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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.172  
changed lines
  Added in v.957

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12