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

Diff of /code/trunk/ChangeLog

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

revision 422 by ph10, Fri Aug 14 16:42:55 2009 UTC revision 937 by ph10, Sun Feb 26 15:58:56 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 8.00 ??-???-??  Version 8.31
5    -----------------------------
6    
7    1.  Fixing a wrong JIT test case and some compiler warnings.
8    
9    2.  Removed a bashism from the RunTest script.
10    
11    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
12        to unsigned type, result still unsigned" that was given by an MS compiler
13        on encountering the code "-sizeof(xxx)".
14    
15    4.  Partial matching support is added to the JIT compiler.
16    
17    5.  Fixed several bugs concerned with partial matching of items that consist
18        of more than one character:
19    
20        (a) /^(..)\1/ did not partially match "aba" because checking references was
21            done on an "all or nothing" basis. This also applied to repeated
22            references.
23    
24        (b) \R did not give a hard partial match if \r was found at the end of the
25            subject.
26    
27        (c) \X did not give a hard partial match after matching one or more
28            characters at the end of the subject.
29    
30        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
31            a partial match for the string "\r".
32    
33        (e) When newline was set to CRLF, the metacharacter "." did not recognize
34            a partial match for a CR character at the end of the subject string.
35    
36    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
37        running pcretest, the text "(JIT)" added to the output whenever JIT is
38        actually used to run the match.
39    
40    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
41        or /S+[+] with a digit between 1 and 7.
42    
43    8.  OP_NOT now supports any UTF character not just single-byte ones.
44    
45    9.  (*MARK) control verb is now supported by the JIT compiler.
46    
47    10. The command "./RunTest list" lists the available tests without actually
48        running any of them. (Because I keep forgetting what they all are.)
49    
50    11. Add PCRE_INFO_MAXLOOKBEHIND.
51    
52    12. Applied a (slightly modified) user-supplied patch that improves performance
53        when the heap is used for recursion (compiled with --disable-stack-for-
54        recursion). Instead of malloc and free for each heap frame each time a
55        logical recursion happens, frames are retained on a chain and re-used where
56        possible. This sometimes gives as much as 30% improvement.
57    
58    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
59        call.
60    
61    14. As documented, (*COMMIT) is now confined to within a positive assertion.
62    
63    15. It is now possible to link pcretest with libedit as an alternative to
64        libreadline.
65    
66    
67    Version 8.30 04-February-2012
68    -----------------------------
69    
70    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
71        name is defined in ctype.h.
72    
73    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
74        only in quite long subpatterns.
75    
76    3.  Removed the function pcre_info(), which has been obsolete and deprecated
77        since it was replaced by pcre_fullinfo() in February 2000.
78    
79    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
80        match a (*MARK), and the match failed at the start of the subject, a
81        reference to memory before the start of the subject could occur. This bug
82        was introduced by fix 17 of release 8.21.
83    
84    5.  A reference to an unset group with zero minimum repetition was giving
85        totally wrong answers (in non-JavaScript-compatibility mode). For example,
86        /(another)?(\1?)test/ matched against "hello world test". This bug was
87        introduced in release 8.13.
88    
89    6.  Add support for 16-bit character strings (a large amount of work involving
90        many changes and refactorings).
91    
92    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
93        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
94        from a file.
95    
96    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
97        rounding is not applied in this particular case).
98    
99    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
100        if they appear, or are escaped, in patterns.
101    
102    10. Get rid of a number of -Wunused-but-set-variable warnings.
103    
104    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
105        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
106        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
107        also returns the mark "x". This bug applied to capturing parentheses,
108        non-capturing parentheses, and atomic parentheses. It also applied to some
109        assertions.
110    
111    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
112        information out of configure.ac instead of relying on pcre.h.generic, which
113        is not stored in the repository.
114    
115    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
116        -lreadline.
117    
118    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
119    
120    15. Applied Graycode's patch to put the top-level frame on the stack rather
121        than the heap when not using the stack for recursion. This gives a
122        performance improvement in many cases when recursion is not deep.
123    
124    16. Experimental code added to "pcretest -C" to output the stack frame size.
125    
126    
127    Version 8.21 12-Dec-2011
128    ------------------------
129    
130    1.  Updating the JIT compiler.
131    
132    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
133        are added as well.
134    
135    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
136        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
137        calling _pcre_jit_exec. Some extra comments are added.
138    
139    4.  (*MARK) settings inside atomic groups that do not contain any capturing
140        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
141        was introduced by change 18 for 8.20.
142    
143    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
144        ECMA-262 standard.
145    
146    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
147        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
148        This bug was probably introduced by change 9 of 8.13.
149    
150    7.  While fixing 6 above, I noticed that a number of other items were being
151        incorrectly rejected as "not fixed length". This arose partly because newer
152        opcodes had not been added to the fixed-length checking code. I have (a)
153        corrected the bug and added tests for these items, and (b) arranged for an
154        error to occur if an unknown opcode is encountered while checking for fixed
155        length instead of just assuming "not fixed length". The items that were
156        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
157        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
158        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
159    
160    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
161        being incorrectly compiled and would have given unpredicatble results.
162    
163    9.  A possessively repeated subpattern with minimum repeat count greater than
164        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
165        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
166        the first (A) could occur when it should not.
167    
168    10. Add a cast and remove a redundant test from the code.
169    
170    11. JIT should use pcre_malloc/pcre_free for allocation.
171    
172    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
173        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
174        is anything other than /usr, -L is still shown).
175    
176    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
177    
178    14. Perl does not support \N without a following name in a [] class; PCRE now
179        also gives an error.
180    
181    15. If a forward reference was repeated with an upper limit of around 2000,
182        it caused the error "internal error: overran compiling workspace". The
183        maximum number of forward references (including repeats) was limited by the
184        internal workspace, and dependent on the LINK_SIZE. The code has been
185        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
186        the default depends on LINK_SIZE. There is a new upper limit (for safety)
187        of around 200,000 forward references. While doing this, I also speeded up
188        the filling in of repeated forward references.
189    
190    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
191        incorrectly expecting the subject to contain another "a" after the start.
192    
193    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
194        in the match, the SKIP should be ignored. This was not happening; instead
195        the SKIP was being treated as NOMATCH. For patterns such as
196        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
197        tested.
198    
199    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
200        now much more compatible with Perl, in particular in cases where the result
201        is a non-match for a non-anchored pattern. For example, if
202        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
203        "m", where previously it did not return a name. A side effect of this
204        change is that for partial matches, the last encountered mark name is
205        returned, as for non matches. A number of tests that were previously not
206        Perl-compatible have been moved into the Perl-compatible test files. The
207        refactoring has had the pleasing side effect of removing one argument from
208        the match() function, thus reducing its stack requirements.
209    
210    19. If the /S+ option was used in pcretest to study a pattern using JIT,
211        subsequent uses of /S (without +) incorrectly behaved like /S+.
212    
213    21. Retrieve executable code size support for the JIT compiler and fixing
214        some warnings.
215    
216    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
217        not work when the shorter character appeared right at the end of the
218        subject string.
219    
220    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
221        systems.
222    
223    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
224        output it when the /M option is used in pcretest.
225    
226    25. The CheckMan script was not being included in the distribution. Also, added
227        an explicit "perl" to run Perl scripts from the PrepareRelease script
228        because this is reportedly needed in Windows.
229    
230    26. If study data was being save in a file and studying had not found a set of
231        "starts with" bytes for the pattern, the data written to the file (though
232        never used) was taken from uninitialized memory and so caused valgrind to
233        complain.
234    
235    27. Updated RunTest.bat as provided by Sheri Pierce.
236    
237    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
238    
239    29. Computation of memory usage for the table of capturing group names was
240        giving an unnecessarily large value.
241    
242    
243    Version 8.20 21-Oct-2011
244    ------------------------
245    
246    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
247        a POSIX class. After further experiments with Perl, which convinced me that
248        Perl has bugs and confusions, a closing square bracket is no longer allowed
249        in a POSIX name. This bug also affected patterns with classes that started
250        with full stops.
251    
252    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
253        captured substring, but while checking the failing first alternative,
254        substring 1 is temporarily captured. If the output vector supplied to
255        pcre_exec() was not big enough for this capture, the yield of the function
256        was still zero ("insufficient space for captured substrings"). This cannot
257        be totally fixed without adding another stack variable, which seems a lot
258        of expense for a edge case. However, I have improved the situation in cases
259        such as /(a)(b)x|abc/ matched against "abc", where the return code
260        indicates that fewer than the maximum number of slots in the ovector have
261        been set.
262    
263    3.  Related to (2) above: when there are more back references in a pattern than
264        slots in the output vector, pcre_exec() uses temporary memory during
265        matching, and copies in the captures as far as possible afterwards. It was
266        using the entire output vector, but this conflicts with the specification
267        that only 2/3 is used for passing back captured substrings. Now it uses
268        only the first 2/3, for compatibility. This is, of course, another edge
269        case.
270    
271    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
272        main code base, and can be used by building with --enable-jit. When this is
273        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
274        runtime --no-jit option is given.
275    
276    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
277        ovector, the return from the function was zero, implying that there were
278        other matches that did not fit. The correct "exactly full" value is now
279        returned.
280    
281    6.  If a subpattern that was called recursively or as a subroutine contained
282        (*PRUNE) or any other control that caused it to give a non-standard return,
283        invalid errors such as "Error -26 (nested recursion at the same subject
284        position)" or even infinite loops could occur.
285    
286    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
287        computing the minimum length on reaching *ACCEPT, and so ended up with the
288        wrong value of 1 rather than 0. Further investigation indicates that
289        computing a minimum subject length in the presence of *ACCEPT is difficult
290        (think back references, subroutine calls), and so I have changed the code
291        so that no minimum is registered for a pattern that contains *ACCEPT.
292    
293    8.  If (*THEN) was present in the first (true) branch of a conditional group,
294        it was not handled as intended. [But see 16 below.]
295    
296    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
297        Sheri Pierce.
298    
299    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
300        the first byte in a match must be "a".
301    
302    11. Change 17 for 8.13 increased the recursion depth for patterns like
303        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
304        pattern contains any instances of (*THEN). If it does not, the old
305        optimizations are restored. It would be nice to do this on a per-group
306        basis, but at the moment that is not feasible.
307    
308    12. In some environments, the output of pcretest -C is CRLF terminated. This
309        broke RunTest's code that checks for the link size. A single white space
310        character after the value is now allowed for.
311    
312    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
313        For "fr", it uses the Windows-specific input and output files.
314    
315    14. If (*THEN) appeared in a group that was called recursively or as a
316        subroutine, it did not work as intended. [But see next item.]
317    
318    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
319        pattern fragments (but not containing any | characters). If A and B are
320        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
321        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
322        D. In other words, Perl considers parentheses that do not contain any |
323        characters to be part of a surrounding alternative, whereas PCRE was
324        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
325        differently. PCRE now behaves in the same way as Perl, except in the case
326        of subroutine/recursion calls such as (?1) which have in any case always
327        been different (but PCRE had them first :-).
328    
329    16. Related to 15 above: Perl does not treat the | in a conditional group as
330        creating alternatives. Such a group is treated in the same way as an
331        ordinary group without any | characters when processing (*THEN). PCRE has
332        been changed to match Perl's behaviour.
333    
334    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
335        RunGrepTest script failed.
336    
337    18. Change 22 for version 13 caused atomic groups to use more stack. This is
338        inevitable for groups that contain captures, but it can lead to a lot of
339        stack use in large patterns. The old behaviour has been restored for atomic
340        groups that do not contain any capturing parentheses.
341    
342    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
343        suppress the check for a minimum subject length at run time. (If it was
344        given to pcre_exec() or pcre_dfa_exec() it did work.)
345    
346    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
347        fail to work when decoding hex characters in data strings in EBCDIC
348        environments.
349    
350    21. It appears that in at least one Mac OS environment, the isxdigit() function
351        is implemented as a macro that evaluates to its argument more than once,
352        contravening the C 90 Standard (I haven't checked a later standard). There
353        was an instance in pcretest which caused it to go wrong when processing
354        \x{...} escapes in subject strings. The has been rewritten to avoid using
355        things like p++ in the argument of isxdigit().
356    
357    
358    Version 8.13 16-Aug-2011
359    ------------------------
360    
361    1.  The Unicode data tables have been updated to Unicode 6.0.0.
362    
363    2.  Two minor typos in pcre_internal.h have been fixed.
364    
365    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
366        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
367        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
368    
369    4.  There were a number of related bugs in the code for matching backrefences
370        caselessly in UTF-8 mode when codes for the characters concerned were
371        different numbers of bytes. For example, U+023A and U+2C65 are an upper
372        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
373        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
374        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
375        2-byte code at the end of the subject (it thought there wasn't enough data
376        left).
377    
378    5.  Comprehensive information about what went wrong is now returned by
379        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
380        as the output vector has at least 2 elements. The offset of the start of
381        the failing character and a reason code are placed in the vector.
382    
383    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
384        now returned is for the first byte of the failing character, instead of the
385        last byte inspected. This is an incompatible change, but I hope it is small
386        enough not to be a problem. It makes the returned offset consistent with
387        pcre_exec() and pcre_dfa_exec().
388    
389    7.  pcretest now gives a text phrase as well as the error number when
390        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
391        failure, the offset and reason code are output.
392    
393    8.  When \R was used with a maximizing quantifier it failed to skip backwards
394        over a \r\n pair if the subsequent match failed. Instead, it just skipped
395        back over a single character (\n). This seems wrong (because it treated the
396        two characters as a single entity when going forwards), conflicts with the
397        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
398        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
399        has been changed.
400    
401    9.  Some internal refactoring has changed the processing so that the handling
402        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
403        time (the PCRE_DOTALL option was changed this way some time ago: version
404        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
405        which was always a bit of a fudge. It also means that there is one less
406        argument for the match() function, which reduces its stack requirements
407        slightly. This change also fixes an incompatibility with Perl: the pattern
408        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
409    
410    10. More internal refactoring has drastically reduced the number of recursive
411        calls to match() for possessively repeated groups such as (abc)++ when
412        using pcre_exec().
413    
414    11. While implementing 10, a number of bugs in the handling of groups were
415        discovered and fixed:
416    
417        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
418        (a|)*(?1) gave a compile-time internal error.
419        ((a|)+)+  did not notice that the outer group could match an empty string.
420        (^a|^)+   was not marked as anchored.
421        (.*a|.*)+ was not marked as matching at start or after a newline.
422    
423    12. Yet more internal refactoring has removed another argument from the match()
424        function. Special calls to this function are now indicated by setting a
425        value in a variable in the "match data" data block.
426    
427    13. Be more explicit in pcre_study() instead of relying on "default" for
428        opcodes that mean there is no starting character; this means that when new
429        ones are added and accidentally left out of pcre_study(), testing should
430        pick them up.
431    
432    14. The -s option of pcretest has been documented for ages as being an old
433        synonym of -m (show memory usage). I have changed it to mean "force study
434        for every regex", that is, assume /S for every regex. This is similar to -i
435        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
436        using it. It makes it easier to run collections of tests with and without
437        study enabled, and thereby test pcre_study() more easily. All the standard
438        tests are now run with and without -s (but some patterns can be marked as
439        "never study" - see 20 below).
440    
441    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
442        restoration of the capturing data to the outer values was not happening
443        correctly.
444    
445    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
446        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
447        pattern had matched an empty string, and so incorrectly returned a no
448        match.
449    
450    17. There was optimizing code for the last branch of non-capturing parentheses,
451        and also for the obeyed branch of a conditional subexpression, which used
452        tail recursion to cut down on stack usage. Unfortunately, now that there is
453        the possibility of (*THEN) occurring in these branches, tail recursion is
454        no longer possible because the return has to be checked for (*THEN). These
455        two optimizations have therefore been removed. [But see 8.20/11 above.]
456    
457    18. If a pattern containing \R was studied, it was assumed that \R always
458        matched two bytes, thus causing the minimum subject length to be
459        incorrectly computed because \R can also match just one byte.
460    
461    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
462        was incorrectly computed.
463    
464    20. If /S is present twice on a test pattern in pcretest input, it now
465        *disables* studying, thereby overriding the use of -s on the command line
466        (see 14 above). This is necessary for one or two tests to keep the output
467        identical in both cases.
468    
469    21. When (*ACCEPT) was used in an assertion that matched an empty string and
470        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
471    
472    22. When an atomic group that contained a capturing parenthesis was
473        successfully matched, but the branch in which it appeared failed, the
474        capturing was not being forgotten if a higher numbered group was later
475        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
476        group 1 to "a", when in fact it should be unset. This applied to multi-
477        branched capturing and non-capturing groups, repeated or not, and also to
478        positive assertions (capturing in negative assertions does not happen
479        in PCRE) and also to nested atomic groups.
480    
481    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
482        subject after a captured substring, to make it easier to tell which of a
483        number of identical substrings has been captured.
484    
485    24. The way atomic groups are processed by pcre_exec() has been changed so that
486        if they are repeated, backtracking one repetition now resets captured
487        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
488        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
489        "aaa". Previously, it would have been "a". As part of this code
490        refactoring, the way recursive calls are handled has also been changed.
491    
492    25. If an assertion condition captured any substrings, they were not passed
493        back unless some other capturing happened later. For example, if
494        (?(?=(a))a) was matched against "a", no capturing was returned.
495    
496    26. When studying a pattern that contained subroutine calls or assertions,
497        the code for finding the minimum length of a possible match was handling
498        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
499        group 1 called group 2 while simultaneously a separate group 2 called group
500        1). A stack overflow occurred in this case. I have fixed this by limiting
501        the recursion depth to 10.
502    
503    27. Updated RunTest.bat in the distribution to the version supplied by Tom
504        Fortmann. This supports explicit test numbers on the command line, and has
505        argument validation and error reporting.
506    
507    28. An instance of \X with an unlimited repeat could fail if at any point the
508        first character it looked at was a mark character.
509    
510    29. Some minor code refactoring concerning Unicode properties and scripts
511        should reduce the stack requirement of match() slightly.
512    
513    30. Added the '=' option to pcretest to check the setting of unused capturing
514        slots at the end of the pattern, which are documented as being -1, but are
515        not included in the return count.
516    
517    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
518        compiled something random. Now it gives a compile-time error (as does
519        Perl).
520    
521    32. A *MARK encountered during the processing of a positive assertion is now
522        recorded and passed back (compatible with Perl).
523    
524    33. If --only-matching or --colour was set on a pcregrep call whose pattern
525        had alternative anchored branches, the search for a second match in a line
526        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
527        matched the line "0102" twice. The same bug affected patterns that started
528        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
529        twice.
530    
531    34. Previously, PCRE did not allow quantification of assertions. However, Perl
532        does, and because of capturing effects, quantifying parenthesized
533        assertions may at times be useful. Quantifiers are now allowed for
534        parenthesized assertions.
535    
536    35. A minor code tidy in pcre_compile() when checking options for \R usage.
537    
538    36. \g was being checked for fancy things in a character class, when it should
539        just be a literal "g".
540    
541    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
542        appearance of a nested POSIX class supersedes an apparent external class.
543        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
544        unescaped square brackets may also appear as part of class names. For
545        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
546        more like Perl. (But see 8.20/1 above.)
547    
548    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
549        was because it thought it was \N{name}, which is not supported).
550    
551    39. Add minix to OS list not supporting the -S option in pcretest.
552    
553    40. PCRE tries to detect cases of infinite recursion at compile time, but it
554        cannot analyze patterns in sufficient detail to catch mutual recursions
555        such as ((?1))((?2)). There is now a runtime test that gives an error if a
556        subgroup is called recursively as a subpattern for a second time at the
557        same position in the subject string. In previous releases this might have
558        been caught by the recursion limit, or it might have run out of stack.
559    
560    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
561        happen only once. PCRE was, however incorrectly giving a compile time error
562        "recursive call could loop indefinitely" because it cannot analyze the
563        pattern in sufficient detail. The compile time test no longer happens when
564        PCRE is compiling a conditional subpattern, but actual runaway loops are
565        now caught at runtime (see 40 above).
566    
567    42. It seems that Perl allows any characters other than a closing parenthesis
568        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
569        has been changed to be the same.
570    
571    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
572        as not to get warnings when autogen.sh is called. Also changed
573        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
574    
575    44. To help people who use pcregrep to scan files containing exceedingly long
576        lines, the following changes have been made:
577    
578        (a) The default value of the buffer size parameter has been increased from
579            8K to 20K. (The actual buffer used is three times this size.)
580    
581        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
582            PCRE is built.
583    
584        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
585            to be set at run time.
586    
587        (d) Numerical values in pcregrep options can be followed by K or M, for
588            example --buffer-size=50K.
589    
590        (e) If a line being scanned overflows pcregrep's buffer, an error is now
591            given and the return code is set to 2.
592    
593    45. Add a pointer to the latest mark to the callout data block.
594    
595    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
596        partial match of an empty string instead of no match. This was specific to
597        the use of ".".
598    
599    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
600        complete match instead of a partial match. This bug was dependent on both
601        the PCRE_UTF8 and PCRE_DOTALL options being set.
602    
603    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
604        starting byte set, because \b was not being ignored.
605    
606    
607    Version 8.12 15-Jan-2011
608    ------------------------
609    
610    1.  Fixed some typos in the markup of the man pages, and wrote a script that
611        checks for such things as part of the documentation building process.
612    
613    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
614        --match-limit and --recursion-limit options (added for 8.11). In
615        particular, this made one of the standard tests fail. (The integer value
616        went into the wrong half of a long int.)
617    
618    3.  If the --colour option was given to pcregrep with -v (invert match), it
619        did strange things, either producing crazy output, or crashing. It should,
620        of course, ignore a request for colour when reporting lines that do not
621        match.
622    
623    4.  Another pcregrep bug caused similar problems if --colour was specified with
624        -M (multiline) and the pattern match finished with a line ending.
625    
626    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
627        matched in multiline mode, the following line was shown as part of the
628        match. This seems wrong, so I have changed it.
629    
630    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
631        the check for further matches in the same line (so they could be coloured)
632        to overrun the end of the current line. If another match was found, it was
633        incorrectly shown (and then shown again when found in the next line).
634    
635    7.  If pcregrep was compiled under Windows, there was a reference to the
636        function pcregrep_exit() before it was defined. I am assuming this was
637        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
638        reported by a user. I've moved the definition above the reference.
639    
640    
641    Version 8.11 10-Dec-2010
642    ------------------------
643    
644    1.  (*THEN) was not working properly if there were untried alternatives prior
645        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
646        backtracked to try for "b" instead of moving to the next alternative branch
647        at the same level (in this case, to look for "c"). The Perl documentation
648        is clear that when (*THEN) is backtracked onto, it goes to the "next
649        alternative in the innermost enclosing group".
650    
651    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
652        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
653        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
654        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
655        (*THEN).
656    
657    3.  If \s appeared in a character class, it removed the VT character from
658        the class, even if it had been included by some previous item, for example
659        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
660        of \s, but is part of the POSIX "space" class.)
661    
662    4.  A partial match never returns an empty string (because you can always
663        match an empty string at the end of the subject); however the checking for
664        an empty string was starting at the "start of match" point. This has been
665        changed to the "earliest inspected character" point, because the returned
666        data for a partial match starts at this character. This means that, for
667        example, /(?<=abc)def/ gives a partial match for the subject "abc"
668        (previously it gave "no match").
669    
670    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
671        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
672        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
673        has an implication that the given string is incomplete (because a partial
674        match is preferred over a full match). For this reason, these items now
675        give a partial match in this situation. [Aside: previously, the one case
676        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
677        match rather than a full match, which was wrong by the old rules, but is
678        now correct.]
679    
680    6.  There was a bug in the handling of #-introduced comments, recognized when
681        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
682        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
683        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
684        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
685        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
686        places in pcre_compile().
687    
688    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
689        comments when looking ahead for named forward references to subpatterns,
690        the only newline sequence it recognized was NL. It now handles newlines
691        according to the set newline convention.
692    
693    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
694        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
695        cater for a lack of strerror(). These oversights have been fixed.
696    
697    9.  Added --match-limit and --recursion-limit to pcregrep.
698    
699    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
700    
701    11. When the -o option was used, pcregrep was setting a return code of 1, even
702        when matches were found, and --line-buffered was not being honoured.
703    
704    12. Added an optional parentheses number to the -o and --only-matching options
705        of pcregrep.
706    
707    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
708        can match an empty string. The code to do it in pcretest and pcredemo
709        needed fixing:
710    
711        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
712            only one byte after an empty string match just before CRLF (this case
713            just got forgotten; "any" and "anycrlf" were OK).
714    
715        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
716            mode when an empty string match preceded an ASCII character followed by
717            a non-ASCII character. (The code for advancing by one character rather
718            than one byte was nonsense.)
719    
720        (c) The pcredemo.c sample program did not have any code at all to handle
721            the cases when CRLF is a valid newline sequence.
722    
723    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
724        as a starting offset was within the subject string. There is now a new
725        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
726        negative or greater than the length of the string. In order to test this,
727        pcretest is extended to allow the setting of negative starting offsets.
728    
729    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
730        starting offset points to the beginning of a UTF-8 character was
731        unnecessarily clumsy. I tidied it up.
732    
733    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
734        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
735    
736    17. Nobody had reported that the --include_dir option, which was added in
737        release 7.7 should have been called --include-dir (hyphen, not underscore)
738        for compatibility with GNU grep. I have changed it to --include-dir, but
739        left --include_dir as an undocumented synonym, and the same for
740        --exclude-dir, though that is not available in GNU grep, at least as of
741        release 2.5.4.
742    
743    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
744        characters from a string of bytes) have been redefined so as not to use
745        loops, in order to improve performance in some environments. At the same
746        time, I abstracted some of the common code into auxiliary macros to save
747        repetition (this should not affect the compiled code).
748    
749    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
750        compile-time error is now given if \c is not followed by an ASCII
751        character, that is, a byte less than 128. (In EBCDIC mode, the code is
752        different, and any byte value is allowed.)
753    
754    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
755        START_OPTIMIZE option, which is now allowed at compile time - but just
756        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
757        to pcregrep and other applications that have no direct access to PCRE
758        options. The new /Y option in pcretest sets this option when calling
759        pcre_compile().
760    
761    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
762        back references. Groups containing recursive back references were forced to
763        be atomic by that change, but in the case of named groups, the amount of
764        memory required was incorrectly computed, leading to "Failed: internal
765        error: code overflow". This has been fixed.
766    
767    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
768        pcretest.c, to avoid build problems in some Borland environments.
769    
770    
771    Version 8.10 25-Jun-2010
772    ------------------------
773    
774    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
775        THEN.
776    
777    2.  (*ACCEPT) was not working when inside an atomic group.
778    
779    3.  Inside a character class, \B is treated as a literal by default, but
780        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
781        causes the error). The code is unchanged, but I tidied the documentation.
782    
783    4.  Inside a character class, PCRE always treated \R and \X as literals,
784        whereas Perl faults them if its -w option is set. I have changed PCRE so
785        that it faults them when PCRE_EXTRA is set.
786    
787    5.  Added support for \N, which always matches any character other than
788        newline. (It is the same as "." when PCRE_DOTALL is not set.)
789    
790    6.  When compiling pcregrep with newer versions of gcc which may have
791        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
792        declared with attribute warn_unused_result" were given. Just casting the
793        result to (void) does not stop the warnings; a more elaborate fudge is
794        needed. I've used a macro to implement this.
795    
796    7.  Minor change to pcretest.c to avoid a compiler warning.
797    
798    8.  Added four artifical Unicode properties to help with an option to make
799        \s etc use properties (see next item). The new properties are: Xan
800        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
801    
802    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
803        use Unicode properties. (*UCP) at the start of a pattern can be used to set
804        this option. Modified pcretest to add /W to test this facility. Added
805        REG_UCP to make it available via the POSIX interface.
806    
807    10. Added --line-buffered to pcregrep.
808    
809    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
810        studied, and the match started with a letter with a code point greater than
811        127 whose first byte was different to the first byte of the other case of
812        the letter, the other case of this starting letter was not recognized
813        (#976).
814    
815    12. If a pattern that was studied started with a repeated Unicode property
816        test, for example, \p{Nd}+, there was the theoretical possibility of
817        setting up an incorrect bitmap of starting bytes, but fortunately it could
818        not have actually happened in practice until change 8 above was made (it
819        added property types that matched character-matching opcodes).
820    
821    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
822        possible starting bytes for non-anchored patterns.
823    
824    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
825        \R, and also a number of cases that involve Unicode properties, both
826        explicit and implicit when PCRE_UCP is set.
827    
828    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
829        input, it could crash or give wrong results if characters with values
830        greater than 0xc0 were present in the subject string. (Detail: it assumed
831        UTF-8 input when processing these items.)
832    
833    16. Added a lot of (int) casts to avoid compiler warnings in systems where
834        size_t is 64-bit (#991).
835    
836    17. Added a check for running out of memory when PCRE is compiled with
837        --disable-stack-for-recursion (#990).
838    
839    18. If the last data line in a file for pcretest does not have a newline on
840        the end, a newline was missing in the output.
841    
842    19. The default pcre_chartables.c file recognizes only ASCII characters (values
843        less than 128) in its various bitmaps. However, there is a facility for
844        generating tables according to the current locale when PCRE is compiled. It
845        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
846        characters, are recognized by isspace() and therefore were getting set in
847        these tables, and indeed these tables seem to approximate to ISO 8859. This
848        caused a problem in UTF-8 mode when pcre_study() was used to create a list
849        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
850        which of course cannot start UTF-8 characters. I have changed the code so
851        that only real ASCII characters (less than 128) and the correct starting
852        bytes for UTF-8 encodings are set for characters greater than 127 when in
853        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
854        altogether.)
855    
856    20. Added the /T option to pcretest so as to be able to run tests with non-
857        standard character tables, thus making it possible to include the tests
858        used for 19 above in the standard set of tests.
859    
860    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
861        reference to a subpattern the other side of a comment that contains an
862        opening parenthesis caused either an internal compiling error, or a
863        reference to the wrong subpattern.
864    
865    
866    Version 8.02 19-Mar-2010
867    ------------------------
868    
869    1.  The Unicode data tables have been updated to Unicode 5.2.0.
870    
871    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
872        configured.
873    
874    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
875        original author of that file, following a query about its status.
876    
877    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
878        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
879    
880    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
881        quantifier applied to a forward-referencing subroutine call, could compile
882        incorrect code or give the error "internal error: previously-checked
883        referenced subpattern not found".
884    
885    6.  Both MS Visual Studio and Symbian OS have problems with initializing
886        variables to point to external functions. For these systems, therefore,
887        pcre_malloc etc. are now initialized to local functions that call the
888        relevant global functions.
889    
890    7.  There were two entries missing in the vectors called coptable and poptable
891        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
892        I've fixed the data, and added a kludgy way of testing at compile time that
893        the lengths are correct (equal to the number of opcodes).
894    
895    8.  Following on from 7, I added a similar kludge to check the length of the
896        eint vector in pcreposix.c.
897    
898    9.  Error texts for pcre_compile() are held as one long string to avoid too
899        much relocation at load time. To find a text, the string is searched,
900        counting zeros. There was no check for running off the end of the string,
901        which could happen if a new error number was added without updating the
902        string.
903    
904    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
905    
906    11. \K was not working if it appeared in an atomic group or in a group that
907        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
908        \K is "not well defined" if used in an assertion. PCRE now accepts it if
909        the assertion is positive, but not if it is negative.
910    
911    12. Change 11 fortuitously reduced the size of the stack frame used in the
912        "match()" function of pcre_exec.c by one pointer. Forthcoming
913        implementation of support for (*MARK) will need an extra pointer on the
914        stack; I have reserved it now, so that the stack frame size does not
915        decrease.
916    
917    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
918        item in branch that calls a recursion is a subroutine call - as in the
919        second branch in the above example - was incorrectly given the compile-
920        time error "recursive call could loop indefinitely" because pcre_compile()
921        was not correctly checking the subroutine for matching a non-empty string.
922    
923    14. The checks for overrunning compiling workspace could trigger after an
924        overrun had occurred. This is a "should never occur" error, but it can be
925        triggered by pathological patterns such as hundreds of nested parentheses.
926        The checks now trigger 100 bytes before the end of the workspace.
927    
928    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
929    
930    
931    Version 8.01 19-Jan-2010
932    ------------------------
933    
934    1.  If a pattern contained a conditional subpattern with only one branch (in
935        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
936        computed the wrong minimum data length (which is of course zero for such
937        subpatterns). This could cause incorrect "no match" results.
938    
939    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
940        the pattern is reset in the first branch, pcre_compile() failed with
941        "internal error: code overflow at offset...". This happened only when
942        the reset was to the original external option setting. (An optimization
943        abstracts leading options settings into an external setting, which was the
944        cause of this.)
945    
946    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
947        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
948        assertion pattern did not match (meaning that the assertion was true), it
949        was incorrectly treated as false if the SKIP had been reached during the
950        matching. This also applied to assertions used as conditions.
951    
952    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
953        assertion subpattern, including such a pattern used as a condition,
954        unpredictable results occurred, instead of the error return
955        PCRE_ERROR_DFA_UITEM.
956    
957    5.  The C++ GlobalReplace function was not working like Perl for the special
958        situation when an empty string is matched. It now does the fancy magic
959        stuff that is necessary.
960    
961    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
962        removed. (These were left over from very, very early versions of PCRE.)
963    
964    7.  Some cosmetic changes to the code to make life easier when compiling it
965        as part of something else:
966    
967        (a) Change DEBUG to PCRE_DEBUG.
968    
969        (b) In pcre_compile(), rename the member of the "branch_chain" structure
970            called "current" as "current_branch", to prevent a collision with the
971            Linux macro when compiled as a kernel module.
972    
973        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
974            prevent a collision with the Linux macro when compiled as a kernel
975            module.
976    
977    8.  In pcre_compile() there are some checks for integer overflows that used to
978        cast potentially large values to (double). This has been changed to that
979        when building, a check for int64_t is made, and if it is found, it is used
980        instead, thus avoiding the use of floating point arithmetic. (There is no
981        other use of FP in PCRE.) If int64_t is not found, the fallback is to
982        double.
983    
984    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
985        2005 (difference between two addresses compared to an unsigned value).
986    
987    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
988        custom one, because of the following reported problem in Windows:
989    
990          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
991              under Win32.
992          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
993              therefore missing the function definition.
994          - The compiler thus generates a "C" signature for the test function.
995          - The linker fails to find the "C" function.
996          - PCRE fails to configure if asked to do so against libbz2.
997    
998    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
999        messages were output:
1000    
1001          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1002          rerunning libtoolize, to keep the correct libtool macros in-tree.
1003          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1004    
1005        I have done both of these things.
1006    
1007    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1008        most of the time, it *can* run out if it is given a pattern that contains a
1009        runaway infinite recursion. I updated the discussion in the pcrestack man
1010        page.
1011    
1012    13. Now that we have gone to the x.xx style of version numbers, the minor
1013        version may start with zero. Using 08 or 09 is a bad idea because users
1014        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1015        interpreted as invalid octal numbers. I've updated the previous comment in
1016        configure.ac, and also added a check that gives an error if 08 or 09 are
1017        used.
1018    
1019    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1020        causing partial matching to fail when the end of the subject matched \W
1021        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1022    
1023    15. There were some discrepancies between the declarations in pcre_internal.h
1024        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1025        their definitions. The declarations used "const uschar *" and the
1026        definitions used USPTR. Even though USPTR is normally defined as "const
1027        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1028        reported that: "This difference in casting confuses some C++ compilers, for
1029        example, SunCC recognizes above declarations as different functions and
1030        generates broken code for hbpcre." I have changed the declarations to use
1031        USPTR.
1032    
1033    16. GNU libtool is named differently on some systems. The autogen.sh script now
1034        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1035        (FreeBSD).
1036    
1037    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1038        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1039        comment: "Figure out how to create a longlong from a string: strtoll and
1040        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1041        instance, but it only takes 2 args instead of 3!"
1042    
1043    18. A subtle bug concerned with back references has been fixed by a change of
1044        specification, with a corresponding code fix. A pattern such as
1045        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1046        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1047        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1048        same bug. Such groups have to be quantified to be useful, or contained
1049        inside another quantified group. (If there's no repetition, the reference
1050        can never match.) The problem arises because, having left the group and
1051        moved on to the rest of the pattern, a later failure that backtracks into
1052        the group uses the captured value from the final iteration of the group
1053        rather than the correct earlier one. I have fixed this in PCRE by forcing
1054        any group that contains a reference to itself to be an atomic group; that
1055        is, there cannot be any backtracking into it once it has completed. This is
1056        similar to recursive and subroutine calls.
1057    
1058    
1059    Version 8.00 19-Oct-09
1060  ----------------------  ----------------------
1061    
1062  1.  The table for translating pcre_compile() error codes into POSIX error codes  1.  The table for translating pcre_compile() error codes into POSIX error codes
1063      was out-of-date, and there was no check on the pcre_compile() error code      was out-of-date, and there was no check on the pcre_compile() error code
1064      being within the table. This could lead to an OK return being given in      being within the table. This could lead to an OK return being given in
1065      error.      error.
1066    
1067  2.  Changed the call to open a subject file in pcregrep from fopen(pathname,  2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1068      "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests      "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1069      in a Windows environment.      in a Windows environment.
1070    
1071  3.  The pcregrep --count option prints the count for each file even when it is  3.  The pcregrep --count option prints the count for each file even when it is
1072      zero, as does GNU grep. However, pcregrep was also printing all files when      zero, as does GNU grep. However, pcregrep was also printing all files when
1073      --files-with-matches was added. Now, when both options are given, it prints      --files-with-matches was added. Now, when both options are given, it prints
1074      counts only for those files that have at least one match. (GNU grep just      counts only for those files that have at least one match. (GNU grep just
1075      prints the file name in this circumstance, but including the count seems      prints the file name in this circumstance, but including the count seems
1076      more useful - otherwise, why use --count?) Also ensured that the      more useful - otherwise, why use --count?) Also ensured that the
1077      combination -clh just lists non-zero counts, with no names.      combination -clh just lists non-zero counts, with no names.
1078    
1079  4.  The long form of the pcregrep -F option was incorrectly implemented as  4.  The long form of the pcregrep -F option was incorrectly implemented as
1080      --fixed_strings instead of --fixed-strings. This is an incompatible change,      --fixed_strings instead of --fixed-strings. This is an incompatible change,
1081      but it seems right to fix it, and I didn't think it was worth preserving      but it seems right to fix it, and I didn't think it was worth preserving
1082      the old behaviour.      the old behaviour.
1083    
1084  5.  The command line items --regex=pattern and --regexp=pattern were not  5.  The command line items --regex=pattern and --regexp=pattern were not
1085      recognized by pcregrep, which required --regex pattern or --regexp pattern      recognized by pcregrep, which required --regex pattern or --regexp pattern
1086      (with a space rather than an '='). The man page documented the '=' forms,      (with a space rather than an '='). The man page documented the '=' forms,
1087      which are compatible with GNU grep; these now work.      which are compatible with GNU grep; these now work.
1088    
1089    6.  No libpcreposix.pc file was created for pkg-config; there was just
1090        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1091    
1092    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1093        when UCP support is not needed, by modifying the Python script that
1094        generates it from Unicode data files. This should not matter if the module
1095        is correctly used as a library, but I received one complaint about 50K of
1096        unwanted data. My guess is that the person linked everything into his
1097        program rather than using a library. Anyway, it does no harm.
1098    
1099    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1100        was a minimum greater than 1 for a wide character in a possessive
1101        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1102        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1103        character. Chaos in the form of incorrect output or a compiling loop could
1104        result.
1105    
1106    9.  The restrictions on what a pattern can contain when partial matching is
1107        requested for pcre_exec() have been removed. All patterns can now be
1108        partially matched by this function. In addition, if there are at least two
1109        slots in the offset vector, the offset of the earliest inspected character
1110        for the match and the offset of the end of the subject are set in them when
1111        PCRE_ERROR_PARTIAL is returned.
1112    
1113    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1114        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1115        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1116        and may be more useful for multi-segment matching.
1117    
1118    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1119        used to be given if ever the end of the subject was reached; now it is
1120        given only if matching could not proceed because another character was
1121        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1122        string "Z", which now yields "no match" instead of "partial match". In the
1123        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1124        final character ended with (*FAIL).
1125    
1126    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1127        if the pattern had a "must contain" character that was already found in the
1128        earlier partial match, unless partial matching was again requested. For
1129        example, with the pattern /dog.(body)?/, the "must contain" character is
1130        "g". If the first part-match was for the string "dog", restarting with
1131        "sbody" failed. This bug has been fixed.
1132    
1133    13. The string returned by pcre_dfa_exec() after a partial match has been
1134        changed so that it starts at the first inspected character rather than the
1135        first character of the match. This makes a difference only if the pattern
1136        starts with a lookbehind assertion or \b or \B (\K is not supported by
1137        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1138        matching functions compatible, and I think it's the right thing to do.
1139    
1140    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1141        so that the demonstration program is easily available in environments where
1142        PCRE has not been installed from source.
1143    
1144    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1145        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1146        library.
1147    
1148    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1149        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1150        is not the first non-POSIX option to be added. Clearly some people find
1151        these options useful.
1152    
1153    17. If a caller to the POSIX matching function regexec() passes a non-zero
1154        value for nmatch with a NULL value for pmatch, the value of
1155        nmatch is forced to zero.
1156    
1157    18. RunGrepTest did not have a test for the availability of the -u option of
1158        the diff command, as RunTest does. It now checks in the same way as
1159        RunTest, and also checks for the -b option.
1160    
1161    19. If an odd number of negated classes containing just a single character
1162        interposed, within parentheses, between a forward reference to a named
1163        subpattern and the definition of the subpattern, compilation crashed with
1164        an internal error, complaining that it could not find the referenced
1165        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1166        [The bug was that it was starting one character too far in when skipping
1167        over the character class, thus treating the ] as data rather than
1168        terminating the class. This meant it could skip too much.]
1169    
1170    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1171        /g option in pcretest when the pattern contains \K, which makes it possible
1172        to have an empty string match not at the start, even when the pattern is
1173        anchored. Updated pcretest and pcredemo to use this option.
1174    
1175    21. If the maximum number of capturing subpatterns in a recursion was greater
1176        than the maximum at the outer level, the higher number was returned, but
1177        with unset values at the outer level. The correct (outer level) value is
1178        now given.
1179    
1180    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1181        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1182        make it do so. The string so far is captured, making this feature
1183        compatible with Perl.
1184    
1185    23. The tests have been re-organized, adding tests 11 and 12, to make it
1186        possible to check the Perl 5.10 features against Perl 5.10.
1187    
1188    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1189        pattern matches a fixed length string. PCRE did not allow this; now it
1190        does. Neither allows recursion.
1191    
1192    25. I finally figured out how to implement a request to provide the minimum
1193        length of subject string that was needed in order to match a given pattern.
1194        (It was back references and recursion that I had previously got hung up
1195        on.) This code has now been added to pcre_study(); it finds a lower bound
1196        to the length of subject needed. It is not necessarily the greatest lower
1197        bound, but using it to avoid searching strings that are too short does give
1198        some useful speed-ups. The value is available to calling programs via
1199        pcre_fullinfo().
1200    
1201    26. While implementing 25, I discovered to my embarrassment that pcretest had
1202        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1203        study optimizations had never been tested with that matching function.
1204        Oops. What is worse, even when it was passed study data, there was a bug in
1205        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1206        were also very few tests of studied patterns with pcre_dfa_exec().
1207    
1208    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1209        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1210        on the other side of the coin, they are no longer allowed to have different
1211        names, because these cannot be distinguished in PCRE, and this has caused
1212        confusion. (This is a difference from Perl.)
1213    
1214    28. When duplicate subpattern names are present (necessarily with different
1215        numbers, as required by 27 above), and a test is made by name in a
1216        conditional pattern, either for a subpattern having been matched, or for
1217        recursion in such a pattern, all the associated numbered subpatterns are
1218        tested, and the overall condition is true if the condition is true for any
1219        one of them. This is the way Perl works, and is also more like the way
1220        testing by number works.
1221    
1222    
1223  Version 7.9 11-Apr-09  Version 7.9 11-Apr-09
1224  ---------------------  ---------------------

Legend:
Removed from v.422  
changed lines
  Added in v.937

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12