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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.208  
changed lines
  Added in v.949

  ViewVC Help
Powered by ViewVC 1.1.5