 8.31 06-July-2012
114    -------------------------
115    
116    1.  Fixing a wrong JIT test case and some compiler warnings.
117    
118    2.  Removed a bashism from the RunTest script.
119    
120    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
121        to unsigned type, result still unsigned" that was given by an MS compiler
122        on encountering the code "-sizeof(xxx)".
123    
124    4.  Partial matching support is added to the JIT compiler.
125    
126    5.  Fixed several bugs concerned with partial matching of items that consist
127        of more than one character:
128    
129        (a) /^(..)\1/ did not partially match "aba" because checking references was
130            done on an "all or nothing" basis. This also applied to repeated
131            references.
132    
133        (b) \R did not give a hard partial match if \r was found at the end of the
134            subject.
135    
136        (c) \X did not give a hard partial match after matching one or more
137            characters at the end of the subject.
138    
139        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
140            a partial match for the string "\r".
141    
142        (e) When newline was set to CRLF, the metacharacter "." did not recognize
143            a partial match for a CR character at the end of the subject string.
144    
145    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
146        running pcretest, the text "(JIT)" added to the output whenever JIT is
147        actually used to run the match.
148    
149    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
150        or /S+[+] with a digit between 1 and 7.
151    
152    8.  OP_NOT now supports any UTF character not just single-byte ones.
153    
154    9.  (*MARK) control verb is now supported by the JIT compiler.
155    
156    10. The command "./RunTest list" lists the available tests without actually
157        running any of them. (Because I keep forgetting what they all are.)
158    
159    11. Add PCRE_INFO_MAXLOOKBEHIND.
160    
161    12. Applied a (slightly modified) user-supplied patch that improves performance
162        when the heap is used for recursion (compiled with --disable-stack-for-
163        recursion). Instead of malloc and free for each heap frame each time a
164        logical recursion happens, frames are retained on a chain and re-used where
165        possible. This sometimes gives as much as 30% improvement.
166    
167    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
168        call.
169    
170    14. As documented, (*COMMIT) is now confined to within a positive assertion.
171    
172    15. It is now possible to link pcretest with libedit as an alternative to
173        libreadline.
174    
175    16. (*COMMIT) control verb is now supported by the JIT compiler.
176    
177    17. The Unicode data tables have been updated to Unicode 6.1.0.
178    
179    18. Added --file-list option to pcregrep.
180    
181    19. Added binary file support to pcregrep, including the -a, --binary-files,
182        -I, and --text options.
183    
184    20. The madvise function is renamed for posix_madvise for QNX compatibility
185        reasons. Fixed by Giuseppe D'Angelo.
186    
187    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
188    
189    22. Changed the option for creating symbolic links for 16-bit man pages from
190        -s to -sf so that re-installing does not cause issues.
191    
192    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
193    
194    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
195        matches in certain environments (the workspace was not being correctly
196        retained). Also added to pcre_dfa_exec() a simple plausibility check on
197        some of the workspace data at the beginning of a restart.
198    
199    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
200        was not doing so when it should - probably a typo introduced by SVN 528
201        (change 8.10/14).
202    
203    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
204        \w+ when the character tables indicated that \x{c4} was a word character.
205        There were several related cases, all because the tests for doing a table
206        lookup were testing for characters less than 127 instead of 255.
207    
208    27. If a pattern contains capturing parentheses that are not used in a match,
209        their slots in the ovector are set to -1. For those that are higher than
210        any matched groups, this happens at the end of processing. In the case when
211        there were back references that the ovector was too small to contain
212        (causing temporary malloc'd memory to be used during matching), and the
213        highest capturing number was not used, memory off the end of the ovector
214        was incorrectly being set to -1. (It was using the size of the temporary
215        memory instead of the true size.)
216    
217    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
218        ovector size, it uses memory at the end of the block that it has got.
219    
220    29. Check for an overlong MARK name and give an error at compile time. The
221        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
222    
223    30. JIT compiler update.
224    
225    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
226        Rill for the patch.
227    
228    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
229    
230    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
231    
232    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
233        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
234        was enabled.
235    
236    35. Improve JIT code generation for greedy plus quantifier.
237    
238    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
239        1 to "aa" instead of to an empty string. The bug affected repeated groups
240        that could potentially match an empty string.
241    
242    37. Optimizing single character iterators in JIT.
243    
244    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
245        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
246        codepoints that are too big for the mode are faulted, and in a UTF mode,
247        disallowed codepoints are also faulted.
248    
249    39. If PCRE was compiled with UTF support, in three places in the DFA
250        matcher there was code that should only have been obeyed in UTF mode, but
251        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
252        processing when bytes with values greater than 127 were present. In 16-bit
253        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
254        both cases the values are those that cannot be the first data item in a UTF
255        character. The three items that might have provoked this were recursions,
256        possessively repeated groups, and atomic groups.
257    
258    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
259        and pcregrep, because some OS require shared objects to be explicitly
260        passed to ld, causing the link step to fail if they are not.
261    
262    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
263        mode, patterns that started with \h* or \R* might be incorrectly matched.
264    
265    
266    Version 8.30 04-February-2012
267    -----------------------------
268    
269    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
270        name is defined in ctype.h.
271    
272    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
273        only in quite long subpatterns.
274    
275    3.  Removed the function pcre_info(), which has been obsolete and deprecated
276        since it was replaced by pcre_fullinfo() in February 2000.
277    
278    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
279        match a (*MARK), and the match failed at the start of the subject, a
280        reference to memory before the start of the subject could occur. This bug
281        was introduced by fix 17 of release 8.21.
282    
283    5.  A reference to an unset group with zero minimum repetition was giving
284        totally wrong answers (in non-JavaScript-compatibility mode). For example,
285        /(another)?(\1?)test/ matched against "hello world test". This bug was
286        introduced in release 8.13.
287    
288    6.  Add support for 16-bit character strings (a large amount of work involving
289        many changes and refactorings).
290    
291    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
292        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
293        from a file.
294    
295    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
296        rounding is not applied in this particular case).
297    
298    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
299        if they appear, or are escaped, in patterns.
300    
301    10. Get rid of a number of -Wunused-but-set-variable warnings.
302    
303    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
304        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
305        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
306        also returns the mark "x". This bug applied to capturing parentheses,
307        non-capturing parentheses, and atomic parentheses. It also applied to some
308        assertions.
309    
310    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
311        information out of configure.ac instead of relying on pcre.h.generic, which
312        is not stored in the repository.
313    
314    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
315        -lreadline.
316    
317    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
318    
319    15. Applied Graycode's patch to put the top-level frame on the stack rather
320        than the heap when not using the stack for recursion. This gives a
321        performance improvement in many cases when recursion is not deep.
322    
323    16. Experimental code added to "pcretest -C" to output the stack frame size.
324    
325    
326    Version 8.21 12-Dec-2011
327    ------------------------
328    
329    1.  Updating the JIT compiler.
330    
331    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
332        are added as well.
333    
334    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
335        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
336        calling _pcre_jit_exec. Some extra comments are added.
337    
338    4.  (*MARK) settings inside atomic groups that do not contain any capturing
339        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
340        was introduced by change 18 for 8.20.
341    
342    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
343        ECMA-262 standard.
344    
345    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
346        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
347        This bug was probably introduced by change 9 of 8.13.
348    
349    7.  While fixing 6 above, I noticed that a number of other items were being
350        incorrectly rejected as "not fixed length". This arose partly because newer
351        opcodes had not been added to the fixed-length checking code. I have (a)
352        corrected the bug and added tests for these items, and (b) arranged for an
353        error to occur if an unknown opcode is encountered while checking for fixed
354        length instead of just assuming "not fixed length". The items that were
355        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
356        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
357        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
358    
359    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
360        being incorrectly compiled and would have given unpredicatble results.
361    
362    9.  A possessively repeated subpattern with minimum repeat count greater than
363        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
364        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
365        the first (A) could occur when it should not.
366    
367    10. Add a cast and remove a redundant test from the code.
368    
369    11. JIT should use pcre_malloc/pcre_free for allocation.
370    
371    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
372        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
373        is anything other than /usr, -L is still shown).
374    
375    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
376    
377    14. Perl does not support \N without a following name in a [] class; PCRE now
378        also gives an error.
379    
380    15. If a forward reference was repeated with an upper limit of around 2000,
381        it caused the error "internal error: overran compiling workspace". The
382        maximum number of forward references (including repeats) was limited by the
383        internal workspace, and dependent on the LINK_SIZE. The code has been
384        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
385        the default depends on LINK_SIZE. There is a new upper limit (for safety)
386        of around 200,000 forward references. While doing this, I also speeded up
387        the filling in of repeated forward references.
388    
389    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
390        incorrectly expecting the subject to contain another "a" after the start.
391    
392    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
393        in the match, the SKIP should be ignored. This was not happening; instead
394        the SKIP was being treated as NOMATCH. For patterns such as
395        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
396        tested.
397    
398    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
399        now much more compatible with Perl, in particular in cases where the result
400        is a non-match for a non-anchored pattern. For example, if
401        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
402        "m", where previously it did not return a name. A side effect of this
403        change is that for partial matches, the last encountered mark name is
404        returned, as for non matches. A number of tests that were previously not
405        Perl-compatible have been moved into the Perl-compatible test files. The
406        refactoring has had the pleasing side effect of removing one argument from
407        the match() function, thus reducing its stack requirements.
408    
409    19. If the /S+ option was used in pcretest to study a pattern using JIT,
410        subsequent uses of /S (without +) incorrectly behaved like /S+.
411    
412    21. Retrieve executable code size support for the JIT compiler and fixing
413        some warnings.
414    
415    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
416        not work when the shorter character appeared right at the end of the
417        subject string.
418    
419    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
420        systems.
421    
422    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
423        output it when the /M option is used in pcretest.
424    
425    25. The CheckMan script was not being included in the distribution. Also, added
426        an explicit "perl" to run Perl scripts from the PrepareRelease script
427        because this is reportedly needed in Windows.
428    
429    26. If study data was being save in a file and studying had not found a set of
430        "starts with" bytes for the pattern, the data written to the file (though
431        never used) was taken from uninitialized memory and so caused valgrind to
432        complain.
433    
434    27. Updated RunTest.bat as provided by Sheri Pierce.
435    
436    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
437    
438    29. Computation of memory usage for the table of capturing group names was
439        giving an unnecessarily large value.
440    
441    
442    Version 8.20 21-Oct-2011
443    ------------------------
444    
445    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
446        a POSIX class. After further experiments with Perl, which convinced me that
447        Perl has bugs and confusions, a closing square bracket is no longer allowed
448        in a POSIX name. This bug also affected patterns with classes that started
449        with full stops.
450    
451    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
452        captured substring, but while checking the failing first alternative,
453        substring 1 is temporarily captured. If the output vector supplied to
454        pcre_exec() was not big enough for this capture, the yield of the function
455        was still zero ("insufficient space for captured substrings"). This cannot
456        be totally fixed without adding another stack variable, which seems a lot
457        of expense for a edge case. However, I have improved the situation in cases
458        such as /(a)(b)x|abc/ matched against "abc", where the return code
459        indicates that fewer than the maximum number of slots in the ovector have
460        been set.
461    
462    3.  Related to (2) above: when there are more back references in a pattern than
463        slots in the output vector, pcre_exec() uses temporary memory during
464        matching, and copies in the captures as far as possible afterwards. It was
465        using the entire output vector, but this conflicts with the specification
466        that only 2/3 is used for passing back captured substrings. Now it uses
467        only the first 2/3, for compatibility. This is, of course, another edge
468        case.
469    
470    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
471        main code base, and can be used by building with --enable-jit. When this is
472        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
473        runtime --no-jit option is given.
474    
475    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
476        ovector, the return from the function was zero, implying that there were
477        other matches that did not fit. The correct "exactly full" value is now
478        returned.
479    
480    6.  If a subpattern that was called recursively or as a subroutine contained
481        (*PRUNE) or any other control that caused it to give a non-standard return,
482        invalid errors such as "Error -26 (nested recursion at the same subject
483        position)" or even infinite loops could occur.
484    
485    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
486        computing the minimum length on reaching *ACCEPT, and so ended up with the
487        wrong value of 1 rather than 0. Further investigation indicates that
488        computing a minimum subject length in the presence of *ACCEPT is difficult
489        (think back references, subroutine calls), and so I have changed the code
490        so that no minimum is registered for a pattern that contains *ACCEPT.
491    
492    8.  If (*THEN) was present in the first (true) branch of a conditional group,
493        it was not handled as intended. [But see 16 below.]
494    
495    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
496        Sheri Pierce.
497    
498    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
499        the first byte in a match must be "a".
500    
501    11. Change 17 for 8.13 increased the recursion depth for patterns like
502        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
503        pattern contains any instances of (*THEN). If it does not, the old
504        optimizations are restored. It would be nice to do this on a per-group
505        basis, but at the moment that is not feasible.
506    
507    12. In some environments, the output of pcretest -C is CRLF terminated. This
508        broke RunTest's code that checks for the link size. A single white space
509        character after the value is now allowed for.
510    
511    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
512        For "fr", it uses the Windows-specific input and output files.
513    
514    14. If (*THEN) appeared in a group that was called recursively or as a
515        subroutine, it did not work as intended. [But see next item.]
516    
517    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
518        pattern fragments (but not containing any | characters). If A and B are
519        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
520        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
521        D. In other words, Perl considers parentheses that do not contain any |
522        characters to be part of a surrounding alternative, whereas PCRE was
523        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
524        differently. PCRE now behaves in the same way as Perl, except in the case
525        of subroutine/recursion calls such as (?1) which have in any case always
526        been different (but PCRE had them first :-).
527    
528    16. Related to 15 above: Perl does not treat the | in a conditional group as
529        creating alternatives. Such a group is treated in the same way as an
530        ordinary group without any | characters when processing (*THEN). PCRE has
531        been changed to match Perl's behaviour.
532    
533    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
534        RunGrepTest script failed.
535    
536    18. Change 22 for version 13 caused atomic groups to use more stack. This is
537        inevitable for groups that contain captures, but it can lead to a lot of
538        stack use in large patterns. The old behaviour has been restored for atomic
539        groups that do not contain any capturing parentheses.
540    
541    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
542        suppress the check for a minimum subject length at run time. (If it was
543        given to pcre_exec() or pcre_dfa_exec() it did work.)
544    
545    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
546        fail to work when decoding hex characters in data strings in EBCDIC
547        environments.
548    
549    21. It appears that in at least one Mac OS environment, the isxdigit() function
550        is implemented as a macro that evaluates to its argument more than once,
551        contravening the C 90 Standard (I haven't checked a later standard). There
552        was an instance in pcretest which caused it to go wrong when processing
553        \x{...} escapes in subject strings. The has been rewritten to avoid using
554        things like p++ in the argument of isxdigit().
555    
556    
557    Version 8.13 16-Aug-2011
558    ------------------------
559    
560    1.  The Unicode data tables have been updated to Unicode 6.0.0.
561    
562    2.  Two minor typos in pcre_internal.h have been fixed.
563    
564    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
565        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
566        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
567    
568    4.  There were a number of related bugs in the code for matching backrefences
569        caselessly in UTF-8 mode when codes for the characters concerned were
570        different numbers of bytes. For example, U+023A and U+2C65 are an upper
571        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
572        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
573        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
574        2-byte code at the end of the subject (it thought there wasn't enough data
575        left).
576    
577    5.  Comprehensive information about what went wrong is now returned by
578        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
579        as the output vector has at least 2 elements. The offset of the start of
580        the failing character and a reason code are placed in the vector.
581    
582    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
583        now returned is for the first byte of the failing character, instead of the
584        last byte inspected. This is an incompatible change, but I hope it is small
585        enough not to be a problem. It makes the returned offset consistent with
586        pcre_exec() and pcre_dfa_exec().
587    
588    7.  pcretest now gives a text phrase as well as the error number when
589        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
590        failure, the offset and reason code are output.
591    
592    8.  When \R was used with a maximizing quantifier it failed to skip backwards
593        over a \r\n pair if the subsequent match failed. Instead, it just skipped
594        back over a single character (\n). This seems wrong (because it treated the
595        two characters as a single entity when going forwards), conflicts with the
596        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
597        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
598        has been changed.
599    
600    9.  Some internal refactoring has changed the processing so that the handling
601        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
602        time (the PCRE_DOTALL option was changed this way some time ago: version
603        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
604        which was always a bit of a fudge. It also means that there is one less
605        argument for the match() function, which reduces its stack requirements
606        slightly. This change also fixes an incompatibility with Perl: the pattern
607        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
608    
609    10. More internal refactoring has drastically reduced the number of recursive
610        calls to match() for possessively repeated groups such as (abc)++ when
611        using pcre_exec().
612    
613    11. While implementing 10, a number of bugs in the handling of groups were
614        discovered and fixed:
615    
616        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
617        (a|)*(?1) gave a compile-time internal error.
618        ((a|)+)+  did not notice that the outer group could match an empty string.
619        (^a|^)+   was not marked as anchored.
620        (.*a|.*)+ was not marked as matching at start or after a newline.
621    
622    12. Yet more internal refactoring has removed another argument from the match()
623        function. Special calls to this function are now indicated by setting a
624        value in a variable in the "match data" data block.
625    
626    13. Be more explicit in pcre_study() instead of relying on "default" for
627        opcodes that mean there is no starting character; this means that when new
628        ones are added and accidentally left out of pcre_study(), testing should
629        pick them up.
630    
631    14. The -s option of pcretest has been documented for ages as being an old
632        synonym of -m (show memory usage). I have changed it to mean "force study
633        for every regex", that is, assume /S for every regex. This is similar to -i
634        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
635        using it. It makes it easier to run collections of tests with and without
636        study enabled, and thereby test pcre_study() more easily. All the standard
637        tests are now run with and without -s (but some patterns can be marked as
638        "never study" - see 20 below).
639    
640    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
641        restoration of the capturing data to the outer values was not happening
642        correctly.
643    
644    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
645        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
646        pattern had matched an empty string, and so incorrectly returned a no
647        match.
648    
649    17. There was optimizing code for the last branch of non-capturing parentheses,
650        and also for the obeyed branch of a conditional subexpression, which used
651        tail recursion to cut down on stack usage. Unfortunately, now that there is
652        the possibility of (*THEN) occurring in these branches, tail recursion is
653        no longer possible because the return has to be checked for (*THEN). These
654        two optimizations have therefore been removed. [But see 8.20/11 above.]
655    
656    18. If a pattern containing \R was studied, it was assumed that \R always
657        matched two bytes, thus causing the minimum subject length to be
658        incorrectly computed because \R can also match just one byte.
659    
660    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
661        was incorrectly computed.
662    
663    20. If /S is present twice on a test pattern in pcretest input, it now
664        *disables* studying, thereby overriding the use of -s on the command line
665        (see 14 above). This is necessary for one or two tests to keep the output
666        identical in both cases.
667    
668    21. When (*ACCEPT) was used in an assertion that matched an empty string and
669        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
670    
671    22. When an atomic group that contained a capturing parenthesis was
672        successfully matched, but the branch in which it appeared failed, the
673        capturing was not being forgotten if a higher numbered group was later
674        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
675        group 1 to "a", when in fact it should be unset. This applied to multi-
676        branched capturing and non-capturing groups, repeated or not, and also to
677        positive assertions (capturing in negative assertions does not happen
678        in PCRE) and also to nested atomic groups.
679    
680    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
681        subject after a captured substring, to make it easier to tell which of a
682        number of identical substrings has been captured.
683    
684    24. The way atomic groups are processed by pcre_exec() has been changed so that
685        if they are repeated, backtracking one repetition now resets captured
686        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
687        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
688        "aaa". Previously, it would have been "a". As part of this code
689        refactoring, the way recursive calls are handled has also been changed.
690    
691    25. If an assertion condition captured any substrings, they were not passed
692        back unless some other capturing happened later. For example, if
693        (?(?=(a))a) was matched against "a", no capturing was returned.
694    
695    26. When studying a pattern that contained subroutine calls or assertions,
696        the code for finding the minimum length of a possible match was handling
697        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
698        group 1 called group 2 while simultaneously a separate group 2 called group
699        1). A stack overflow occurred in this case. I have fixed this by limiting
700        the recursion depth to 10.
701    
702    27. Updated RunTest.bat in the distribution to the version supplied by Tom
703        Fortmann. This supports explicit test numbers on the command line, and has
704        argument validation and error reporting.
705    
706    28. An instance of \X with an unlimited repeat could fail if at any point the
707        first character it looked at was a mark character.
708    
709    29. Some minor code refactoring concerning Unicode properties and scripts
710        should reduce the stack requirement of match() slightly.
711    
712    30. Added the '=' option to pcretest to check the setting of unused capturing
713        slots at the end of the pattern, which are documented as being -1, but are
714        not included in the return count.
715    
716    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
717        compiled something random. Now it gives a compile-time error (as does
718        Perl).
719    
720    32. A *MARK encountered during the processing of a positive assertion is now
721        recorded and passed back (compatible with Perl).
722    
723    33. If --only-matching or --colour was set on a pcregrep call whose pattern
724        had alternative anchored branches, the search for a second match in a line
725        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
726        matched the line "0102" twice. The same bug affected patterns that started
727        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
728        twice.
729    
730    34. Previously, PCRE did not allow quantification of assertions. However, Perl
731        does, and because of capturing effects, quantifying parenthesized
732        assertions may at times be useful. Quantifiers are now allowed for
733        parenthesized assertions.
734    
735    35. A minor code tidy in pcre_compile() when checking options for \R usage.
736    
737    36. \g was being checked for fancy things in a character class, when it should
738        just be a literal "g".
739    
740    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
741        appearance of a nested POSIX class supersedes an apparent external class.
742        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
743        unescaped square brackets may also appear as part of class names. For
744        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
745        more like Perl. (But see 8.20/1 above.)
746    
747    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
748        was because it thought it was \N{name}, which is not supported).
749    
750    39. Add minix to OS list not supporting the -S option in pcretest.
751    
752    40. PCRE tries to detect cases of infinite recursion at compile time, but it
753        cannot analyze patterns in sufficient detail to catch mutual recursions
754        such as ((?1))((?2)). There is now a runtime test that gives an error if a
755        subgroup is called recursively as a subpattern for a second time at the
756        same position in the subject string. In previous releases this might have
757        been caught by the recursion limit, or it might have run out of stack.
758    
759    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
760        happen only once. PCRE was, however incorrectly giving a compile time error
761        "recursive call could loop indefinitely" because it cannot analyze the
762        pattern in sufficient detail. The compile time test no longer happens when
763        PCRE is compiling a conditional subpattern, but actual runaway loops are
764        now caught at runtime (see 40 above).
765    
766    42. It seems that Perl allows any characters other than a closing parenthesis
767        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
768        has been changed to be the same.
769    
770    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
771        as not to get warnings when autogen.sh is called. Also changed
772        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
773    
774    44. To help people who use pcregrep to scan files containing exceedingly long
775        lines, the following changes have been made:
776    
777        (a) The default value of the buffer size parameter has been increased from
778            8K to 20K. (The actual buffer used is three times this size.)
779    
780        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
781            PCRE is built.
782    
783        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
784            to be set at run time.
785    
786        (d) Numerical values in pcregrep options can be followed by K or M, for
787            example --buffer-size=50K.
788    
789        (e) If a line being scanned overflows pcregrep's buffer, an error is now
790            given and the return code is set to 2.
791    
792    45. Add a pointer to the latest mark to the callout data block.
793    
794    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
795        partial match of an empty string instead of no match. This was specific to
796        the use of ".".
797    
798    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
799        complete match instead of a partial match. This bug was dependent on both
800        the PCRE_UTF8 and PCRE_DOTALL options being set.
801    
802    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
803        starting byte set, because \b was not being ignored.
804    
805    
806    Version 8.12 15-Jan-2011
807    ------------------------
808    
809    1.  Fixed some typos in the markup of the man pages, and wrote a script that
810        checks for such things as part of the documentation building process.
811    
812    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
813        --match-limit and --recursion-limit options (added for 8.11). In
814        particular, this made one of the standard tests fail. (The integer value
815        went into the wrong half of a long int.)
816    
817    3.  If the --colour option was given to pcregrep with -v (invert match), it
818        did strange things, either producing crazy output, or crashing. It should,
819        of course, ignore a request for colour when reporting lines that do not
820        match.
821    
822    4.  Another pcregrep bug caused similar problems if --colour was specified with
823        -M (multiline) and the pattern match finished with a line ending.
824    
825    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
826        matched in multiline mode, the following line was shown as part of the
827        match. This seems wrong, so I have changed it.
828    
829    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
830        the check for further matches in the same line (so they could be coloured)
831        to overrun the end of the current line. If another match was found, it was
832        incorrectly shown (and then shown again when found in the next line).
833    
834    7.  If pcregrep was compiled under Windows, there was a reference to the
835        function pcregrep_exit() before it was defined. I am assuming this was
836        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
837        reported by a user. I've moved the definition above the reference.
838    
839    
840    Version 8.11 10-Dec-2010
841    ------------------------
842    
843    1.  (*THEN) was not working properly if there were untried alternatives prior
844        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
845        backtracked to try for "b" instead of moving to the next alternative branch
846        at the same level (in this case, to look for "c"). The Perl documentation
847        is clear that when (*THEN) is backtracked onto, it goes to the "next
848        alternative in the innermost enclosing group".
849    
850    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
851        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
852        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
853        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
854        (*THEN).
855    
856    3.  If \s appeared in a character class, it removed the VT character from
857        the class, even if it had been included by some previous item, for example
858        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
859        of \s, but is part of the POSIX "space" class.)
860    
861    4.  A partial match never returns an empty string (because you can always
862        match an empty string at the end of the subject); however the checking for
863        an empty string was starting at the "start of match" point. This has been
864        changed to the "earliest inspected character" point, because the returned
865        data for a partial match starts at this character. This means that, for
866        example, /(?<=abc)def/ gives a partial match for the subject "abc"
867        (previously it gave "no match").
868    
869    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
870        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
871        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
872        has an implication that the given string is incomplete (because a partial
873        match is preferred over a full match). For this reason, these items now
874        give a partial match in this situation. [Aside: previously, the one case
875        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
876        match rather than a full match, which was wrong by the old rules, but is
877        now correct.]
878    
879    6.  There was a bug in the handling of #-introduced comments, recognized when
880        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
881        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
882        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
883        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
884        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
885        places in pcre_compile().
886    
887    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
888        comments when looking ahead for named forward references to subpatterns,
889        the only newline sequence it recognized was NL. It now handles newlines
890        according to the set newline convention.
891    
892    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
893        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
894        cater for a lack of strerror(). These oversights have been fixed.
895    
896    9.  Added --match-limit and --recursion-limit to pcregrep.
897    
898    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
899    
900    11. When the -o option was used, pcregrep was setting a return code of 1, even
901        when matches were found, and --line-buffered was not being honoured.
902    
903    12. Added an optional parentheses number to the -o and --only-matching options
904        of pcregrep.
905    
906    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
907        can match an empty string. The code to do it in pcretest and pcredemo
908        needed fixing:
909    
910        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
911            only one byte after an empty string match just before CRLF (this case
912            just got forgotten; "any" and "anycrlf" were OK).
913    
914        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
915            mode when an empty string match preceded an ASCII character followed by
916            a non-ASCII character. (The code for advancing by one character rather
917            than one byte was nonsense.)
918    
919        (c) The pcredemo.c sample program did not have any code at all to handle
920            the cases when CRLF is a valid newline sequence.
921    
922    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
923        as a starting offset was within the subject string. There is now a new
924        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
925        negative or greater than the length of the string. In order to test this,
926        pcretest is extended to allow the setting of negative starting offsets.
927    
928    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
929        starting offset points to the beginning of a UTF-8 character was
930        unnecessarily clumsy. I tidied it up.
931    
932    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
933        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
934    
935    17. Nobody had reported that the --include_dir option, which was added in
936        release 7.7 should have been called --include-dir (hyphen, not underscore)
937        for compatibility with GNU grep. I have changed it to --include-dir, but
938        left --include_dir as an undocumented synonym, and the same for
939        --exclude-dir, though that is not available in GNU grep, at least as of
940        release 2.5.4.
941    
942    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
943        characters from a string of bytes) have been redefined so as not to use
944        loops, in order to improve performance in some environments. At the same
945        time, I abstracted some of the common code into auxiliary macros to save
946        repetition (this should not affect the compiled code).
947    
948    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
949        compile-time error is now given if \c is not followed by an ASCII
950        character, that is, a byte less than 128. (In EBCDIC mode, the code is
951        different, and any byte value is allowed.)
952    
953    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
954        START_OPTIMIZE option, which is now allowed at compile time - but just
955        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
956        to pcregrep and other applications that have no direct access to PCRE
957        options. The new /Y option in pcretest sets this option when calling
958        pcre_compile().
959    
960    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
961        back references. Groups containing recursive back references were forced to
962        be atomic by that change, but in the case of named groups, the amount of
963        memory required was incorrectly computed, leading to "Failed: internal
964        error: code overflow". This has been fixed.
965    
966    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
967        pcretest.c, to avoid build problems in some Borland environments.
968    
969    
970    Version 8.10 25-Jun-2010
971    ------------------------
972    
973    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
974        THEN.
975    
976    2.  (*ACCEPT) was not working when inside an atomic group.
977    
978    3.  Inside a character class, \B is treated as a literal by default, but
979        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
980        causes the error). The code is unchanged, but I tidied the documentation.
981    
982    4.  Inside a character class, PCRE always treated \R and \X as literals,
983        whereas Perl faults them if its -w option is set. I have changed PCRE so
984        that it faults them when PCRE_EXTRA is set.
985    
986    5.  Added support for \N, which always matches any character other than
987        newline. (It is the same as "." when PCRE_DOTALL is not set.)
988    
989    6.  When compiling pcregrep with newer versions of gcc which may have
990        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
991        declared with attribute warn_unused_result" were given. Just casting the
992        result to (void) does not stop the warnings; a more elaborate fudge is
993        needed. I've used a macro to implement this.
994    
995    7.  Minor change to pcretest.c to avoid a compiler warning.
996    
997    8.  Added four artifical Unicode properties to help with an option to make
998        \s etc use properties (see next item). The new properties are: Xan
999        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1000    
1001    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1002        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1003        this option. Modified pcretest to add /W to test this facility. Added
1004        REG_UCP to make it available via the POSIX interface.
1005    
1006    10. Added --line-buffered to pcregrep.
1007    
1008    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1009        studied, and the match started with a letter with a code point greater than
1010        127 whose first byte was different to the first byte of the other case of
1011        the letter, the other case of this starting letter was not recognized
1012        (#976).
1013    
1014    12. If a pattern that was studied started with a repeated Unicode property
1015        test, for example, \p{Nd}+, there was the theoretical possibility of
1016        setting up an incorrect bitmap of starting bytes, but fortunately it could
1017        not have actually happened in practice until change 8 above was made (it
1018        added property types that matched character-matching opcodes).
1019    
1020    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1021        possible starting bytes for non-anchored patterns.
1022    
1023    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1024        \R, and also a number of cases that involve Unicode properties, both
1025        explicit and implicit when PCRE_UCP is set.
1026    
1027    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1028        input, it could crash or give wrong results if characters with values
1029        greater than 0xc0 were present in the subject string. (Detail: it assumed
1030        UTF-8 input when processing these items.)
1031    
1032    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1033        size_t is 64-bit (#991).
1034    
1035    17. Added a check for running out of memory when PCRE is compiled with
1036        --disable-stack-for-recursion (#990).
1037    
1038    18. If the last data line in a file for pcretest does not have a newline on
1039        the end, a newline was missing in the output.
1040    
1041    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1042        less than 128) in its various bitmaps. However, there is a facility for
1043        generating tables according to the current locale when PCRE is compiled. It
1044        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1045        characters, are recognized by isspace() and therefore were getting set in
1046        these tables, and indeed these tables seem to approximate to ISO 8859. This
1047        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1048        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1049        which of course cannot start UTF-8 characters. I have changed the code so
1050        that only real ASCII characters (less than 128) and the correct starting
1051        bytes for UTF-8 encodings are set for characters greater than 127 when in
1052        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1053        altogether.)
1054    
1055    20. Added the /T option to pcretest so as to be able to run tests with non-
1056        standard character tables, thus making it possible to include the tests
1057        used for 19 above in the standard set of tests.
1058    
1059    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1060        reference to a subpattern the other side of a comment that contains an
1061        opening parenthesis caused either an internal compiling error, or a
1062        reference to the wrong subpattern.
1063    
1064    
1065    Version 8.02 19-Mar-2010
1066    ------------------------
1067    
1068    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1069    
1070    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1071        configured.
1072    
1073    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1074        original author of that file, following a query about its status.
1075    
1076    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1077        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1078    
1079    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1080        quantifier applied to a forward-referencing subroutine call, could compile
1081        incorrect code or give the error "internal error: previously-checked
1082        referenced subpattern not found".
1083    
1084    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1085        variables to point to external functions. For these systems, therefore,
1086        pcre_malloc etc. are now initialized to local functions that call the
1087        relevant global functions.
1088    
1089    7.  There were two entries missing in the vectors called coptable and poptable
1090        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1091        I've fixed the data, and added a kludgy way of testing at compile time that
1092        the lengths are correct (equal to the number of opcodes).
1093    
1094    8.  Following on from 7, I added a similar kludge to check the length of the
1095        eint vector in pcreposix.c.
1096    
1097    9.  Error texts for pcre_compile() are held as one long string to avoid too
1098        much relocation at load time. To find a text, the string is searched,
1099        counting zeros. There was no check for running off the end of the string,
1100        which could happen if a new error number was added without updating the
1101        string.
1102    
1103    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1104    
1105    11. \K was not working if it appeared in an atomic group or in a group that
1106        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1107        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1108        the assertion is positive, but not if it is negative.
1109    
1110    12. Change 11 fortuitously reduced the size of the stack frame used in the
1111        "match()" function of pcre_exec.c by one pointer. Forthcoming
1112        implementation of support for (*MARK) will need an extra pointer on the
1113        stack; I have reserved it now, so that the stack frame size does not
1114        decrease.
1115    
1116    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1117        item in branch that calls a recursion is a subroutine call - as in the
1118        second branch in the above example - was incorrectly given the compile-
1119        time error "recursive call could loop indefinitely" because pcre_compile()
1120        was not correctly checking the subroutine for matching a non-empty string.
1121    
1122    14. The checks for overrunning compiling workspace could trigger after an
1123        overrun had occurred. This is a "should never occur" error, but it can be
1124        triggered by pathological patterns such as hundreds of nested parentheses.
1125        The checks now trigger 100 bytes before the end of the workspace.
1126    
1127    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1128    
1129    
1130    Version 8.01 19-Jan-2010
1131    ------------------------
1132    
1133    1.  If a pattern contained a conditional subpattern with only one branch (in
1134        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1135        computed the wrong minimum data length (which is of course zero for such
1136        subpatterns). This could cause incorrect "no match" results.
1137    
1138    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1139        the pattern is reset in the first branch, pcre_compile() failed with
1140        "internal error: code overflow at offset...". This happened only when
1141        the reset was to the original external option setting. (An optimization
1142        abstracts leading options settings into an external setting, which was the
1143        cause of this.)
1144    
1145    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1146        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1147        assertion pattern did not match (meaning that the assertion was true), it
1148        was incorrectly treated as false if the SKIP had been reached during the
1149        matching. This also applied to assertions used as conditions.
1150    
1151    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1152        assertion subpattern, including such a pattern used as a condition,
1153        unpredictable results occurred, instead of the error return
1154        PCRE_ERROR_DFA_UITEM.
1155    
1156    5.  The C++ GlobalReplace function was not working like Perl for the special
1157        situation when an empty string is matched. It now does the fancy magic
1158        stuff that is necessary.
1159    
1160    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1161        removed. (These were left over from very, very early versions of PCRE.)
1162    
1163    7.  Some cosmetic changes to the code to make life easier when compiling it
1164        as part of something else:
1165    
1166        (a) Change DEBUG to PCRE_DEBUG.
1167    
1168        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1169            called "current" as "current_branch", to prevent a collision with the
1170            Linux macro when compiled as a kernel module.
1171    
1172        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1173            prevent a collision with the Linux macro when compiled as a kernel
1174            module.
1175    
1176    8.  In pcre_compile() there are some checks for integer overflows that used to
1177        cast potentially large values to (double). This has been changed to that
1178        when building, a check for int64_t is made, and if it is found, it is used
1179        instead, thus avoiding the use of floating point arithmetic. (There is no
1180        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1181        double.
1182    
1183    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1184        2005 (difference between two addresses compared to an unsigned value).
1185    
1186    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1187        custom one, because of the following reported problem in Windows:
1188    
1189          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1190              under Win32.
1191          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1192              therefore missing the function definition.
1193          - The compiler thus generates a "C" signature for the test function.
1194          - The linker fails to find the "C" function.
1195          - PCRE fails to configure if asked to do so against libbz2.
1196    
1197    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1198        messages were output:
1199    
1200          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1201          rerunning libtoolize, to keep the correct libtool macros in-tree.
1202          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1203    
1204        I have done both of these things.
1205    
1206    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1207        most of the time, it *can* run out if it is given a pattern that contains a
1208        runaway infinite recursion. I updated the discussion in the pcrestack man
1209        page.
1210    
1211    13. Now that we have gone to the x.xx style of version numbers, the minor
1212        version may start with zero. Using 08 or 09 is a bad idea because users
1213        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1214        interpreted as invalid octal numbers. I've updated the previous comment in
1215        configure.ac, and also added a check that gives an error if 08 or 09 are
1216        used.
1217    
1218    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1219        causing partial matching to fail when the end of the subject matched \W
1220        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1221    
1222    15. There were some discrepancies between the declarations in pcre_internal.h
1223        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1224        their definitions. The declarations used "const uschar *" and the
1225        definitions used USPTR. Even though USPTR is normally defined as "const
1226        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1227        reported that: "This difference in casting confuses some C++ compilers, for
1228        example, SunCC recognizes above declarations as different functions and
1229        generates broken code for hbpcre." I have changed the declarations to use
1230        USPTR.
1231    
1232    16. GNU libtool is named differently on some systems. The autogen.sh script now
1233        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1234        (FreeBSD).
1235    
1236    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1237        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1238        comment: "Figure out how to create a longlong from a string: strtoll and
1239        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1240        instance, but it only takes 2 args instead of 3!"
1241    
1242    18. A subtle bug concerned with back references has been fixed by a change of
1243        specification, with a corresponding code fix. A pattern such as
1244        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1245        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1246        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1247        same bug. Such groups have to be quantified to be useful, or contained
1248        inside another quantified group. (If there's no repetition, the reference
1249        can never match.) The problem arises because, having left the group and
1250        moved on to the rest of the pattern, a later failure that backtracks into
1251        the group uses the captured value from the final iteration of the group
1252        rather than the correct earlier one. I have fixed this in PCRE by forcing
1253        any group that contains a reference to itself to be an atomic group; that
1254        is, there cannot be any backtracking into it once it has completed. This is
1255        similar to recursive and subroutine calls.
1256    
1257    
1258    Version 8.00 19-Oct-09
1259    ----------------------
1260    
1261    1.  The table for translating pcre_compile() error codes into POSIX error codes
1262        was out-of-date, and there was no check on the pcre_compile() error code
1263        being within the table. This could lead to an OK return being given in
1264        error.
1265    
1266    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1267        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1268        in a Windows environment.
1269    
1270    3.  The pcregrep --count option prints the count for each file even when it is
1271        zero, as does GNU grep. However, pcregrep was also printing all files when
1272        --files-with-matches was added. Now, when both options are given, it prints
1273        counts only for those files that have at least one match. (GNU grep just
1274        prints the file name in this circumstance, but including the count seems
1275        more useful - otherwise, why use --count?) Also ensured that the
1276        combination -clh just lists non-zero counts, with no names.
1277    
1278    4.  The long form of the pcregrep -F option was incorrectly implemented as
1279        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1280        but it seems right to fix it, and I didn't think it was worth preserving
1281        the old behaviour.
1282    
1283    5.  The command line items --regex=pattern and --regexp=pattern were not
1284        recognized by pcregrep, which required --regex pattern or --regexp pattern
1285        (with a space rather than an '='). The man page documented the '=' forms,
1286        which are compatible with GNU grep; these now work.
1287    
1288    6.  No libpcreposix.pc file was created for pkg-config; there was just
1289        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1290    
1291    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1292        when UCP support is not needed, by modifying the Python script that
1293        generates it from Unicode data files. This should not matter if the module
1294        is correctly used as a library, but I received one complaint about 50K of
1295        unwanted data. My guess is that the person linked everything into his
1296        program rather than using a library. Anyway, it does no harm.
1297    
1298    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1299        was a minimum greater than 1 for a wide character in a possessive
1300        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1301        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1302        character. Chaos in the form of incorrect output or a compiling loop could
1303        result.
1304    
1305    9.  The restrictions on what a pattern can contain when partial matching is
1306        requested for pcre_exec() have been removed. All patterns can now be
1307        partially matched by this function. In addition, if there are at least two
1308        slots in the offset vector, the offset of the earliest inspected character
1309        for the match and the offset of the end of the subject are set in them when
1310        PCRE_ERROR_PARTIAL is returned.
1311    
1312    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1313        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1314        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1315        and may be more useful for multi-segment matching.
1316    
1317    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1318        used to be given if ever the end of the subject was reached; now it is
1319        given only if matching could not proceed because another character was
1320        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1321        string "Z", which now yields "no match" instead of "partial match". In the
1322        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1323        final character ended with (*FAIL).
1324    
1325    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1326        if the pattern had a "must contain" character that was already found in the
1327        earlier partial match, unless partial matching was again requested. For
1328        example, with the pattern /dog.(body)?/, the "must contain" character is
1329        "g". If the first part-match was for the string "dog", restarting with
1330        "sbody" failed. This bug has been fixed.
1331    
1332    13. The string returned by pcre_dfa_exec() after a partial match has been
1333        changed so that it starts at the first inspected character rather than the
1334        first character of the match. This makes a difference only if the pattern
1335        starts with a lookbehind assertion or \b or \B (\K is not supported by
1336        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1337        matching functions compatible, and I think it's the right thing to do.
1338    
1339    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1340        so that the demonstration program is easily available in environments where
1341        PCRE has not been installed from source.
1342    
1343    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1344        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1345        library.
1346    
1347    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1348        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1349        is not the first non-POSIX option to be added. Clearly some people find
1350        these options useful.
1351    
1352    17. If a caller to the POSIX matching function regexec() passes a non-zero
1353        value for nmatch with a NULL value for pmatch, the value of
1354        nmatch is forced to zero.
1355    
1356    18. RunGrepTest did not have a test for the availability of the -u option of
1357        the diff command, as RunTest does. It now checks in the same way as
1358        RunTest, and also checks for the -b option.
1359    
1360    19. If an odd number of negated classes containing just a single character
1361        interposed, within parentheses, between a forward reference to a named
1362        subpattern and the definition of the subpattern, compilation crashed with
1363        an internal error, complaining that it could not find the referenced
1364        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1365        [The bug was that it was starting one character too far in when skipping
1366        over the character class, thus treating the ] as data rather than
1367        terminating the class. This meant it could skip too much.]
1368    
1369    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1370        /g option in pcretest when the pattern contains \K, which makes it possible
1371        to have an empty string match not at the start, even when the pattern is
1372        anchored. Updated pcretest and pcredemo to use this option.
1373    
1374    21. If the maximum number of capturing subpatterns in a recursion was greater
1375        than the maximum at the outer level, the higher number was returned, but
1376        with unset values at the outer level. The correct (outer level) value is
1377        now given.
1378    
1379    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1380        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1381        make it do so. The string so far is captured, making this feature
1382        compatible with Perl.
1383    
1384    23. The tests have been re-organized, adding tests 11 and 12, to make it
1385        possible to check the Perl 5.10 features against Perl 5.10.
1386    
1387    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1388        pattern matches a fixed length string. PCRE did not allow this; now it
1389        does. Neither allows recursion.
1390    
1391    25. I finally figured out how to implement a request to provide the minimum
1392        length of subject string that was needed in order to match a given pattern.
1393        (It was back references and recursion that I had previously got hung up
1394        on.) This code has now been added to pcre_study(); it finds a lower bound
1395        to the length of subject needed. It is not necessarily the greatest lower
1396        bound, but using it to avoid searching strings that are too short does give
1397        some useful speed-ups. The value is available to calling programs via
1398        pcre_fullinfo().
1399    
1400    26. While implementing 25, I discovered to my embarrassment that pcretest had
1401        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1402        study optimizations had never been tested with that matching function.
1403        Oops. What is worse, even when it was passed study data, there was a bug in
1404        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1405        were also very few tests of studied patterns with pcre_dfa_exec().
1406    
1407    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1408        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1409        on the other side of the coin, they are no longer allowed to have different
1410        names, because these cannot be distinguished in PCRE, and this has caused
1411        confusion. (This is a difference from Perl.)
1412    
1413    28. When duplicate subpattern names are present (necessarily with different
1414        numbers, as required by 27 above), and a test is made by name in a
1415        conditional pattern, either for a subpattern having been matched, or for
1416        recursion in such a pattern, all the associated numbered subpatterns are
1417        tested, and the overall condition is true if the condition is true for any
1418        one of them. This is the way Perl works, and is also more like the way
1419        testing by number works.
1420    
1421    
1422    Version 7.9 11-Apr-09
1423    ---------------------
1424    
1425    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1426        (pcretest), all targets were linked against these libraries. This included
1427        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1428        libraries. This caused unwanted dependencies to be created. This problem
1429        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1430        pcretest is linked with readline.
1431    
1432    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1433        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1434        moved outside it again, because FALSE and TRUE are already defined in AIX,
1435        but BOOL is not.
1436    
1437    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1438        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1439    
1440    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1441        hyphens) following file names and line numbers when outputting matching
1442        lines. This is not true; no spaces are inserted. I have also clarified the
1443        wording for the --colour (or --color) option.
1444    
1445    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1446        was not coloured; this is different to GNU grep, so I have changed it to be
1447        the same.
1448    
1449    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1450        each matching line was coloured. Now it goes on to look for further matches
1451        of any of the test patterns, which is the same behaviour as GNU grep.
1452    
1453    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1454        doesn't make sense to accept an empty string match in pcregrep, so I have
1455        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1456        seems to be how GNU grep behaves.
1457    
1458    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1459        start or after a newline", because the conditional assertion was not being
1460        correctly handled. The rule now is that both the assertion and what follows
1461        in the first alternative must satisfy the test.
1462    
1463    9.  If auto-callout was enabled in a pattern with a conditional group whose
1464        condition was an assertion, PCRE could crash during matching, both with
1465        pcre_exec() and pcre_dfa_exec().
1466    
1467    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1468        used for matching.
1469    
1470    11. Unicode property support in character classes was not working for
1471        characters (bytes) greater than 127 when not in UTF-8 mode.
1472    
1473    12. Added the -M command line option to pcretest.
1474    
1475    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1476    
1477    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1478    
1479    16. Added comments and documentation about mis-use of no_arg in the C++
1480        wrapper.
1481    
1482    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1483        from Martin Jerabek that uses macro names for all relevant character and
1484        string constants.
1485    
1486    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1487        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1488        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1489        these, but not everybody uses configure.
1490    
1491    19. A conditional group that had only one branch was not being correctly
1492        recognized as an item that could match an empty string. This meant that an
1493        enclosing group might also not be so recognized, causing infinite looping
1494        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1495        with the subject "ab", where knowledge that the repeated group can match
1496        nothing is needed in order to break the loop.
1497    
1498    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1499        exec(), but without supplying a callout function, matching went wrong.
1500    
1501    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1502        leak if the size of the offset vector was greater than 30. When the vector
1503        is smaller, the saved offsets during recursion go onto a local stack
1504        vector, but for larger vectors malloc() is used. It was failing to free
1505        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1506        error, in fact).
1507    
1508    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1509        heapframe that is used only when UTF-8 support is enabled. This caused no
1510        problem, but was untidy.
1511    
1512    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1513        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1514        included within another project.
1515    
1516    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1517        slightly modified by me:
1518    
1519          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1520              not building pcregrep.
1521    
1522          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1523              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1524    
1525    25. Forward references, both numeric and by name, in patterns that made use of
1526        duplicate group numbers, could behave incorrectly or give incorrect errors,
1527        because when scanning forward to find the reference group, PCRE was not
1528        taking into account the duplicate group numbers. A pattern such as
1529        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1530    
1531    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1532        the feature of a custom pointer more persuasive (as requested by a user).
1533    
1534    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1535        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1536        pre-defined.
1537    
1538    28. Added support for (*UTF8) at the start of a pattern.
1539    
1540    29. Arrange for flags added by the "release type" setting in CMake to be shown
1541        in the configuration summary.
1542    
1543    
1544    Version 7.8 05-Sep-08
1545    ---------------------
1546    
1547    1.  Replaced UCP searching code with optimized version as implemented for Ad
1548        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1549        stage table and inline lookup instead of a function, giving speed ups of 2
1550        to 5 times on some simple patterns that I tested. Permission was given to
1551        distribute the MultiStage2.py script that generates the tables (it's not in
1552        the tarball, but is in the Subversion repository).
1553    
1554    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1555        scripts.
1556    
1557    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1558        a group with a zero qualifier. The result of the study could be incorrect,
1559        or the function might crash, depending on the pattern.
1560    
1561    4.  Caseless matching was not working for non-ASCII characters in back
1562        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1563        It now works when Unicode Property Support is available.
1564    
1565    5.  In pcretest, an escape such as \x{de} in the data was always generating
1566        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1567        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1568        truncation.
1569    
1570    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1571    
1572    7.  Added two (int) casts to pcregrep when printing the difference of two
1573        pointers, in case they are 64-bit values.
1574    
1575    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1576        test 2 if it fails.
1577    
1578    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1579        and a #define of that name to empty if it is not externally set. This is to
1580        allow users of MSVC to set it if necessary.
1581    
1582    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1583        the convenience functions in the pcre_get.c source file.
1584    
1585    11. An option change at the start of a pattern that had top-level alternatives
1586        could cause overwriting and/or a crash. This command provoked a crash in
1587        some environments:
1588    
1589          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1590    
1591        This potential security problem was recorded as CVE-2008-2371.
1592    
1593    12. For a pattern where the match had to start at the beginning or immediately
1594        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1595        pcre_dfa_exec() could read past the end of the passed subject if there was
1596        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1597        pcretest so that it places the subject at the end of its malloc-ed buffer.
1598    
1599    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1600        exec() might read past the end of the data buffer in UTF-8 mode.
1601    
1602    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1603        the data contained the byte 0x85 as part of a UTF-8 character within its
1604        first line. This applied both to normal and DFA matching.
1605    
1606    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1607        /^[^d]*?$/8 failed to match "abc".
1608    
1609    16. Added a missing copyright notice to pcrecpp_internal.h.
1610    
1611    17. Make it more clear in the documentation that values returned from
1612        pcre_exec() in ovector are byte offsets, not character counts.
1613    
1614    18. Tidied a few places to stop certain compilers from issuing warnings.
1615    
1616    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1617        supplied by Stefan Weber. I made a further small update for 7.8 because
1618        there is a change of source arrangements: the pcre_searchfuncs.c module is
1619        replaced by pcre_ucd.c.
1620    
1621    
1622    Version 7.7 07-May-08
1623    ---------------------
1624    
1625    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1626        a string to a long long, pretend we don't even have a long long." This is
1627        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1628    
1629    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1630        pre-7.6 versions, which defined a global no_arg variable instead of putting
1631        it in the RE class. (See also #8 below.)
1632    
1633    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1634        Lopes.
1635    
1636    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1637    
1638        (1) The include/exclude patterns were being applied to the whole pathnames
1639            of files, instead of just to the final components.
1640    
1641        (2) If there was more than one level of directory, the subdirectories were
1642            skipped unless they satisfied the include/exclude conditions. This is
1643            inconsistent with GNU grep (and could even be seen as contrary to the
1644            pcregrep specification - which I improved to make it absolutely clear).
1645            The action now is always to scan all levels of directory, and just
1646            apply the include/exclude patterns to regular files.
1647    
1648    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1649        --exclude_dir in the tests to avoid scanning .svn directories.
1650    
1651    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1652        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1653        doesn't support NULs in patterns.
1654    
1655    7.  Added some missing "const"s to declarations of static tables in
1656        pcre_compile.c and pcre_dfa_exec.c.
1657    
1658    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1659        caused by fix #2  above. (Subsequently also a second patch to fix the
1660        first patch. And a third patch - this was a messy problem.)
1661    
1662    9.  Applied Craig's patch to remove the use of push_back().
1663    
1664    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1665        matching function regexec().
1666    
1667    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1668        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1669        references. PCRE supports relative numbers with this syntax (I don't think
1670        Oniguruma does).
1671    
1672    12. Previously, a group with a zero repeat such as (...){0} was completely
1673        omitted from the compiled regex. However, this means that if the group
1674        was called as a subroutine from elsewhere in the pattern, things went wrong
1675        (an internal error was given). Such groups are now left in the compiled
1676        pattern, with a new opcode that causes them to be skipped at execution
1677        time.
1678    
1679    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1680        to the way PCRE behaves:
1681    
1682        (a) A lone ] character is dis-allowed (Perl treats it as data).
1683    
1684        (b) A back reference to an unmatched subpattern matches an empty string
1685            (Perl fails the current match path).
1686    
1687        (c) A data ] in a character class must be notated as \] because if the
1688            first data character in a class is ], it defines an empty class. (In
1689            Perl it is not possible to have an empty class.) The empty class []
1690            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1691            The negative empty class [^] matches any one character, independently
1692            of the DOTALL setting.
1693    
1694    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1695        non-existent subpattern following a character class starting with ']' and
1696        containing () gave an internal compiling error instead of "reference to
1697        non-existent subpattern". Fortunately, when the pattern did exist, the
1698        compiled code was correct. (When scanning forwards to check for the
1699        existencd of the subpattern, it was treating the data ']' as terminating
1700        the class, so got the count wrong. When actually compiling, the reference
1701        was subsequently set up correctly.)
1702    
1703    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1704        it was being rejected as not supported by pcre_dfa_exec(), even though
1705        other assertions are supported. I have made pcre_dfa_exec() support
1706        (*FAIL).
1707    
1708    16. The implementation of 13c above involved the invention of a new opcode,
1709        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1710        cannot be changed at match time, I realized I could make a small
1711        improvement to matching performance by compiling OP_ALLANY instead of
1712        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1713        on the OP_ANY path.
1714    
1715    17. Compiling pcretest on Windows with readline support failed without the
1716        following two fixes: (1) Make the unistd.h include conditional on
1717        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1718    
1719    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1720        ncurses library to be included for pcretest when ReadLine support is
1721        requested, but also to allow for it to be overridden. This patch came from
1722        Daniel Bergström.
1723    
1724    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1725        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1726        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1727        spotting this.
1728    
1729    
1730    Version 7.6 28-Jan-08
1731    ---------------------
1732    
1733    1.  A character class containing a very large number of characters with
1734        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1735        overflow.
1736    
1737    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1738        HAVE_LONG_LONG is not defined.
1739    
1740    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1741        bring them up to date and include new features. This patch includes:
1742    
1743        - Fixed PH's badly added libz and libbz2 support.
1744        - Fixed a problem with static linking.
1745        - Added pcredemo. [But later removed - see 7 below.]
1746        - Fixed dftables problem and added an option.
1747        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1748            HAVE_LONG_LONG.
1749        - Added readline support for pcretest.
1750        - Added an listing of the option settings after cmake has run.
1751    
1752    4.  A user submitted a patch to Makefile that makes it easy to create
1753        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1754        Makefile.am that cause it to include this special target, without
1755        affecting anything else. Note that the same mingw target plus all
1756        the other distribution libraries and programs are now supported
1757        when configuring with CMake (see 6 below) instead of with
1758        Configure/Make.
1759    
1760    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1761        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1762        exported in the Windows port". It has not yet been confirmed that the patch
1763        solves the problem, but it does no harm.
1764    
1765    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1766        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1767        with CMake, and also correct the comment about stack recursion.
1768    
1769    7.  Remove the automatic building of pcredemo from the ./configure system and
1770        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1771        of a program that users should build themselves after PCRE is installed, so
1772        building it automatically is not really right. What is more, it gave
1773        trouble in some build environments.
1774    
1775    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1776    
1777    
1778    Version 7.5 10-Jan-08
1779    ---------------------
1780    
1781    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1782        values in parens when parsing an RE using the C++ wrapper."
1783    
1784    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1785        Characters greater than 255 were excluded from the class instead of being
1786        included.
1787    
1788    3.  The same bug as (2) above applied to negated POSIX classes such as
1789        [:^space:].
1790    
1791    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1792        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1793        I have changed it.
1794    
1795    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1796        first named subpattern) and a construct such as (?&a) would reference the
1797        first named subpattern whose name started with "a" (in other words, the
1798        length check was missing). Both these problems are fixed. "Subpattern name
1799        expected" is now given for (?&) (a zero-length name), and this patch also
1800        makes it give the same error for \k'' (previously it complained that that
1801        was a reference to a non-existent subpattern).
1802    
1803    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1804        this is right because (?- can be followed by option settings as well as by
1805        digits. I have, however, made the messages clearer.
1806    
1807    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1808        than the number used in the conditional) now cause a compile-time error.
1809        This is actually not compatible with Perl, which accepts such patterns, but
1810        treats the conditional as always being FALSE (as PCRE used to), but it
1811        seems to me that giving a diagnostic is better.
1812    
1813    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1814        and messages.
1815    
1816    9.  Fix two occurrences of "backslash" in comments that should have been
1817        "backspace".
1818    
1819    10. Remove two redundant lines of code that can never be obeyed (their function
1820        was moved elsewhere).
1821    
1822    11. The program that makes PCRE's Unicode character property table had a bug
1823        which caused it to generate incorrect table entries for sequences of
1824        characters that have the same character type, but are in different scripts.
1825        It amalgamated them into a single range, with the script of the first of
1826        them. In other words, some characters were in the wrong script. There were
1827        thirteen such cases, affecting characters in the following ranges:
1828    
1829          U+002b0 - U+002c1
1830          U+0060c - U+0060d
1831          U+0061e - U+00612
1832          U+0064b - U+0065e
1833          U+0074d - U+0076d
1834          U+01800 - U+01805
1835          U+01d00 - U+01d77
1836          U+01d9b - U+01dbf
1837          U+0200b - U+0200f
1838          U+030fc - U+030fe
1839          U+03260 - U+0327f
1840          U+0fb46 - U+0fbb1
1841          U+10450 - U+1049d
1842    
1843    12. The -o option (show only the matching part of a line) for pcregrep was not
1844        compatible with GNU grep in that, if there was more than one match in a
1845        line, it showed only the first of them. It now behaves in the same way as
1846        GNU grep.
1847    
1848    13. If the -o and -v options were combined for pcregrep, it printed a blank
1849        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1850        does the same. The return code can be used to tell if there were any
1851        non-matching lines.
1852    
1853    14. Added --file-offsets and --line-offsets to pcregrep.
1854    
1855    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1856        infinitely looping recursion. The bug was that positive lookaheads were not
1857        being skipped when checking for a possible empty match (negative lookaheads
1858        and both kinds of lookbehind were skipped).
1859    
1860    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1861        inclusion of <windows.h> to before rather than after the definition of
1862        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1863    
1864    17. Specifying a possessive quantifier with a specific limit for a Unicode
1865        character property caused pcre_compile() to compile bad code, which led at
1866        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1867        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1868        caused the error; without that there was no problem.
1869    
1870    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1871    
1872    19. Added --enable-pcretest-libreadline.
1873    
1874    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1875        RE::GlobalReplace(). As a result, the number of replacements returned was
1876        double what it should be. I removed one of the increments, but Craig sent a
1877        later patch that removed the other one (the right fix) and added unit tests
1878        that check the return values (which was not done before).
1879    
1880    21. Several CMake things:
1881    
1882        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1883            the names libpcre and libpcreposix, not just pcre and pcreposix.
1884    
1885        (2) The above change means that pcretest and pcregrep are now correctly
1886            linked with the newly-built libraries, not previously installed ones.
1887    
1888        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1889    
1890    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1891        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1892        UTF-8 newline character). The key issue is that the pattern starts .*;
1893        this means that the match must be either at the beginning, or after a
1894        newline. The bug was in the code for advancing after a failed match and
1895        checking that the new position followed a newline. It was not taking
1896        account of UTF-8 characters correctly.
1897    
1898    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1899        character classes. PCRE was not treating the sequence [:...:] as a
1900        character class unless the ... were all letters. Perl, however, seems to
1901        allow any characters between [: and :], though of course it rejects as
1902        unknown any "names" that contain non-letters, because all the known class
1903        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1904        for example, whereas PCRE did not - it did not recognize a POSIX character
1905        class. This seemed a bit dangerous, so the code has been changed to be
1906        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1907        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1908        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1909        Perl does, and where it didn't before.
1910    
1911    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1912        Windows environments %n is disabled by default.
1913    
1914    
1915    Version 7.4 21-Sep-07
1916    ---------------------
1917    
1918    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1919        means that a class such as [\s] counted as "explicit reference to CR or
1920        LF". That isn't really right - the whole point of the change was to try to
1921        help when there was an actual mention of one of the two characters. So now
1922        the change happens only if \r or \n (or a literal CR or LF) character is
1923        encountered.
1924    
1925    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1926        of both had grown to the point where there were only 3 bits left.
1927        Fortunately, there was spare space in the data structure, and so I have
1928        moved the internal flags into a new 16-bit field to free up more option
1929        bits.
1930    
1931    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1932        but did not set the internal JCHANGED flag - either of these is enough to
1933        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1934        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1935        start sets both bits.
1936    
1937    4.  Added options (at build time, compile time, exec time) to change \R from
1938        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1939    
1940    5.  doc/pcresyntax.html was missing from the distribution.
1941    
1942    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1943        compatibility, even though it is no longer used.
1944    
1945    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1946        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1947        windows.h file is present (where different names are used). [This was
1948        reversed later after testing - see 16 below.]
1949    
1950    8.  Changed all #include <config.h> to #include "config.h". There were also
1951        some further <pcre.h> cases that I changed to "pcre.h".
1952    
1953    9.  When pcregrep was used with the --colour option, it missed the line ending
1954        sequence off the lines that it output.
1955    
1956    10. It was pointed out to me that arrays of string pointers cause lots of
1957        relocations when a shared library is dynamically loaded. A technique of
1958        using a single long string with a table of offsets can drastically reduce
1959        these. I have refactored PCRE in four places to do this. The result is
1960        dramatic:
1961    
1962          Originally:                          290
1963          After changing UCP table:            187
1964          After changing error message table:   43
1965          After changing table of "verbs"       36
1966          After changing table of Posix names   22
1967    
1968        Thanks to the folks working on Gregex for glib for this insight.
1969    
1970    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1971        unicode-properties was also set.
1972    
1973    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1974    
1975    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1976        checked only for CRLF.
1977    
1978    14. Added casts to pcretest.c to avoid compiler warnings.
1979    
1980    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1981    
1982    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1983        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1984        entirely. This removes changes made in 7 above.
1985    
1986    17. The CMake files have been updated, and there is now more information about
1987        building with CMake in the NON-UNIX-USE document.
1988    
1989    
1990    Version 7.3 28-Aug-07
1991  ---------------------  ---------------------
1992    
1993   1. In the rejigging of the build system that eventually resulted in 7.1, the   1. In the rejigging of the build system that eventually resulted in 7.1, the
# Line 9  Version 7.3 05-Jul-07 Line 1995  Version 7.3 05-Jul-07
1995      brackets there is not right, since it causes compilers to look for an      brackets there is not right, since it causes compilers to look for an
1996      installed pcre.h, not the version that is in the source that is being      installed pcre.h, not the version that is in the source that is being
1997      compiled (which of course may be different). I have changed it back to:      compiled (which of course may be different). I have changed it back to:
1998    
1999        #include "pcre.h"        #include "pcre.h"
2000    
2001      I have a vague recollection that the change was concerned with compiling in      I have a vague recollection that the change was concerned with compiling in
2002      different directories, but in the new build system, that is taken care of      different directories, but in the new build system, that is taken care of
2003      by the VPATH setting the Makefile.      by the VPATH setting the Makefile.
2004    
2005   2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed   2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2006      when the subject happened to end in the byte 0x85 (e.g. if the last      when the subject happened to end in the byte 0x85 (e.g. if the last
2007      character was \x{1ec5}). *Character* 0x85 is one of the "any" newline      character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
# Line 23  Version 7.3 05-Jul-07 Line 2009  Version 7.3 05-Jul-07
2009      of another character. The bug was that, for an unlimited repeat of . in      of another character. The bug was that, for an unlimited repeat of . in
2010      not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by      not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2011      characters when looking for a newline.      characters when looking for a newline.
   
  3. A small performance improvement in the DOTALL UTF-8 mode .* case.  
   
  4. Debugging: adjusted the names of opcodes for different kinds of parentheses  
     in debug output.  
   
  5. Arrange to use "%I64d" instead of "%lld" for long printing in the pcrecpp  
     unittest when running under MinGW.  
   
  6. ESC_K was left out of the EBCDIC table.  
2012    
2013     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2014    
2015     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2016        in debug output.
2017    
2018     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2019        long printing in the pcrecpp unittest when running under MinGW.
2020    
2021     6. ESC_K was left out of the EBCDIC table.
2022    
2023     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2024        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2025        limit also applies to "virtual nesting" when a pattern is recursive, and in
2026        this case 1000 isn't so big. I have been able to remove this limit at the
2027        expense of backing off one optimization in certain circumstances. Normally,
2028        when pcre_exec() would call its internal match() function recursively and
2029        immediately return the result unconditionally, it uses a "tail recursion"
2030        feature to save stack. However, when a subpattern that can match an empty
2031        string has an unlimited repetition quantifier, it no longer makes this
2032        optimization. That gives it a stack frame in which to save the data for
2033        checking that an empty string has been matched. Previously this was taken
2034        from the 1000-entry workspace that had been reserved. So now there is no
2035        explicit limit, but more stack is used.
2036    
2037     8. Applied Daniel's patches to solve problems with the import/export magic
2038        syntax that is required for Windows, and which was going wrong for the
2039        pcreposix and pcrecpp parts of the library. These were overlooked when this
2040        problem was solved for the main library.
2041    
2042     9. There were some crude static tests to avoid integer overflow when computing
2043        the size of patterns that contain repeated groups with explicit upper
2044        limits. As the maximum quantifier is 65535, the maximum group length was
2045        set at 30,000 so that the product of these two numbers did not overflow a
2046        32-bit integer. However, it turns out that people want to use groups that
2047        are longer than 30,000 bytes (though not repeat them that many times).
2048        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2049        made it possible to implement the integer overflow checks in a much more
2050        dynamic way, which I have now done. The artificial limitation on group
2051        length has been removed - we now have only the limit on the total length of
2052        the compiled pattern, which depends on the LINK_SIZE setting.
2053    
2054    10. Fixed a bug in the documentation for get/copy named substring when
2055        duplicate names are permitted. If none of the named substrings are set, the
2056        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2057        empty string.
2058    
2059    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2060        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2061        because the ] is interpreted as the first data character and the
2062        terminating ] is not found. PCRE has been made compatible with Perl in this
2063        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2064        cause memory overwriting.
2065    
2066    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2067        string has been matched (to stop an infinite loop). It was not recognizing
2068        a conditional subpattern that could match an empty string if that
2069        subpattern was within another subpattern. For example, it looped when
2070        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2071        condition was not nested. This bug has been fixed.
2072    
2073    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2074        past the start of the subject in the presence of bytes with the top bit
2075        set, for example "\x8aBCD".
2076    
2077    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2078        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2079    
2080    14. Optimized (?!) to (*FAIL).
2081    
2082    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2083        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2084        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2085        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2086        does: it's just the validity check that is more restrictive.
2087    
2088    16. Inserted checks for integer overflows during escape sequence (backslash)
2089        processing, and also fixed erroneous offset values for syntax errors during
2090        backslash processing.
2091    
2092    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2093        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2094    
2095    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2096        caused an overrun.
2097    
2098    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2099        something other than just ASCII characters) inside a group that had an
2100        unlimited repeat caused a loop at compile time (while checking to see
2101        whether the group could match an empty string).
2102    
2103    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2104        [\P{Any}] did so. (Error in the code for printing property names.)
2105    
2106    21. An orphan \E inside a character class could cause a crash.
2107    
2108    22. A repeated capturing bracket such as (A)? could cause a wild memory
2109        reference during compilation.
2110    
2111    23. There are several functions in pcre_compile() that scan along a compiled
2112        expression for various reasons (e.g. to see if it's fixed length for look
2113        behind). There were bugs in these functions when a repeated \p or \P was
2114        present in the pattern. These operators have additional parameters compared
2115        with \d, etc, and these were not being taken into account when moving along
2116        the compiled data. Specifically:
2117    
2118        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2119            length.
2120    
2121        (b) An item such as \pL+ within a repeated group could cause crashes or
2122            loops.
2123    
2124        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2125            "reference to non-existent subpattern" error.
2126    
2127        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2128    
2129    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2130        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2131    
2132    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2133    
2134    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2135        character were causing crashes (broken optimization).
2136    
2137    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2138        \p or \P) caused a compile-time loop.
2139    
2140    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2141        break. For example, the unstudied pattern [\r\n]A does not match the string
2142        "\r\nA" because change 7.0/46 below moves the current point on by two
2143        characters after failing to match at the start. However, the pattern \nA
2144        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2145        the same is true. There doesn't seem any very clean way out of this, but
2146        what I have chosen to do makes the common cases work: PCRE now takes note
2147        of whether there can be an explicit match for \r or \n anywhere in the
2148        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2149        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2150        pattern has explicit CR or LF references.
2151    
2152    29. Added (*CR) etc for changing newline setting at start of pattern.
2153    
2154    
2155  Version 7.2 19-Jun-07  Version 7.2 19-Jun-07

Legend:
Removed from v.195  
changed lines
  Added in v.1045

  ViewVC Help
Powered by ViewVC 1.1.5