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

Diff of /code/trunk/ChangeLog

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

revision 45 by nigel, Sat Feb 24 21:39:25 2007 UTC revision 976 by ph10, Sat Jun 16 17:53:17 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.31 02-June-2012
5    -------------------------
6    
7    1.  Fixing a wrong JIT test case and some compiler warnings.
8    
9    2.  Removed a bashism from the RunTest script.
10    
11    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
12        to unsigned type, result still unsigned" that was given by an MS compiler
13        on encountering the code "-sizeof(xxx)".
14    
15    4.  Partial matching support is added to the JIT compiler.
16    
17    5.  Fixed several bugs concerned with partial matching of items that consist
18        of more than one character:
19    
20        (a) /^(..)\1/ did not partially match "aba" because checking references was
21            done on an "all or nothing" basis. This also applied to repeated
22            references.
23    
24        (b) \R did not give a hard partial match if \r was found at the end of the
25            subject.
26    
27        (c) \X did not give a hard partial match after matching one or more
28            characters at the end of the subject.
29    
30        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
31            a partial match for the string "\r".
32    
33        (e) When newline was set to CRLF, the metacharacter "." did not recognize
34            a partial match for a CR character at the end of the subject string.
35    
36    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
37        running pcretest, the text "(JIT)" added to the output whenever JIT is
38        actually used to run the match.
39    
40    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
41        or /S+[+] with a digit between 1 and 7.
42    
43    8.  OP_NOT now supports any UTF character not just single-byte ones.
44    
45    9.  (*MARK) control verb is now supported by the JIT compiler.
46    
47    10. The command "./RunTest list" lists the available tests without actually
48        running any of them. (Because I keep forgetting what they all are.)
49    
50    11. Add PCRE_INFO_MAXLOOKBEHIND.
51    
52    12. Applied a (slightly modified) user-supplied patch that improves performance
53        when the heap is used for recursion (compiled with --disable-stack-for-
54        recursion). Instead of malloc and free for each heap frame each time a
55        logical recursion happens, frames are retained on a chain and re-used where
56        possible. This sometimes gives as much as 30% improvement.
57    
58    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
59        call.
60    
61    14. As documented, (*COMMIT) is now confined to within a positive assertion.
62    
63    15. It is now possible to link pcretest with libedit as an alternative to
64        libreadline.
65    
66    16. (*COMMIT) control verb is now supported by the JIT compiler.
67    
68    17. The Unicode data tables have been updated to Unicode 6.1.0.
69    
70    18. Added --file-list option to pcregrep.
71    
72    19. Added binary file support to pcregrep, including the -a, --binary-files,
73        -I, and --text options.
74    
75    20. The madvise function is renamed for posix_madvise for QNX compatibility
76        reasons. Fixed by Giuseppe D'Angelo.
77    
78    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
79    
80    22. Changed the option for creating symbolic links for 16-bit man pages from
81        -s to -sf so that re-installing does not cause issues.
82    
83    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
84    
85    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
86        matches in certain environments (the workspace was not being correctly
87        retained). Also added to pcre_dfa_exec() a simple plausibility check on
88        some of the workspace data at the beginning of a restart.
89    
90    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
91        was not doing so when it should - probably a typo introduced by SVN 528
92        (change 8.10/14).
93    
94    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
95        \w+ when the character tables indicated that \x{c4} was a word character.
96        There were several related cases, all because the tests for doing a table
97        lookup were testing for characters less than 127 instead of 255.
98    
99    27. If a pattern contains capturing parentheses that are not used in a match,
100        their slots in the ovector are set to -1. For those that are higher than
101        any matched groups, this happens at the end of processing. In the case when
102        there were back references that the ovector was too small to contain
103        (causing temporary malloc'd memory to be used during matching), and the
104        highest capturing number was not used, memory off the end of the ovector
105        was incorrectly being set to -1. (It was using the size of the temporary
106        memory instead of the true size.)
107    
108    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
109        ovector size, it uses memory at the end of the block that it has got.
110    
111    29. Check for an overlong MARK name and give an error at compile time. The
112        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
113    
114    30. JIT compiler update.
115    
116    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
117        Rill for the patch.
118    
119    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
120    
121    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
122    
123    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
124        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
125        was enabled.
126    
127    35. Improve JIT code generation for greedy plus quantifier.
128    
129    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
130        1 to "aa" instead of to an empty string. The bug affected repeated groups
131        that could potentially match an empty string.
132    
133    
134    Version 8.30 04-February-2012
135    -----------------------------
136    
137    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
138        name is defined in ctype.h.
139    
140    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
141        only in quite long subpatterns.
142    
143    3.  Removed the function pcre_info(), which has been obsolete and deprecated
144        since it was replaced by pcre_fullinfo() in February 2000.
145    
146    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
147        match a (*MARK), and the match failed at the start of the subject, a
148        reference to memory before the start of the subject could occur. This bug
149        was introduced by fix 17 of release 8.21.
150    
151    5.  A reference to an unset group with zero minimum repetition was giving
152        totally wrong answers (in non-JavaScript-compatibility mode). For example,
153        /(another)?(\1?)test/ matched against "hello world test". This bug was
154        introduced in release 8.13.
155    
156    6.  Add support for 16-bit character strings (a large amount of work involving
157        many changes and refactorings).
158    
159    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
160        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
161        from a file.
162    
163    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
164        rounding is not applied in this particular case).
165    
166    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
167        if they appear, or are escaped, in patterns.
168    
169    10. Get rid of a number of -Wunused-but-set-variable warnings.
170    
171    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
172        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
173        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
174        also returns the mark "x". This bug applied to capturing parentheses,
175        non-capturing parentheses, and atomic parentheses. It also applied to some
176        assertions.
177    
178    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
179        information out of configure.ac instead of relying on pcre.h.generic, which
180        is not stored in the repository.
181    
182    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
183        -lreadline.
184    
185    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
186    
187    15. Applied Graycode's patch to put the top-level frame on the stack rather
188        than the heap when not using the stack for recursion. This gives a
189        performance improvement in many cases when recursion is not deep.
190    
191    16. Experimental code added to "pcretest -C" to output the stack frame size.
192    
193    
194    Version 8.21 12-Dec-2011
195    ------------------------
196    
197    1.  Updating the JIT compiler.
198    
199    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
200        are added as well.
201    
202    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
203        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
204        calling _pcre_jit_exec. Some extra comments are added.
205    
206    4.  (*MARK) settings inside atomic groups that do not contain any capturing
207        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
208        was introduced by change 18 for 8.20.
209    
210    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
211        ECMA-262 standard.
212    
213    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
214        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
215        This bug was probably introduced by change 9 of 8.13.
216    
217    7.  While fixing 6 above, I noticed that a number of other items were being
218        incorrectly rejected as "not fixed length". This arose partly because newer
219        opcodes had not been added to the fixed-length checking code. I have (a)
220        corrected the bug and added tests for these items, and (b) arranged for an
221        error to occur if an unknown opcode is encountered while checking for fixed
222        length instead of just assuming "not fixed length". The items that were
223        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
224        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
225        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
226    
227    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
228        being incorrectly compiled and would have given unpredicatble results.
229    
230    9.  A possessively repeated subpattern with minimum repeat count greater than
231        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
232        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
233        the first (A) could occur when it should not.
234    
235    10. Add a cast and remove a redundant test from the code.
236    
237    11. JIT should use pcre_malloc/pcre_free for allocation.
238    
239    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
240        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
241        is anything other than /usr, -L is still shown).
242    
243    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
244    
245    14. Perl does not support \N without a following name in a [] class; PCRE now
246        also gives an error.
247    
248    15. If a forward reference was repeated with an upper limit of around 2000,
249        it caused the error "internal error: overran compiling workspace". The
250        maximum number of forward references (including repeats) was limited by the
251        internal workspace, and dependent on the LINK_SIZE. The code has been
252        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
253        the default depends on LINK_SIZE. There is a new upper limit (for safety)
254        of around 200,000 forward references. While doing this, I also speeded up
255        the filling in of repeated forward references.
256    
257    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
258        incorrectly expecting the subject to contain another "a" after the start.
259    
260    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
261        in the match, the SKIP should be ignored. This was not happening; instead
262        the SKIP was being treated as NOMATCH. For patterns such as
263        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
264        tested.
265    
266    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
267        now much more compatible with Perl, in particular in cases where the result
268        is a non-match for a non-anchored pattern. For example, if
269        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
270        "m", where previously it did not return a name. A side effect of this
271        change is that for partial matches, the last encountered mark name is
272        returned, as for non matches. A number of tests that were previously not
273        Perl-compatible have been moved into the Perl-compatible test files. The
274        refactoring has had the pleasing side effect of removing one argument from
275        the match() function, thus reducing its stack requirements.
276    
277    19. If the /S+ option was used in pcretest to study a pattern using JIT,
278        subsequent uses of /S (without +) incorrectly behaved like /S+.
279    
280    21. Retrieve executable code size support for the JIT compiler and fixing
281        some warnings.
282    
283    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
284        not work when the shorter character appeared right at the end of the
285        subject string.
286    
287    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
288        systems.
289    
290    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
291        output it when the /M option is used in pcretest.
292    
293    25. The CheckMan script was not being included in the distribution. Also, added
294        an explicit "perl" to run Perl scripts from the PrepareRelease script
295        because this is reportedly needed in Windows.
296    
297    26. If study data was being save in a file and studying had not found a set of
298        "starts with" bytes for the pattern, the data written to the file (though
299        never used) was taken from uninitialized memory and so caused valgrind to
300        complain.
301    
302    27. Updated RunTest.bat as provided by Sheri Pierce.
303    
304    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
305    
306    29. Computation of memory usage for the table of capturing group names was
307        giving an unnecessarily large value.
308    
309    
310    Version 8.20 21-Oct-2011
311    ------------------------
312    
313    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
314        a POSIX class. After further experiments with Perl, which convinced me that
315        Perl has bugs and confusions, a closing square bracket is no longer allowed
316        in a POSIX name. This bug also affected patterns with classes that started
317        with full stops.
318    
319    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
320        captured substring, but while checking the failing first alternative,
321        substring 1 is temporarily captured. If the output vector supplied to
322        pcre_exec() was not big enough for this capture, the yield of the function
323        was still zero ("insufficient space for captured substrings"). This cannot
324        be totally fixed without adding another stack variable, which seems a lot
325        of expense for a edge case. However, I have improved the situation in cases
326        such as /(a)(b)x|abc/ matched against "abc", where the return code
327        indicates that fewer than the maximum number of slots in the ovector have
328        been set.
329    
330    3.  Related to (2) above: when there are more back references in a pattern than
331        slots in the output vector, pcre_exec() uses temporary memory during
332        matching, and copies in the captures as far as possible afterwards. It was
333        using the entire output vector, but this conflicts with the specification
334        that only 2/3 is used for passing back captured substrings. Now it uses
335        only the first 2/3, for compatibility. This is, of course, another edge
336        case.
337    
338    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
339        main code base, and can be used by building with --enable-jit. When this is
340        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
341        runtime --no-jit option is given.
342    
343    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
344        ovector, the return from the function was zero, implying that there were
345        other matches that did not fit. The correct "exactly full" value is now
346        returned.
347    
348    6.  If a subpattern that was called recursively or as a subroutine contained
349        (*PRUNE) or any other control that caused it to give a non-standard return,
350        invalid errors such as "Error -26 (nested recursion at the same subject
351        position)" or even infinite loops could occur.
352    
353    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
354        computing the minimum length on reaching *ACCEPT, and so ended up with the
355        wrong value of 1 rather than 0. Further investigation indicates that
356        computing a minimum subject length in the presence of *ACCEPT is difficult
357        (think back references, subroutine calls), and so I have changed the code
358        so that no minimum is registered for a pattern that contains *ACCEPT.
359    
360    8.  If (*THEN) was present in the first (true) branch of a conditional group,
361        it was not handled as intended. [But see 16 below.]
362    
363    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
364        Sheri Pierce.
365    
366    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
367        the first byte in a match must be "a".
368    
369    11. Change 17 for 8.13 increased the recursion depth for patterns like
370        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
371        pattern contains any instances of (*THEN). If it does not, the old
372        optimizations are restored. It would be nice to do this on a per-group
373        basis, but at the moment that is not feasible.
374    
375    12. In some environments, the output of pcretest -C is CRLF terminated. This
376        broke RunTest's code that checks for the link size. A single white space
377        character after the value is now allowed for.
378    
379    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
380        For "fr", it uses the Windows-specific input and output files.
381    
382    14. If (*THEN) appeared in a group that was called recursively or as a
383        subroutine, it did not work as intended. [But see next item.]
384    
385    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
386        pattern fragments (but not containing any | characters). If A and B are
387        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
388        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
389        D. In other words, Perl considers parentheses that do not contain any |
390        characters to be part of a surrounding alternative, whereas PCRE was
391        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
392        differently. PCRE now behaves in the same way as Perl, except in the case
393        of subroutine/recursion calls such as (?1) which have in any case always
394        been different (but PCRE had them first :-).
395    
396    16. Related to 15 above: Perl does not treat the | in a conditional group as
397        creating alternatives. Such a group is treated in the same way as an
398        ordinary group without any | characters when processing (*THEN). PCRE has
399        been changed to match Perl's behaviour.
400    
401    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
402        RunGrepTest script failed.
403    
404    18. Change 22 for version 13 caused atomic groups to use more stack. This is
405        inevitable for groups that contain captures, but it can lead to a lot of
406        stack use in large patterns. The old behaviour has been restored for atomic
407        groups that do not contain any capturing parentheses.
408    
409    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
410        suppress the check for a minimum subject length at run time. (If it was
411        given to pcre_exec() or pcre_dfa_exec() it did work.)
412    
413    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
414        fail to work when decoding hex characters in data strings in EBCDIC
415        environments.
416    
417    21. It appears that in at least one Mac OS environment, the isxdigit() function
418        is implemented as a macro that evaluates to its argument more than once,
419        contravening the C 90 Standard (I haven't checked a later standard). There
420        was an instance in pcretest which caused it to go wrong when processing
421        \x{...} escapes in subject strings. The has been rewritten to avoid using
422        things like p++ in the argument of isxdigit().
423    
424    
425    Version 8.13 16-Aug-2011
426    ------------------------
427    
428    1.  The Unicode data tables have been updated to Unicode 6.0.0.
429    
430    2.  Two minor typos in pcre_internal.h have been fixed.
431    
432    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
433        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
434        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
435    
436    4.  There were a number of related bugs in the code for matching backrefences
437        caselessly in UTF-8 mode when codes for the characters concerned were
438        different numbers of bytes. For example, U+023A and U+2C65 are an upper
439        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
440        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
441        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
442        2-byte code at the end of the subject (it thought there wasn't enough data
443        left).
444    
445    5.  Comprehensive information about what went wrong is now returned by
446        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
447        as the output vector has at least 2 elements. The offset of the start of
448        the failing character and a reason code are placed in the vector.
449    
450    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
451        now returned is for the first byte of the failing character, instead of the
452        last byte inspected. This is an incompatible change, but I hope it is small
453        enough not to be a problem. It makes the returned offset consistent with
454        pcre_exec() and pcre_dfa_exec().
455    
456    7.  pcretest now gives a text phrase as well as the error number when
457        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
458        failure, the offset and reason code are output.
459    
460    8.  When \R was used with a maximizing quantifier it failed to skip backwards
461        over a \r\n pair if the subsequent match failed. Instead, it just skipped
462        back over a single character (\n). This seems wrong (because it treated the
463        two characters as a single entity when going forwards), conflicts with the
464        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
465        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
466        has been changed.
467    
468    9.  Some internal refactoring has changed the processing so that the handling
469        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
470        time (the PCRE_DOTALL option was changed this way some time ago: version
471        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
472        which was always a bit of a fudge. It also means that there is one less
473        argument for the match() function, which reduces its stack requirements
474        slightly. This change also fixes an incompatibility with Perl: the pattern
475        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
476    
477    10. More internal refactoring has drastically reduced the number of recursive
478        calls to match() for possessively repeated groups such as (abc)++ when
479        using pcre_exec().
480    
481    11. While implementing 10, a number of bugs in the handling of groups were
482        discovered and fixed:
483    
484        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
485        (a|)*(?1) gave a compile-time internal error.
486        ((a|)+)+  did not notice that the outer group could match an empty string.
487        (^a|^)+   was not marked as anchored.
488        (.*a|.*)+ was not marked as matching at start or after a newline.
489    
490    12. Yet more internal refactoring has removed another argument from the match()
491        function. Special calls to this function are now indicated by setting a
492        value in a variable in the "match data" data block.
493    
494    13. Be more explicit in pcre_study() instead of relying on "default" for
495        opcodes that mean there is no starting character; this means that when new
496        ones are added and accidentally left out of pcre_study(), testing should
497        pick them up.
498    
499    14. The -s option of pcretest has been documented for ages as being an old
500        synonym of -m (show memory usage). I have changed it to mean "force study
501        for every regex", that is, assume /S for every regex. This is similar to -i
502        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
503        using it. It makes it easier to run collections of tests with and without
504        study enabled, and thereby test pcre_study() more easily. All the standard
505        tests are now run with and without -s (but some patterns can be marked as
506        "never study" - see 20 below).
507    
508    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
509        restoration of the capturing data to the outer values was not happening
510        correctly.
511    
512    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
513        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
514        pattern had matched an empty string, and so incorrectly returned a no
515        match.
516    
517    17. There was optimizing code for the last branch of non-capturing parentheses,
518        and also for the obeyed branch of a conditional subexpression, which used
519        tail recursion to cut down on stack usage. Unfortunately, now that there is
520        the possibility of (*THEN) occurring in these branches, tail recursion is
521        no longer possible because the return has to be checked for (*THEN). These
522        two optimizations have therefore been removed. [But see 8.20/11 above.]
523    
524    18. If a pattern containing \R was studied, it was assumed that \R always
525        matched two bytes, thus causing the minimum subject length to be
526        incorrectly computed because \R can also match just one byte.
527    
528    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
529        was incorrectly computed.
530    
531    20. If /S is present twice on a test pattern in pcretest input, it now
532        *disables* studying, thereby overriding the use of -s on the command line
533        (see 14 above). This is necessary for one or two tests to keep the output
534        identical in both cases.
535    
536    21. When (*ACCEPT) was used in an assertion that matched an empty string and
537        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
538    
539    22. When an atomic group that contained a capturing parenthesis was
540        successfully matched, but the branch in which it appeared failed, the
541        capturing was not being forgotten if a higher numbered group was later
542        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
543        group 1 to "a", when in fact it should be unset. This applied to multi-
544        branched capturing and non-capturing groups, repeated or not, and also to
545        positive assertions (capturing in negative assertions does not happen
546        in PCRE) and also to nested atomic groups.
547    
548    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
549        subject after a captured substring, to make it easier to tell which of a
550        number of identical substrings has been captured.
551    
552    24. The way atomic groups are processed by pcre_exec() has been changed so that
553        if they are repeated, backtracking one repetition now resets captured
554        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
555        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
556        "aaa". Previously, it would have been "a". As part of this code
557        refactoring, the way recursive calls are handled has also been changed.
558    
559    25. If an assertion condition captured any substrings, they were not passed
560        back unless some other capturing happened later. For example, if
561        (?(?=(a))a) was matched against "a", no capturing was returned.
562    
563    26. When studying a pattern that contained subroutine calls or assertions,
564        the code for finding the minimum length of a possible match was handling
565        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
566        group 1 called group 2 while simultaneously a separate group 2 called group
567        1). A stack overflow occurred in this case. I have fixed this by limiting
568        the recursion depth to 10.
569    
570    27. Updated RunTest.bat in the distribution to the version supplied by Tom
571        Fortmann. This supports explicit test numbers on the command line, and has
572        argument validation and error reporting.
573    
574    28. An instance of \X with an unlimited repeat could fail if at any point the
575        first character it looked at was a mark character.
576    
577    29. Some minor code refactoring concerning Unicode properties and scripts
578        should reduce the stack requirement of match() slightly.
579    
580    30. Added the '=' option to pcretest to check the setting of unused capturing
581        slots at the end of the pattern, which are documented as being -1, but are
582        not included in the return count.
583    
584    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
585        compiled something random. Now it gives a compile-time error (as does
586        Perl).
587    
588    32. A *MARK encountered during the processing of a positive assertion is now
589        recorded and passed back (compatible with Perl).
590    
591    33. If --only-matching or --colour was set on a pcregrep call whose pattern
592        had alternative anchored branches, the search for a second match in a line
593        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
594        matched the line "0102" twice. The same bug affected patterns that started
595        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
596        twice.
597    
598    34. Previously, PCRE did not allow quantification of assertions. However, Perl
599        does, and because of capturing effects, quantifying parenthesized
600        assertions may at times be useful. Quantifiers are now allowed for
601        parenthesized assertions.
602    
603    35. A minor code tidy in pcre_compile() when checking options for \R usage.
604    
605    36. \g was being checked for fancy things in a character class, when it should
606        just be a literal "g".
607    
608    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
609        appearance of a nested POSIX class supersedes an apparent external class.
610        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
611        unescaped square brackets may also appear as part of class names. For
612        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
613        more like Perl. (But see 8.20/1 above.)
614    
615    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
616        was because it thought it was \N{name}, which is not supported).
617    
618    39. Add minix to OS list not supporting the -S option in pcretest.
619    
620    40. PCRE tries to detect cases of infinite recursion at compile time, but it
621        cannot analyze patterns in sufficient detail to catch mutual recursions
622        such as ((?1))((?2)). There is now a runtime test that gives an error if a
623        subgroup is called recursively as a subpattern for a second time at the
624        same position in the subject string. In previous releases this might have
625        been caught by the recursion limit, or it might have run out of stack.
626    
627    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
628        happen only once. PCRE was, however incorrectly giving a compile time error
629        "recursive call could loop indefinitely" because it cannot analyze the
630        pattern in sufficient detail. The compile time test no longer happens when
631        PCRE is compiling a conditional subpattern, but actual runaway loops are
632        now caught at runtime (see 40 above).
633    
634    42. It seems that Perl allows any characters other than a closing parenthesis
635        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
636        has been changed to be the same.
637    
638    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
639        as not to get warnings when autogen.sh is called. Also changed
640        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
641    
642    44. To help people who use pcregrep to scan files containing exceedingly long
643        lines, the following changes have been made:
644    
645        (a) The default value of the buffer size parameter has been increased from
646            8K to 20K. (The actual buffer used is three times this size.)
647    
648        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
649            PCRE is built.
650    
651        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
652            to be set at run time.
653    
654        (d) Numerical values in pcregrep options can be followed by K or M, for
655            example --buffer-size=50K.
656    
657        (e) If a line being scanned overflows pcregrep's buffer, an error is now
658            given and the return code is set to 2.
659    
660    45. Add a pointer to the latest mark to the callout data block.
661    
662    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
663        partial match of an empty string instead of no match. This was specific to
664        the use of ".".
665    
666    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
667        complete match instead of a partial match. This bug was dependent on both
668        the PCRE_UTF8 and PCRE_DOTALL options being set.
669    
670    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
671        starting byte set, because \b was not being ignored.
672    
673    
674    Version 8.12 15-Jan-2011
675    ------------------------
676    
677    1.  Fixed some typos in the markup of the man pages, and wrote a script that
678        checks for such things as part of the documentation building process.
679    
680    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
681        --match-limit and --recursion-limit options (added for 8.11). In
682        particular, this made one of the standard tests fail. (The integer value
683        went into the wrong half of a long int.)
684    
685    3.  If the --colour option was given to pcregrep with -v (invert match), it
686        did strange things, either producing crazy output, or crashing. It should,
687        of course, ignore a request for colour when reporting lines that do not
688        match.
689    
690    4.  Another pcregrep bug caused similar problems if --colour was specified with
691        -M (multiline) and the pattern match finished with a line ending.
692    
693    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
694        matched in multiline mode, the following line was shown as part of the
695        match. This seems wrong, so I have changed it.
696    
697    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
698        the check for further matches in the same line (so they could be coloured)
699        to overrun the end of the current line. If another match was found, it was
700        incorrectly shown (and then shown again when found in the next line).
701    
702    7.  If pcregrep was compiled under Windows, there was a reference to the
703        function pcregrep_exit() before it was defined. I am assuming this was
704        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
705        reported by a user. I've moved the definition above the reference.
706    
707    
708    Version 8.11 10-Dec-2010
709    ------------------------
710    
711    1.  (*THEN) was not working properly if there were untried alternatives prior
712        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
713        backtracked to try for "b" instead of moving to the next alternative branch
714        at the same level (in this case, to look for "c"). The Perl documentation
715        is clear that when (*THEN) is backtracked onto, it goes to the "next
716        alternative in the innermost enclosing group".
717    
718    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
719        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
720        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
721        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
722        (*THEN).
723    
724    3.  If \s appeared in a character class, it removed the VT character from
725        the class, even if it had been included by some previous item, for example
726        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
727        of \s, but is part of the POSIX "space" class.)
728    
729    4.  A partial match never returns an empty string (because you can always
730        match an empty string at the end of the subject); however the checking for
731        an empty string was starting at the "start of match" point. This has been
732        changed to the "earliest inspected character" point, because the returned
733        data for a partial match starts at this character. This means that, for
734        example, /(?<=abc)def/ gives a partial match for the subject "abc"
735        (previously it gave "no match").
736    
737    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
738        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
739        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
740        has an implication that the given string is incomplete (because a partial
741        match is preferred over a full match). For this reason, these items now
742        give a partial match in this situation. [Aside: previously, the one case
743        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
744        match rather than a full match, which was wrong by the old rules, but is
745        now correct.]
746    
747    6.  There was a bug in the handling of #-introduced comments, recognized when
748        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
749        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
750        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
751        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
752        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
753        places in pcre_compile().
754    
755    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
756        comments when looking ahead for named forward references to subpatterns,
757        the only newline sequence it recognized was NL. It now handles newlines
758        according to the set newline convention.
759    
760    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
761        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
762        cater for a lack of strerror(). These oversights have been fixed.
763    
764    9.  Added --match-limit and --recursion-limit to pcregrep.
765    
766    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
767    
768    11. When the -o option was used, pcregrep was setting a return code of 1, even
769        when matches were found, and --line-buffered was not being honoured.
770    
771    12. Added an optional parentheses number to the -o and --only-matching options
772        of pcregrep.
773    
774    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
775        can match an empty string. The code to do it in pcretest and pcredemo
776        needed fixing:
777    
778        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
779            only one byte after an empty string match just before CRLF (this case
780            just got forgotten; "any" and "anycrlf" were OK).
781    
782        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
783            mode when an empty string match preceded an ASCII character followed by
784            a non-ASCII character. (The code for advancing by one character rather
785            than one byte was nonsense.)
786    
787        (c) The pcredemo.c sample program did not have any code at all to handle
788            the cases when CRLF is a valid newline sequence.
789    
790    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
791        as a starting offset was within the subject string. There is now a new
792        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
793        negative or greater than the length of the string. In order to test this,
794        pcretest is extended to allow the setting of negative starting offsets.
795    
796    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
797        starting offset points to the beginning of a UTF-8 character was
798        unnecessarily clumsy. I tidied it up.
799    
800    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
801        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
802    
803    17. Nobody had reported that the --include_dir option, which was added in
804        release 7.7 should have been called --include-dir (hyphen, not underscore)
805        for compatibility with GNU grep. I have changed it to --include-dir, but
806        left --include_dir as an undocumented synonym, and the same for
807        --exclude-dir, though that is not available in GNU grep, at least as of
808        release 2.5.4.
809    
810    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
811        characters from a string of bytes) have been redefined so as not to use
812        loops, in order to improve performance in some environments. At the same
813        time, I abstracted some of the common code into auxiliary macros to save
814        repetition (this should not affect the compiled code).
815    
816    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
817        compile-time error is now given if \c is not followed by an ASCII
818        character, that is, a byte less than 128. (In EBCDIC mode, the code is
819        different, and any byte value is allowed.)
820    
821    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
822        START_OPTIMIZE option, which is now allowed at compile time - but just
823        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
824        to pcregrep and other applications that have no direct access to PCRE
825        options. The new /Y option in pcretest sets this option when calling
826        pcre_compile().
827    
828    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
829        back references. Groups containing recursive back references were forced to
830        be atomic by that change, but in the case of named groups, the amount of
831        memory required was incorrectly computed, leading to "Failed: internal
832        error: code overflow". This has been fixed.
833    
834    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
835        pcretest.c, to avoid build problems in some Borland environments.
836    
837    
838    Version 8.10 25-Jun-2010
839    ------------------------
840    
841    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
842        THEN.
843    
844    2.  (*ACCEPT) was not working when inside an atomic group.
845    
846    3.  Inside a character class, \B is treated as a literal by default, but
847        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
848        causes the error). The code is unchanged, but I tidied the documentation.
849    
850    4.  Inside a character class, PCRE always treated \R and \X as literals,
851        whereas Perl faults them if its -w option is set. I have changed PCRE so
852        that it faults them when PCRE_EXTRA is set.
853    
854    5.  Added support for \N, which always matches any character other than
855        newline. (It is the same as "." when PCRE_DOTALL is not set.)
856    
857    6.  When compiling pcregrep with newer versions of gcc which may have
858        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
859        declared with attribute warn_unused_result" were given. Just casting the
860        result to (void) does not stop the warnings; a more elaborate fudge is
861        needed. I've used a macro to implement this.
862    
863    7.  Minor change to pcretest.c to avoid a compiler warning.
864    
865    8.  Added four artifical Unicode properties to help with an option to make
866        \s etc use properties (see next item). The new properties are: Xan
867        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
868    
869    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
870        use Unicode properties. (*UCP) at the start of a pattern can be used to set
871        this option. Modified pcretest to add /W to test this facility. Added
872        REG_UCP to make it available via the POSIX interface.
873    
874    10. Added --line-buffered to pcregrep.
875    
876    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
877        studied, and the match started with a letter with a code point greater than
878        127 whose first byte was different to the first byte of the other case of
879        the letter, the other case of this starting letter was not recognized
880        (#976).
881    
882    12. If a pattern that was studied started with a repeated Unicode property
883        test, for example, \p{Nd}+, there was the theoretical possibility of
884        setting up an incorrect bitmap of starting bytes, but fortunately it could
885        not have actually happened in practice until change 8 above was made (it
886        added property types that matched character-matching opcodes).
887    
888    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
889        possible starting bytes for non-anchored patterns.
890    
891    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
892        \R, and also a number of cases that involve Unicode properties, both
893        explicit and implicit when PCRE_UCP is set.
894    
895    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
896        input, it could crash or give wrong results if characters with values
897        greater than 0xc0 were present in the subject string. (Detail: it assumed
898        UTF-8 input when processing these items.)
899    
900    16. Added a lot of (int) casts to avoid compiler warnings in systems where
901        size_t is 64-bit (#991).
902    
903    17. Added a check for running out of memory when PCRE is compiled with
904        --disable-stack-for-recursion (#990).
905    
906    18. If the last data line in a file for pcretest does not have a newline on
907        the end, a newline was missing in the output.
908    
909    19. The default pcre_chartables.c file recognizes only ASCII characters (values
910        less than 128) in its various bitmaps. However, there is a facility for
911        generating tables according to the current locale when PCRE is compiled. It
912        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
913        characters, are recognized by isspace() and therefore were getting set in
914        these tables, and indeed these tables seem to approximate to ISO 8859. This
915        caused a problem in UTF-8 mode when pcre_study() was used to create a list
916        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
917        which of course cannot start UTF-8 characters. I have changed the code so
918        that only real ASCII characters (less than 128) and the correct starting
919        bytes for UTF-8 encodings are set for characters greater than 127 when in
920        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
921        altogether.)
922    
923    20. Added the /T option to pcretest so as to be able to run tests with non-
924        standard character tables, thus making it possible to include the tests
925        used for 19 above in the standard set of tests.
926    
927    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
928        reference to a subpattern the other side of a comment that contains an
929        opening parenthesis caused either an internal compiling error, or a
930        reference to the wrong subpattern.
931    
932    
933    Version 8.02 19-Mar-2010
934    ------------------------
935    
936    1.  The Unicode data tables have been updated to Unicode 5.2.0.
937    
938    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
939        configured.
940    
941    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
942        original author of that file, following a query about its status.
943    
944    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
945        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
946    
947    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
948        quantifier applied to a forward-referencing subroutine call, could compile
949        incorrect code or give the error "internal error: previously-checked
950        referenced subpattern not found".
951    
952    6.  Both MS Visual Studio and Symbian OS have problems with initializing
953        variables to point to external functions. For these systems, therefore,
954        pcre_malloc etc. are now initialized to local functions that call the
955        relevant global functions.
956    
957    7.  There were two entries missing in the vectors called coptable and poptable
958        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
959        I've fixed the data, and added a kludgy way of testing at compile time that
960        the lengths are correct (equal to the number of opcodes).
961    
962    8.  Following on from 7, I added a similar kludge to check the length of the
963        eint vector in pcreposix.c.
964    
965    9.  Error texts for pcre_compile() are held as one long string to avoid too
966        much relocation at load time. To find a text, the string is searched,
967        counting zeros. There was no check for running off the end of the string,
968        which could happen if a new error number was added without updating the
969        string.
970    
971    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
972    
973    11. \K was not working if it appeared in an atomic group or in a group that
974        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
975        \K is "not well defined" if used in an assertion. PCRE now accepts it if
976        the assertion is positive, but not if it is negative.
977    
978    12. Change 11 fortuitously reduced the size of the stack frame used in the
979        "match()" function of pcre_exec.c by one pointer. Forthcoming
980        implementation of support for (*MARK) will need an extra pointer on the
981        stack; I have reserved it now, so that the stack frame size does not
982        decrease.
983    
984    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
985        item in branch that calls a recursion is a subroutine call - as in the
986        second branch in the above example - was incorrectly given the compile-
987        time error "recursive call could loop indefinitely" because pcre_compile()
988        was not correctly checking the subroutine for matching a non-empty string.
989    
990    14. The checks for overrunning compiling workspace could trigger after an
991        overrun had occurred. This is a "should never occur" error, but it can be
992        triggered by pathological patterns such as hundreds of nested parentheses.
993        The checks now trigger 100 bytes before the end of the workspace.
994    
995    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
996    
997    
998    Version 8.01 19-Jan-2010
999    ------------------------
1000    
1001    1.  If a pattern contained a conditional subpattern with only one branch (in
1002        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1003        computed the wrong minimum data length (which is of course zero for such
1004        subpatterns). This could cause incorrect "no match" results.
1005    
1006    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1007        the pattern is reset in the first branch, pcre_compile() failed with
1008        "internal error: code overflow at offset...". This happened only when
1009        the reset was to the original external option setting. (An optimization
1010        abstracts leading options settings into an external setting, which was the
1011        cause of this.)
1012    
1013    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1014        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1015        assertion pattern did not match (meaning that the assertion was true), it
1016        was incorrectly treated as false if the SKIP had been reached during the
1017        matching. This also applied to assertions used as conditions.
1018    
1019    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1020        assertion subpattern, including such a pattern used as a condition,
1021        unpredictable results occurred, instead of the error return
1022        PCRE_ERROR_DFA_UITEM.
1023    
1024    5.  The C++ GlobalReplace function was not working like Perl for the special
1025        situation when an empty string is matched. It now does the fancy magic
1026        stuff that is necessary.
1027    
1028    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1029        removed. (These were left over from very, very early versions of PCRE.)
1030    
1031    7.  Some cosmetic changes to the code to make life easier when compiling it
1032        as part of something else:
1033    
1034        (a) Change DEBUG to PCRE_DEBUG.
1035    
1036        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1037            called "current" as "current_branch", to prevent a collision with the
1038            Linux macro when compiled as a kernel module.
1039    
1040        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1041            prevent a collision with the Linux macro when compiled as a kernel
1042            module.
1043    
1044    8.  In pcre_compile() there are some checks for integer overflows that used to
1045        cast potentially large values to (double). This has been changed to that
1046        when building, a check for int64_t is made, and if it is found, it is used
1047        instead, thus avoiding the use of floating point arithmetic. (There is no
1048        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1049        double.
1050    
1051    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1052        2005 (difference between two addresses compared to an unsigned value).
1053    
1054    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1055        custom one, because of the following reported problem in Windows:
1056    
1057          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1058              under Win32.
1059          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1060              therefore missing the function definition.
1061          - The compiler thus generates a "C" signature for the test function.
1062          - The linker fails to find the "C" function.
1063          - PCRE fails to configure if asked to do so against libbz2.
1064    
1065    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1066        messages were output:
1067    
1068          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1069          rerunning libtoolize, to keep the correct libtool macros in-tree.
1070          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1071    
1072        I have done both of these things.
1073    
1074    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1075        most of the time, it *can* run out if it is given a pattern that contains a
1076        runaway infinite recursion. I updated the discussion in the pcrestack man
1077        page.
1078    
1079    13. Now that we have gone to the x.xx style of version numbers, the minor
1080        version may start with zero. Using 08 or 09 is a bad idea because users
1081        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1082        interpreted as invalid octal numbers. I've updated the previous comment in
1083        configure.ac, and also added a check that gives an error if 08 or 09 are
1084        used.
1085    
1086    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1087        causing partial matching to fail when the end of the subject matched \W
1088        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1089    
1090    15. There were some discrepancies between the declarations in pcre_internal.h
1091        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1092        their definitions. The declarations used "const uschar *" and the
1093        definitions used USPTR. Even though USPTR is normally defined as "const
1094        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1095        reported that: "This difference in casting confuses some C++ compilers, for
1096        example, SunCC recognizes above declarations as different functions and
1097        generates broken code for hbpcre." I have changed the declarations to use
1098        USPTR.
1099    
1100    16. GNU libtool is named differently on some systems. The autogen.sh script now
1101        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1102        (FreeBSD).
1103    
1104    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1105        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1106        comment: "Figure out how to create a longlong from a string: strtoll and
1107        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1108        instance, but it only takes 2 args instead of 3!"
1109    
1110    18. A subtle bug concerned with back references has been fixed by a change of
1111        specification, with a corresponding code fix. A pattern such as
1112        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1113        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1114        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1115        same bug. Such groups have to be quantified to be useful, or contained
1116        inside another quantified group. (If there's no repetition, the reference
1117        can never match.) The problem arises because, having left the group and
1118        moved on to the rest of the pattern, a later failure that backtracks into
1119        the group uses the captured value from the final iteration of the group
1120        rather than the correct earlier one. I have fixed this in PCRE by forcing
1121        any group that contains a reference to itself to be an atomic group; that
1122        is, there cannot be any backtracking into it once it has completed. This is
1123        similar to recursive and subroutine calls.
1124    
1125    
1126    Version 8.00 19-Oct-09
1127    ----------------------
1128    
1129    1.  The table for translating pcre_compile() error codes into POSIX error codes
1130        was out-of-date, and there was no check on the pcre_compile() error code
1131        being within the table. This could lead to an OK return being given in
1132        error.
1133    
1134    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1135        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1136        in a Windows environment.
1137    
1138    3.  The pcregrep --count option prints the count for each file even when it is
1139        zero, as does GNU grep. However, pcregrep was also printing all files when
1140        --files-with-matches was added. Now, when both options are given, it prints
1141        counts only for those files that have at least one match. (GNU grep just
1142        prints the file name in this circumstance, but including the count seems
1143        more useful - otherwise, why use --count?) Also ensured that the
1144        combination -clh just lists non-zero counts, with no names.
1145    
1146    4.  The long form of the pcregrep -F option was incorrectly implemented as
1147        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1148        but it seems right to fix it, and I didn't think it was worth preserving
1149        the old behaviour.
1150    
1151    5.  The command line items --regex=pattern and --regexp=pattern were not
1152        recognized by pcregrep, which required --regex pattern or --regexp pattern
1153        (with a space rather than an '='). The man page documented the '=' forms,
1154        which are compatible with GNU grep; these now work.
1155    
1156    6.  No libpcreposix.pc file was created for pkg-config; there was just
1157        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1158    
1159    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1160        when UCP support is not needed, by modifying the Python script that
1161        generates it from Unicode data files. This should not matter if the module
1162        is correctly used as a library, but I received one complaint about 50K of
1163        unwanted data. My guess is that the person linked everything into his
1164        program rather than using a library. Anyway, it does no harm.
1165    
1166    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1167        was a minimum greater than 1 for a wide character in a possessive
1168        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1169        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1170        character. Chaos in the form of incorrect output or a compiling loop could
1171        result.
1172    
1173    9.  The restrictions on what a pattern can contain when partial matching is
1174        requested for pcre_exec() have been removed. All patterns can now be
1175        partially matched by this function. In addition, if there are at least two
1176        slots in the offset vector, the offset of the earliest inspected character
1177        for the match and the offset of the end of the subject are set in them when
1178        PCRE_ERROR_PARTIAL is returned.
1179    
1180    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1181        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1182        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1183        and may be more useful for multi-segment matching.
1184    
1185    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1186        used to be given if ever the end of the subject was reached; now it is
1187        given only if matching could not proceed because another character was
1188        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1189        string "Z", which now yields "no match" instead of "partial match". In the
1190        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1191        final character ended with (*FAIL).
1192    
1193    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1194        if the pattern had a "must contain" character that was already found in the
1195        earlier partial match, unless partial matching was again requested. For
1196        example, with the pattern /dog.(body)?/, the "must contain" character is
1197        "g". If the first part-match was for the string "dog", restarting with
1198        "sbody" failed. This bug has been fixed.
1199    
1200    13. The string returned by pcre_dfa_exec() after a partial match has been
1201        changed so that it starts at the first inspected character rather than the
1202        first character of the match. This makes a difference only if the pattern
1203        starts with a lookbehind assertion or \b or \B (\K is not supported by
1204        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1205        matching functions compatible, and I think it's the right thing to do.
1206    
1207    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1208        so that the demonstration program is easily available in environments where
1209        PCRE has not been installed from source.
1210    
1211    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1212        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1213        library.
1214    
1215    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1216        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1217        is not the first non-POSIX option to be added. Clearly some people find
1218        these options useful.
1219    
1220    17. If a caller to the POSIX matching function regexec() passes a non-zero
1221        value for nmatch with a NULL value for pmatch, the value of
1222        nmatch is forced to zero.
1223    
1224    18. RunGrepTest did not have a test for the availability of the -u option of
1225        the diff command, as RunTest does. It now checks in the same way as
1226        RunTest, and also checks for the -b option.
1227    
1228    19. If an odd number of negated classes containing just a single character
1229        interposed, within parentheses, between a forward reference to a named
1230        subpattern and the definition of the subpattern, compilation crashed with
1231        an internal error, complaining that it could not find the referenced
1232        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1233        [The bug was that it was starting one character too far in when skipping
1234        over the character class, thus treating the ] as data rather than
1235        terminating the class. This meant it could skip too much.]
1236    
1237    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1238        /g option in pcretest when the pattern contains \K, which makes it possible
1239        to have an empty string match not at the start, even when the pattern is
1240        anchored. Updated pcretest and pcredemo to use this option.
1241    
1242    21. If the maximum number of capturing subpatterns in a recursion was greater
1243        than the maximum at the outer level, the higher number was returned, but
1244        with unset values at the outer level. The correct (outer level) value is
1245        now given.
1246    
1247    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1248        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1249        make it do so. The string so far is captured, making this feature
1250        compatible with Perl.
1251    
1252    23. The tests have been re-organized, adding tests 11 and 12, to make it
1253        possible to check the Perl 5.10 features against Perl 5.10.
1254    
1255    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1256        pattern matches a fixed length string. PCRE did not allow this; now it
1257        does. Neither allows recursion.
1258    
1259    25. I finally figured out how to implement a request to provide the minimum
1260        length of subject string that was needed in order to match a given pattern.
1261        (It was back references and recursion that I had previously got hung up
1262        on.) This code has now been added to pcre_study(); it finds a lower bound
1263        to the length of subject needed. It is not necessarily the greatest lower
1264        bound, but using it to avoid searching strings that are too short does give
1265        some useful speed-ups. The value is available to calling programs via
1266        pcre_fullinfo().
1267    
1268    26. While implementing 25, I discovered to my embarrassment that pcretest had
1269        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1270        study optimizations had never been tested with that matching function.
1271        Oops. What is worse, even when it was passed study data, there was a bug in
1272        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1273        were also very few tests of studied patterns with pcre_dfa_exec().
1274    
1275    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1276        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1277        on the other side of the coin, they are no longer allowed to have different
1278        names, because these cannot be distinguished in PCRE, and this has caused
1279        confusion. (This is a difference from Perl.)
1280    
1281    28. When duplicate subpattern names are present (necessarily with different
1282        numbers, as required by 27 above), and a test is made by name in a
1283        conditional pattern, either for a subpattern having been matched, or for
1284        recursion in such a pattern, all the associated numbered subpatterns are
1285        tested, and the overall condition is true if the condition is true for any
1286        one of them. This is the way Perl works, and is also more like the way
1287        testing by number works.
1288    
1289    
1290    Version 7.9 11-Apr-09
1291    ---------------------
1292    
1293    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1294        (pcretest), all targets were linked against these libraries. This included
1295        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1296        libraries. This caused unwanted dependencies to be created. This problem
1297        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1298        pcretest is linked with readline.
1299    
1300    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1301        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1302        moved outside it again, because FALSE and TRUE are already defined in AIX,
1303        but BOOL is not.
1304    
1305    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1306        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1307    
1308    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1309        hyphens) following file names and line numbers when outputting matching
1310        lines. This is not true; no spaces are inserted. I have also clarified the
1311        wording for the --colour (or --color) option.
1312    
1313    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1314        was not coloured; this is different to GNU grep, so I have changed it to be
1315        the same.
1316    
1317    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1318        each matching line was coloured. Now it goes on to look for further matches
1319        of any of the test patterns, which is the same behaviour as GNU grep.
1320    
1321    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1322        doesn't make sense to accept an empty string match in pcregrep, so I have
1323        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1324        seems to be how GNU grep behaves.
1325    
1326    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1327        start or after a newline", because the conditional assertion was not being
1328        correctly handled. The rule now is that both the assertion and what follows
1329        in the first alternative must satisfy the test.
1330    
1331    9.  If auto-callout was enabled in a pattern with a conditional group whose
1332        condition was an assertion, PCRE could crash during matching, both with
1333        pcre_exec() and pcre_dfa_exec().
1334    
1335    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1336        used for matching.
1337    
1338    11. Unicode property support in character classes was not working for
1339        characters (bytes) greater than 127 when not in UTF-8 mode.
1340    
1341    12. Added the -M command line option to pcretest.
1342    
1343    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1344    
1345    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1346    
1347    16. Added comments and documentation about mis-use of no_arg in the C++
1348        wrapper.
1349    
1350    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1351        from Martin Jerabek that uses macro names for all relevant character and
1352        string constants.
1353    
1354    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1355        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1356        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1357        these, but not everybody uses configure.
1358    
1359    19. A conditional group that had only one branch was not being correctly
1360        recognized as an item that could match an empty string. This meant that an
1361        enclosing group might also not be so recognized, causing infinite looping
1362        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1363        with the subject "ab", where knowledge that the repeated group can match
1364        nothing is needed in order to break the loop.
1365    
1366    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1367        exec(), but without supplying a callout function, matching went wrong.
1368    
1369    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1370        leak if the size of the offset vector was greater than 30. When the vector
1371        is smaller, the saved offsets during recursion go onto a local stack
1372        vector, but for larger vectors malloc() is used. It was failing to free
1373        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1374        error, in fact).
1375    
1376    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1377        heapframe that is used only when UTF-8 support is enabled. This caused no
1378        problem, but was untidy.
1379    
1380    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1381        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1382        included within another project.
1383    
1384    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1385        slightly modified by me:
1386    
1387          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1388              not building pcregrep.
1389    
1390          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1391              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1392    
1393    25. Forward references, both numeric and by name, in patterns that made use of
1394        duplicate group numbers, could behave incorrectly or give incorrect errors,
1395        because when scanning forward to find the reference group, PCRE was not
1396        taking into account the duplicate group numbers. A pattern such as
1397        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1398    
1399    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1400        the feature of a custom pointer more persuasive (as requested by a user).
1401    
1402    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1403        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1404        pre-defined.
1405    
1406    28. Added support for (*UTF8) at the start of a pattern.
1407    
1408    29. Arrange for flags added by the "release type" setting in CMake to be shown
1409        in the configuration summary.
1410    
1411    
1412    Version 7.8 05-Sep-08
1413    ---------------------
1414    
1415    1.  Replaced UCP searching code with optimized version as implemented for Ad
1416        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1417        stage table and inline lookup instead of a function, giving speed ups of 2
1418        to 5 times on some simple patterns that I tested. Permission was given to
1419        distribute the MultiStage2.py script that generates the tables (it's not in
1420        the tarball, but is in the Subversion repository).
1421    
1422    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1423        scripts.
1424    
1425    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1426        a group with a zero qualifier. The result of the study could be incorrect,
1427        or the function might crash, depending on the pattern.
1428    
1429    4.  Caseless matching was not working for non-ASCII characters in back
1430        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1431        It now works when Unicode Property Support is available.
1432    
1433    5.  In pcretest, an escape such as \x{de} in the data was always generating
1434        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1435        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1436        truncation.
1437    
1438    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1439    
1440    7.  Added two (int) casts to pcregrep when printing the difference of two
1441        pointers, in case they are 64-bit values.
1442    
1443    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1444        test 2 if it fails.
1445    
1446    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1447        and a #define of that name to empty if it is not externally set. This is to
1448        allow users of MSVC to set it if necessary.
1449    
1450    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1451        the convenience functions in the pcre_get.c source file.
1452    
1453    11. An option change at the start of a pattern that had top-level alternatives
1454        could cause overwriting and/or a crash. This command provoked a crash in
1455        some environments:
1456    
1457          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1458    
1459        This potential security problem was recorded as CVE-2008-2371.
1460    
1461    12. For a pattern where the match had to start at the beginning or immediately
1462        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1463        pcre_dfa_exec() could read past the end of the passed subject if there was
1464        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1465        pcretest so that it places the subject at the end of its malloc-ed buffer.
1466    
1467    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1468        exec() might read past the end of the data buffer in UTF-8 mode.
1469    
1470    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1471        the data contained the byte 0x85 as part of a UTF-8 character within its
1472        first line. This applied both to normal and DFA matching.
1473    
1474    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1475        /^[^d]*?$/8 failed to match "abc".
1476    
1477    16. Added a missing copyright notice to pcrecpp_internal.h.
1478    
1479    17. Make it more clear in the documentation that values returned from
1480        pcre_exec() in ovector are byte offsets, not character counts.
1481    
1482    18. Tidied a few places to stop certain compilers from issuing warnings.
1483    
1484    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1485        supplied by Stefan Weber. I made a further small update for 7.8 because
1486        there is a change of source arrangements: the pcre_searchfuncs.c module is
1487        replaced by pcre_ucd.c.
1488    
1489    
1490    Version 7.7 07-May-08
1491    ---------------------
1492    
1493    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1494        a string to a long long, pretend we don't even have a long long." This is
1495        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1496    
1497    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1498        pre-7.6 versions, which defined a global no_arg variable instead of putting
1499        it in the RE class. (See also #8 below.)
1500    
1501    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1502        Lopes.
1503    
1504    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1505    
1506        (1) The include/exclude patterns were being applied to the whole pathnames
1507            of files, instead of just to the final components.
1508    
1509        (2) If there was more than one level of directory, the subdirectories were
1510            skipped unless they satisfied the include/exclude conditions. This is
1511            inconsistent with GNU grep (and could even be seen as contrary to the
1512            pcregrep specification - which I improved to make it absolutely clear).
1513            The action now is always to scan all levels of directory, and just
1514            apply the include/exclude patterns to regular files.
1515    
1516    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1517        --exclude_dir in the tests to avoid scanning .svn directories.
1518    
1519    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1520        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1521        doesn't support NULs in patterns.
1522    
1523    7.  Added some missing "const"s to declarations of static tables in
1524        pcre_compile.c and pcre_dfa_exec.c.
1525    
1526    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1527        caused by fix #2  above. (Subsequently also a second patch to fix the
1528        first patch. And a third patch - this was a messy problem.)
1529    
1530    9.  Applied Craig's patch to remove the use of push_back().
1531    
1532    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1533        matching function regexec().
1534    
1535    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1536        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1537        references. PCRE supports relative numbers with this syntax (I don't think
1538        Oniguruma does).
1539    
1540    12. Previously, a group with a zero repeat such as (...){0} was completely
1541        omitted from the compiled regex. However, this means that if the group
1542        was called as a subroutine from elsewhere in the pattern, things went wrong
1543        (an internal error was given). Such groups are now left in the compiled
1544        pattern, with a new opcode that causes them to be skipped at execution
1545        time.
1546    
1547    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1548        to the way PCRE behaves:
1549    
1550        (a) A lone ] character is dis-allowed (Perl treats it as data).
1551    
1552        (b) A back reference to an unmatched subpattern matches an empty string
1553            (Perl fails the current match path).
1554    
1555        (c) A data ] in a character class must be notated as \] because if the
1556            first data character in a class is ], it defines an empty class. (In
1557            Perl it is not possible to have an empty class.) The empty class []
1558            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1559            The negative empty class [^] matches any one character, independently
1560            of the DOTALL setting.
1561    
1562    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1563        non-existent subpattern following a character class starting with ']' and
1564        containing () gave an internal compiling error instead of "reference to
1565        non-existent subpattern". Fortunately, when the pattern did exist, the
1566        compiled code was correct. (When scanning forwards to check for the
1567        existencd of the subpattern, it was treating the data ']' as terminating
1568        the class, so got the count wrong. When actually compiling, the reference
1569        was subsequently set up correctly.)
1570    
1571    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1572        it was being rejected as not supported by pcre_dfa_exec(), even though
1573        other assertions are supported. I have made pcre_dfa_exec() support
1574        (*FAIL).
1575    
1576    16. The implementation of 13c above involved the invention of a new opcode,
1577        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1578        cannot be changed at match time, I realized I could make a small
1579        improvement to matching performance by compiling OP_ALLANY instead of
1580        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1581        on the OP_ANY path.
1582    
1583    17. Compiling pcretest on Windows with readline support failed without the
1584        following two fixes: (1) Make the unistd.h include conditional on
1585        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1586    
1587    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1588        ncurses library to be included for pcretest when ReadLine support is
1589        requested, but also to allow for it to be overridden. This patch came from
1590        Daniel Bergström.
1591    
1592    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1593        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1594        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1595        spotting this.
1596    
1597    
1598    Version 7.6 28-Jan-08
1599    ---------------------
1600    
1601    1.  A character class containing a very large number of characters with
1602        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1603        overflow.
1604    
1605    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1606        HAVE_LONG_LONG is not defined.
1607    
1608    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1609        bring them up to date and include new features. This patch includes:
1610    
1611        - Fixed PH's badly added libz and libbz2 support.
1612        - Fixed a problem with static linking.
1613        - Added pcredemo. [But later removed - see 7 below.]
1614        - Fixed dftables problem and added an option.
1615        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1616            HAVE_LONG_LONG.
1617        - Added readline support for pcretest.
1618        - Added an listing of the option settings after cmake has run.
1619    
1620    4.  A user submitted a patch to Makefile that makes it easy to create
1621        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1622        Makefile.am that cause it to include this special target, without
1623        affecting anything else. Note that the same mingw target plus all
1624        the other distribution libraries and programs are now supported
1625        when configuring with CMake (see 6 below) instead of with
1626        Configure/Make.
1627    
1628    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1629        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1630        exported in the Windows port". It has not yet been confirmed that the patch
1631        solves the problem, but it does no harm.
1632    
1633    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1634        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1635        with CMake, and also correct the comment about stack recursion.
1636    
1637    7.  Remove the automatic building of pcredemo from the ./configure system and
1638        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1639        of a program that users should build themselves after PCRE is installed, so
1640        building it automatically is not really right. What is more, it gave
1641        trouble in some build environments.
1642    
1643    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1644    
1645    
1646    Version 7.5 10-Jan-08
1647    ---------------------
1648    
1649    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1650        values in parens when parsing an RE using the C++ wrapper."
1651    
1652    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1653        Characters greater than 255 were excluded from the class instead of being
1654        included.
1655    
1656    3.  The same bug as (2) above applied to negated POSIX classes such as
1657        [:^space:].
1658    
1659    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1660        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1661        I have changed it.
1662    
1663    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1664        first named subpattern) and a construct such as (?&a) would reference the
1665        first named subpattern whose name started with "a" (in other words, the
1666        length check was missing). Both these problems are fixed. "Subpattern name
1667        expected" is now given for (?&) (a zero-length name), and this patch also
1668        makes it give the same error for \k'' (previously it complained that that
1669        was a reference to a non-existent subpattern).
1670    
1671    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1672        this is right because (?- can be followed by option settings as well as by
1673        digits. I have, however, made the messages clearer.
1674    
1675    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1676        than the number used in the conditional) now cause a compile-time error.
1677        This is actually not compatible with Perl, which accepts such patterns, but
1678        treats the conditional as always being FALSE (as PCRE used to), but it
1679        seems to me that giving a diagnostic is better.
1680    
1681    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1682        and messages.
1683    
1684    9.  Fix two occurrences of "backslash" in comments that should have been
1685        "backspace".
1686    
1687    10. Remove two redundant lines of code that can never be obeyed (their function
1688        was moved elsewhere).
1689    
1690    11. The program that makes PCRE's Unicode character property table had a bug
1691        which caused it to generate incorrect table entries for sequences of
1692        characters that have the same character type, but are in different scripts.
1693        It amalgamated them into a single range, with the script of the first of
1694        them. In other words, some characters were in the wrong script. There were
1695        thirteen such cases, affecting characters in the following ranges:
1696    
1697          U+002b0 - U+002c1
1698          U+0060c - U+0060d
1699          U+0061e - U+00612
1700          U+0064b - U+0065e
1701          U+0074d - U+0076d
1702          U+01800 - U+01805
1703          U+01d00 - U+01d77
1704          U+01d9b - U+01dbf
1705          U+0200b - U+0200f
1706          U+030fc - U+030fe
1707          U+03260 - U+0327f
1708          U+0fb46 - U+0fbb1
1709          U+10450 - U+1049d
1710    
1711    12. The -o option (show only the matching part of a line) for pcregrep was not
1712        compatible with GNU grep in that, if there was more than one match in a
1713        line, it showed only the first of them. It now behaves in the same way as
1714        GNU grep.
1715    
1716    13. If the -o and -v options were combined for pcregrep, it printed a blank
1717        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1718        does the same. The return code can be used to tell if there were any
1719        non-matching lines.
1720    
1721    14. Added --file-offsets and --line-offsets to pcregrep.
1722    
1723    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1724        infinitely looping recursion. The bug was that positive lookaheads were not
1725        being skipped when checking for a possible empty match (negative lookaheads
1726        and both kinds of lookbehind were skipped).
1727    
1728    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1729        inclusion of <windows.h> to before rather than after the definition of
1730        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1731    
1732    17. Specifying a possessive quantifier with a specific limit for a Unicode
1733        character property caused pcre_compile() to compile bad code, which led at
1734        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1735        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1736        caused the error; without that there was no problem.
1737    
1738    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1739    
1740    19. Added --enable-pcretest-libreadline.
1741    
1742    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1743        RE::GlobalReplace(). As a result, the number of replacements returned was
1744        double what it should be. I removed one of the increments, but Craig sent a
1745        later patch that removed the other one (the right fix) and added unit tests
1746        that check the return values (which was not done before).
1747    
1748    21. Several CMake things:
1749    
1750        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1751            the names libpcre and libpcreposix, not just pcre and pcreposix.
1752    
1753        (2) The above change means that pcretest and pcregrep are now correctly
1754            linked with the newly-built libraries, not previously installed ones.
1755    
1756        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1757    
1758    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1759        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1760        UTF-8 newline character). The key issue is that the pattern starts .*;
1761        this means that the match must be either at the beginning, or after a
1762        newline. The bug was in the code for advancing after a failed match and
1763        checking that the new position followed a newline. It was not taking
1764        account of UTF-8 characters correctly.
1765    
1766    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1767        character classes. PCRE was not treating the sequence [:...:] as a
1768        character class unless the ... were all letters. Perl, however, seems to
1769        allow any characters between [: and :], though of course it rejects as
1770        unknown any "names" that contain non-letters, because all the known class
1771        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1772        for example, whereas PCRE did not - it did not recognize a POSIX character
1773        class. This seemed a bit dangerous, so the code has been changed to be
1774        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1775        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1776        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1777        Perl does, and where it didn't before.
1778    
1779    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1780        Windows environments %n is disabled by default.
1781    
1782    
1783    Version 7.4 21-Sep-07
1784    ---------------------
1785    
1786    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1787        means that a class such as [\s] counted as "explicit reference to CR or
1788        LF". That isn't really right - the whole point of the change was to try to
1789        help when there was an actual mention of one of the two characters. So now
1790        the change happens only if \r or \n (or a literal CR or LF) character is
1791        encountered.
1792    
1793    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1794        of both had grown to the point where there were only 3 bits left.
1795        Fortunately, there was spare space in the data structure, and so I have
1796        moved the internal flags into a new 16-bit field to free up more option
1797        bits.
1798    
1799    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1800        but did not set the internal JCHANGED flag - either of these is enough to
1801        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1802        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1803        start sets both bits.
1804    
1805    4.  Added options (at build time, compile time, exec time) to change \R from
1806        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1807    
1808    5.  doc/pcresyntax.html was missing from the distribution.
1809    
1810    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1811        compatibility, even though it is no longer used.
1812    
1813    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1814        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1815        windows.h file is present (where different names are used). [This was
1816        reversed later after testing - see 16 below.]
1817    
1818    8.  Changed all #include <config.h> to #include "config.h". There were also
1819        some further <pcre.h> cases that I changed to "pcre.h".
1820    
1821    9.  When pcregrep was used with the --colour option, it missed the line ending
1822        sequence off the lines that it output.
1823    
1824    10. It was pointed out to me that arrays of string pointers cause lots of
1825        relocations when a shared library is dynamically loaded. A technique of
1826        using a single long string with a table of offsets can drastically reduce
1827        these. I have refactored PCRE in four places to do this. The result is
1828        dramatic:
1829    
1830          Originally:                          290
1831          After changing UCP table:            187
1832          After changing error message table:   43
1833          After changing table of "verbs"       36
1834          After changing table of Posix names   22
1835    
1836        Thanks to the folks working on Gregex for glib for this insight.
1837    
1838    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1839        unicode-properties was also set.
1840    
1841    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1842    
1843    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1844        checked only for CRLF.
1845    
1846    14. Added casts to pcretest.c to avoid compiler warnings.
1847    
1848    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1849    
1850    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1851        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1852        entirely. This removes changes made in 7 above.
1853    
1854    17. The CMake files have been updated, and there is now more information about
1855        building with CMake in the NON-UNIX-USE document.
1856    
1857    
1858    Version 7.3 28-Aug-07
1859    ---------------------
1860    
1861     1. In the rejigging of the build system that eventually resulted in 7.1, the
1862        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1863        brackets there is not right, since it causes compilers to look for an
1864        installed pcre.h, not the version that is in the source that is being
1865        compiled (which of course may be different). I have changed it back to:
1866    
1867          #include "pcre.h"
1868    
1869        I have a vague recollection that the change was concerned with compiling in
1870        different directories, but in the new build system, that is taken care of
1871        by the VPATH setting the Makefile.
1872    
1873     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1874        when the subject happened to end in the byte 0x85 (e.g. if the last
1875        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1876        characters but of course it shouldn't be taken as a newline when it is part
1877        of another character. The bug was that, for an unlimited repeat of . in
1878        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1879        characters when looking for a newline.
1880    
1881     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1882    
1883     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1884        in debug output.
1885    
1886     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1887        long printing in the pcrecpp unittest when running under MinGW.
1888    
1889     6. ESC_K was left out of the EBCDIC table.
1890    
1891     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1892        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1893        limit also applies to "virtual nesting" when a pattern is recursive, and in
1894        this case 1000 isn't so big. I have been able to remove this limit at the
1895        expense of backing off one optimization in certain circumstances. Normally,
1896        when pcre_exec() would call its internal match() function recursively and
1897        immediately return the result unconditionally, it uses a "tail recursion"
1898        feature to save stack. However, when a subpattern that can match an empty
1899        string has an unlimited repetition quantifier, it no longer makes this
1900        optimization. That gives it a stack frame in which to save the data for
1901        checking that an empty string has been matched. Previously this was taken
1902        from the 1000-entry workspace that had been reserved. So now there is no
1903        explicit limit, but more stack is used.
1904    
1905     8. Applied Daniel's patches to solve problems with the import/export magic
1906        syntax that is required for Windows, and which was going wrong for the
1907        pcreposix and pcrecpp parts of the library. These were overlooked when this
1908        problem was solved for the main library.
1909    
1910     9. There were some crude static tests to avoid integer overflow when computing
1911        the size of patterns that contain repeated groups with explicit upper
1912        limits. As the maximum quantifier is 65535, the maximum group length was
1913        set at 30,000 so that the product of these two numbers did not overflow a
1914        32-bit integer. However, it turns out that people want to use groups that
1915        are longer than 30,000 bytes (though not repeat them that many times).
1916        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1917        made it possible to implement the integer overflow checks in a much more
1918        dynamic way, which I have now done. The artificial limitation on group
1919        length has been removed - we now have only the limit on the total length of
1920        the compiled pattern, which depends on the LINK_SIZE setting.
1921    
1922    10. Fixed a bug in the documentation for get/copy named substring when
1923        duplicate names are permitted. If none of the named substrings are set, the
1924        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1925        empty string.
1926    
1927    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1928        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1929        because the ] is interpreted as the first data character and the
1930        terminating ] is not found. PCRE has been made compatible with Perl in this
1931        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1932        cause memory overwriting.
1933    
1934    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1935        string has been matched (to stop an infinite loop). It was not recognizing
1936        a conditional subpattern that could match an empty string if that
1937        subpattern was within another subpattern. For example, it looped when
1938        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1939        condition was not nested. This bug has been fixed.
1940    
1941    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1942        past the start of the subject in the presence of bytes with the top bit
1943        set, for example "\x8aBCD".
1944    
1945    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1946        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1947    
1948    14. Optimized (?!) to (*FAIL).
1949    
1950    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1951        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1952        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1953        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1954        does: it's just the validity check that is more restrictive.
1955    
1956    16. Inserted checks for integer overflows during escape sequence (backslash)
1957        processing, and also fixed erroneous offset values for syntax errors during
1958        backslash processing.
1959    
1960    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1961        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1962    
1963    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1964        caused an overrun.
1965    
1966    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1967        something other than just ASCII characters) inside a group that had an
1968        unlimited repeat caused a loop at compile time (while checking to see
1969        whether the group could match an empty string).
1970    
1971    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1972        [\P{Any}] did so. (Error in the code for printing property names.)
1973    
1974    21. An orphan \E inside a character class could cause a crash.
1975    
1976    22. A repeated capturing bracket such as (A)? could cause a wild memory
1977        reference during compilation.
1978    
1979    23. There are several functions in pcre_compile() that scan along a compiled
1980        expression for various reasons (e.g. to see if it's fixed length for look
1981        behind). There were bugs in these functions when a repeated \p or \P was
1982        present in the pattern. These operators have additional parameters compared
1983        with \d, etc, and these were not being taken into account when moving along
1984        the compiled data. Specifically:
1985    
1986        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1987            length.
1988    
1989        (b) An item such as \pL+ within a repeated group could cause crashes or
1990            loops.
1991    
1992        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1993            "reference to non-existent subpattern" error.
1994    
1995        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1996    
1997    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1998        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1999    
2000    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2001    
2002    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2003        character were causing crashes (broken optimization).
2004    
2005    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2006        \p or \P) caused a compile-time loop.
2007    
2008    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2009        break. For example, the unstudied pattern [\r\n]A does not match the string
2010        "\r\nA" because change 7.0/46 below moves the current point on by two
2011        characters after failing to match at the start. However, the pattern \nA
2012        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2013        the same is true. There doesn't seem any very clean way out of this, but
2014        what I have chosen to do makes the common cases work: PCRE now takes note
2015        of whether there can be an explicit match for \r or \n anywhere in the
2016        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2017        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2018        pattern has explicit CR or LF references.
2019    
2020    29. Added (*CR) etc for changing newline setting at start of pattern.
2021    
2022    
2023    Version 7.2 19-Jun-07
2024    ---------------------
2025    
2026     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2027        which is apparently normally available under Windows.
2028    
2029     2. Re-jig the pcregrep tests with different newline settings in an attempt
2030        to make them independent of the local environment's newline setting.
2031    
2032     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2033    
2034     4. Some of the "internals" tests were previously cut out when the link size
2035        was not 2, because the output contained actual offsets. The recent new
2036        "Z" feature of pcretest means that these can be cut out, making the tests
2037        usable with all link sizes.
2038    
2039     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2040        stack recursion. This gives a massive performance boost under BSD, but just
2041        a small improvement under Linux. However, it saves one field in the frame
2042        in all cases.
2043    
2044     6. Added more features from the forthcoming Perl 5.10:
2045    
2046        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2047            recursion call. It refers to the nth most recently opened parentheses.
2048    
2049        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2050            to be opened parentheses.
2051    
2052        (c) Conditions that refer to capturing parentheses can be specified
2053            relatively, for example, (?(-2)... or (?(+3)...
2054    
2055        (d) \K resets the start of the current match so that everything before
2056            is not part of it.
2057    
2058        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2059    
2060        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2061            reference syntax.
2062    
2063        (g) (?| introduces a group in which the numbering of parentheses in each
2064            alternative starts with the same number.
2065    
2066        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2067    
2068     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2069        PCRE_INFO_JCHANGED.
2070    
2071     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2072        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2073        for detecting groups that can match an empty string.
2074    
2075     9. A pattern with a very large number of alternatives (more than several
2076        hundred) was running out of internal workspace during the pre-compile
2077        phase, where pcre_compile() figures out how much memory will be needed. A
2078        bit of new cunning has reduced the workspace needed for groups with
2079        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2080        workspace instead of running out of the 4096 that are available.
2081    
2082    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2083    
2084    11. Applied patch from Google to remove an optimization that didn't quite work.
2085        The report of the bug said:
2086    
2087          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2088          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2089          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2090    
2091    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2092        it matched the wrong number of bytes.
2093    
2094    
2095    Version 7.1 24-Apr-07
2096    ---------------------
2097    
2098     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2099        that is more "standard", making use of automake and other Autotools. There
2100        is some re-arrangement of the files and adjustment of comments consequent
2101        on this.
2102    
2103     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2104        for recursive directory scanning broke on some systems because the files
2105        are not scanned in any specific order and on different systems the order
2106        was different. A call to "sort" has been inserted into RunGrepTest for the
2107        approprate test as a short-term fix. In the longer term there may be an
2108        alternative.
2109    
2110     3. I had an email from Eric Raymond about problems translating some of PCRE's
2111        man pages to HTML (despite the fact that I distribute HTML pages, some
2112        people do their own conversions for various reasons). The problems
2113        concerned the use of low-level troff macros .br and .in. I have therefore
2114        removed all such uses from the man pages (some were redundant, some could
2115        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2116        HTML has been updated to handle .nf/.fi and to complain if it encounters
2117        .br or .in.
2118    
2119     4. Updated comments in configure.ac that get placed in config.h.in and also
2120        arranged for config.h to be included in the distribution, with the name
2121        config.h.generic, for the benefit of those who have to compile without
2122        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2123    
2124     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2125        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2126        makevp.bat for the current PCRE, using the additional files
2127        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2128    
2129     6. A Windows user reported a minor discrepancy with test 2, which turned out
2130        to be caused by a trailing space on an input line that had got lost in his
2131        copy. The trailing space was an accident, so I've just removed it.
2132    
2133     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2134        that is needed.
2135    
2136     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2137        as "const" (a) because they are and (b) because it helps the PHP
2138        maintainers who have recently made a script to detect big data structures
2139        in the php code that should be moved to the .rodata section. I remembered
2140        to update Builducptable as well, so it won't revert if ucptable.h is ever
2141        re-created.
2142    
2143     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2144        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2145        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2146        support is not required. This saves 1.5-2K of code, which is important in
2147        some applications.
2148    
2149        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2150        so as not to refer to the tables, even though these functions will never be
2151        called when UTF-8 support is disabled. Otherwise there are problems with a
2152        shared library.
2153    
2154    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2155    
2156        (a) It was defining its arguments as char * instead of void *.
2157    
2158        (b) It was assuming that all moves were upwards in memory; this was true
2159            a long time ago when I wrote it, but is no longer the case.
2160    
2161        The emulated memove() is provided for those environments that have neither
2162        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2163        is clearly not the case, as these two bugs were recently reported.
2164    
2165    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2166        and Detrail to create the HTML documentation, the .txt form of the man
2167        pages, and it removes trailing spaces from listed files. It also creates
2168        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2169        case, it wraps all the #defines with #ifndefs. This script should be run
2170        before "make dist".
2171    
2172    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2173        with Unicode property support.
2174    
2175        (a) For a maximizing quantifier, if the two different cases of the
2176            character were of different lengths in their UTF-8 codings (there are
2177            some cases like this - I found 11), and the matching function had to
2178            back up over a mixture of the two cases, it incorrectly assumed they
2179            were both the same length.
2180    
2181        (b) When PCRE was configured to use the heap rather than the stack for
2182            recursion during matching, it was not correctly preserving the data for
2183            the other case of a UTF-8 character when checking ahead for a match
2184            while processing a minimizing repeat. If the check also involved
2185            matching a wide character, but failed, corruption could cause an
2186            erroneous result when trying to check for a repeat of the original
2187            character.
2188    
2189    13. Some tidying changes to the testing mechanism:
2190    
2191        (a) The RunTest script now detects the internal link size and whether there
2192            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2193            values substituted by "configure". (The RunGrepTest script already did
2194            this for UTF-8.) The configure.ac script no longer substitutes the
2195            relevant variables.
2196    
2197        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2198            with length and offset values. This means that the output is different
2199            for different internal link sizes. Test 2 is skipped for link sizes
2200            other than 2 because of this, bypassing the problem. Unfortunately,
2201            there was also a test in test 3 (the locale tests) that used /B and
2202            failed for link sizes other than 2. Rather than cut the whole test out,
2203            I have added a new /Z option to pcretest that replaces the length and
2204            offset values with spaces. This is now used to make test 3 independent
2205            of link size. (Test 2 will be tidied up later.)
2206    
2207    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2208        segmentation fault instead of returning the appropriate error message.
2209    
2210    15. In multiline mode when the newline sequence was set to "any", the pattern
2211        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2212        This doesn't seem right; it now treats the CRLF combination as the line
2213        ending, and so does not match in that case. It's only a pattern such as ^$
2214        that would hit this one: something like ^ABC$ would have failed after \r
2215        and then tried again after \r\n.
2216    
2217    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2218        in an attempt to make files that differ only in their line terminators
2219        compare equal. This works on Linux.
2220    
2221    17. Under certain error circumstances pcregrep might try to free random memory
2222        as it exited. This is now fixed, thanks to valgrind.
2223    
2224    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2225        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2226        was because its rules for how to advance for /g after matching an empty
2227        string at the end of a line did not allow for this case. They now check for
2228        it specially.
2229    
2230    20. pcretest is supposed to handle patterns and data of any length, by
2231        extending its buffers when necessary. It was getting this wrong when the
2232        buffer for a data line had to be extended.
2233    
2234    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2235        CRLF as a newline sequence.
2236    
2237    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2238        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2239        I have nevertheless tidied it up.
2240    
2241    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2242    
2243    24. Added a man page for pcre-config.
2244    
2245    
2246    Version 7.0 19-Dec-06
2247    ---------------------
2248    
2249     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2250        moving to gcc 4.1.1.
2251    
2252     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2253        sys/time.h, which is documented as needed for this function. It doesn't
2254        seem to matter on Linux, but it showed up on some releases of OS X.
2255    
2256     3. It seems that there are systems where bytes whose values are greater than
2257        127 match isprint() in the "C" locale. The "C" locale should be the
2258        default when a C program starts up. In most systems, only ASCII printing
2259        characters match isprint(). This difference caused the output from pcretest
2260        to vary, making some of the tests fail. I have changed pcretest so that:
2261    
2262        (a) When it is outputting text in the compiled version of a pattern, bytes
2263            other than 32-126 are always shown as hex escapes.
2264    
2265        (b) When it is outputting text that is a matched part of a subject string,
2266            it does the same, unless a different locale has been set for the match
2267            (using the /L modifier). In this case, it uses isprint() to decide.
2268    
2269     4. Fixed a major bug that caused incorrect computation of the amount of memory
2270        required for a compiled pattern when options that changed within the
2271        pattern affected the logic of the preliminary scan that determines the
2272        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2273        that the computed length was too small. The symptoms of this bug were
2274        either the PCRE error "internal error: code overflow" from pcre_compile(),
2275        or a glibc crash with a message such as "pcretest: free(): invalid next
2276        size (fast)". Examples of patterns that provoked this bug (shown in
2277        pcretest format) are:
2278    
2279          /(?-x: )/x
2280          /(?x)(?-x: \s*#\s*)/
2281          /((?i)[\x{c0}])/8
2282          /(?i:[\x{c0}])/8
2283    
2284        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2285        is now done differently.
2286    
2287     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2288        wrapper classes; (b) implement a new function in the C++ scanner that is
2289        more efficient than the old way of doing things because it avoids levels of
2290        recursion in the regex matching; (c) add a paragraph to the documentation
2291        for the FullMatch() function.
2292    
2293     6. The escape sequence \n was being treated as whatever was defined as
2294        "newline". Not only was this contrary to the documentation, which states
2295        that \n is character 10 (hex 0A), but it also went horribly wrong when
2296        "newline" was defined as CRLF. This has been fixed.
2297    
2298     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2299        was being set to -1 for the "end of line" case (supposedly a value that no
2300        character can have). Though this value is never used (the check for end of
2301        line is "zero bytes in current character"), it caused compiler complaints.
2302        I've changed it to 0xffffffff.
2303    
2304     8. In pcre_version.c, the version string was being built by a sequence of
2305        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2306        string (as it is for production releases) called a macro with an empty
2307        argument. The C standard says the result of this is undefined. The gcc
2308        compiler treats it as an empty string (which was what was wanted) but it is
2309        reported that Visual C gives an error. The source has been hacked around to
2310        avoid this problem.
2311    
2312     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2313        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2314        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2315        of them did).
2316    
2317    10. Originally, pcretest opened its input and output without "b"; then I was
2318        told that "b" was needed in some environments, so it was added for release
2319        5.0 to both the input and output. (It makes no difference on Unix-like
2320        systems.) Later I was told that it is wrong for the input on Windows. I've
2321        now abstracted the modes into two macros, to make it easier to fiddle with
2322        them, and removed "b" from the input mode under Windows.
2323    
2324    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2325    
2326    12. Added -help and --help to pcretest as an official way of being reminded
2327        of the options.
2328    
2329    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2330        and pcrecpp.cc because they annoy compilers at high warning levels.
2331    
2332    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2333    
2334    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2335        scripts are not C programs :-) and which was not noticed because it works
2336        on Linux.
2337    
2338    16. pcretest is supposed to handle any length of pattern and data line (as one
2339        line or as a continued sequence of lines) by extending its input buffer if
2340        necessary. This feature was broken for very long pattern lines, leading to
2341        a string of junk being passed to pcre_compile() if the pattern was longer
2342        than about 50K.
2343    
2344    17. I have done a major re-factoring of the way pcre_compile() computes the
2345        amount of memory needed for a compiled pattern. Previously, there was code
2346        that made a preliminary scan of the pattern in order to do this. That was
2347        OK when PCRE was new, but as the facilities have expanded, it has become
2348        harder and harder to keep it in step with the real compile phase, and there
2349        have been a number of bugs (see for example, 4 above). I have now found a
2350        cunning way of running the real compile function in a "fake" mode that
2351        enables it to compute how much memory it would need, while actually only
2352        ever using a few hundred bytes of working memory and without too many
2353        tests of the mode. This should make future maintenance and development
2354        easier. A side effect of this work is that the limit of 200 on the nesting
2355        depth of parentheses has been removed (though this was never a serious
2356        limitation, I suspect). However, there is a downside: pcre_compile() now
2357        runs more slowly than before (30% or more, depending on the pattern). I
2358        hope this isn't a big issue. There is no effect on runtime performance.
2359    
2360    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2361        newline (only possible for the last line of a file) and it was a
2362        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2363    
2364    19. Added additional timing features to pcretest. (1) The -tm option now times
2365        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2366        separate command line item, by a number that specifies the number of
2367        repeats to use when timing. The default is 50000; this gives better
2368        precision, but takes uncomfortably long for very large patterns.
2369    
2370    20. Extended pcre_study() to be more clever in cases where a branch of a
2371        subpattern has no definite first character. For example, (a*|b*)[cd] would
2372        previously give no result from pcre_study(). Now it recognizes that the
2373        first character must be a, b, c, or d.
2374    
2375    21. There was an incorrect error "recursive call could loop indefinitely" if
2376        a subpattern (or the entire pattern) that was being tested for matching an
2377        empty string contained only one non-empty item after a nested subpattern.
2378        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2379        incorrectly, because the \d was being skipped in the check.
2380    
2381    22. The pcretest program now has a new pattern option /B and a command line
2382        option -b, which is equivalent to adding /B to every pattern. This causes
2383        it to show the compiled bytecode, without the additional information that
2384        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2385        is the same as /B/I).
2386    
2387    23. A new optimization is now able automatically to treat some sequences such
2388        as a*b as a*+b. More specifically, if something simple (such as a character
2389        or a simple class like \d) has an unlimited quantifier, and is followed by
2390        something that cannot possibly match the quantified thing, the quantifier
2391        is automatically "possessified".
2392    
2393    24. A recursive reference to a subpattern whose number was greater than 39
2394        went wrong under certain circumstances in UTF-8 mode. This bug could also
2395        have affected the operation of pcre_study().
2396    
2397    25. Realized that a little bit of performance could be had by replacing
2398        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2399    
2400    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2401    
2402    27. Possessive quantifiers such as a++ were previously implemented by turning
2403        them into atomic groups such as ($>a+). Now they have their own opcodes,
2404        which improves performance. This includes the automatically created ones
2405        from 23 above.
2406    
2407    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2408        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2409        the first matched character to be a colon. This applied both to named and
2410        numbered groups.
2411    
2412    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2413    
2414    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2415        building PCRE on Windows easier, so I have included it in the distribution.
2416    
2417    31. There is now a check in pcretest against a ridiculously large number being
2418        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2419        loop, the loop is abandoned.
2420    
2421    32. Forward references to subpatterns in conditions such as (?(2)...) where
2422        subpattern 2 is defined later cause pcre_compile() to search forwards in
2423        the pattern for the relevant set of parentheses. This search went wrong
2424        when there were unescaped parentheses in a character class, parentheses
2425        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2426    
2427    33. "Subroutine" calls and backreferences were previously restricted to
2428        referencing subpatterns earlier in the regex. This restriction has now
2429        been removed.
2430    
2431    34. Added a number of extra features that are going to be in Perl 5.10. On the
2432        whole, these are just syntactic alternatives for features that PCRE had
2433        previously implemented using the Python syntax or my own invention. The
2434        other formats are all retained for compatibility.
2435    
2436        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2437            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2438            also .NET compatible.
2439    
2440        (b) A recursion or subroutine call to a named group can now be defined as
2441            (?&name) as well as (?P>name).
2442    
2443        (c) A backreference to a named group can now be defined as \k<name> or
2444            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2445            5.10, are also .NET compatible.
2446    
2447        (d) A conditional reference to a named group can now use the syntax
2448            (?(<name>) or (?('name') as well as (?(name).
2449    
2450        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2451            groups (named and numbered) that are never evaluated inline, but can be
2452            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2453            is always false. There may be only one alternative in such a group.
2454    
2455        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2456            as the simple (?(R). The condition is true only if the most recent
2457            recursion is that of the given number or name. It does not search out
2458            through the entire recursion stack.
2459    
2460        (g) The escape \gN or \g{N} has been added, where N is a positive or
2461            negative number, specifying an absolute or relative reference.
2462    
2463    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2464        some "unreachable code" warnings.
2465    
2466    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2467        things, this adds five new scripts.
2468    
2469    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2470        There were also incompatibilities regarding the handling of \Q..\E inside
2471        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2472        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2473    
2474    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2475        matches an empty string, and forcibly breaks the loop. There were bugs in
2476        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2477        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2478        separate and independent bugs (that affected different cases) have been
2479        fixed.
2480    
2481    39. Refactored the code to abolish the use of different opcodes for small
2482        capturing bracket numbers. This is a tidy that I avoided doing when I
2483        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2484        The new approach is not only tidier, it makes it possible to reduce the
2485        memory needed to fix the previous bug (38).
2486    
2487    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2488        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2489        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2490        mode.
2491    
2492    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2493        report.
2494    
2495    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2496        copy construction and assignment in the C++ wrapper.
2497    
2498    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2499        couple of bugs that could have given wrong results in the "--newline=crlf"
2500        case.
2501    
2502    44. Added a number of casts and did some reorganization of signed/unsigned int
2503        variables following suggestions from Dair Grant. Also renamed the variable
2504        "this" as "item" because it is a C++ keyword.
2505    
2506    45. Arranged for dftables to add
2507    
2508          #include "pcre_internal.h"
2509    
2510        to pcre_chartables.c because without it, gcc 4.x may remove the array
2511        definition from the final binary if PCRE is built into a static library and
2512        dead code stripping is activated.
2513    
2514    46. For an unanchored pattern, if a match attempt fails at the start of a
2515        newline sequence, and the newline setting is CRLF or ANY, and the next two
2516        characters are CRLF, advance by two characters instead of one.
2517    
2518    
2519    Version 6.7 04-Jul-06
2520    ---------------------
2521    
2522     1. In order to handle tests when input lines are enormously long, pcretest has
2523        been re-factored so that it automatically extends its buffers when
2524        necessary. The code is crude, but this _is_ just a test program. The
2525        default size has been increased from 32K to 50K.
2526    
2527     2. The code in pcre_study() was using the value of the re argument before
2528        testing it for NULL. (Of course, in any sensible call of the function, it
2529        won't be NULL.)
2530    
2531     3. The memmove() emulation function in pcre_internal.h, which is used on
2532        systems that lack both memmove() and bcopy() - that is, hardly ever -
2533        was missing a "static" storage class specifier.
2534    
2535     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2536        containing an extended class (one that cannot be represented by a bitmap
2537        because it contains high-valued characters or Unicode property items, e.g.
2538        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2539        pattern, but PCRE should not loop if you do not (it no longer does).
2540        [Detail: two cases were found: (a) a repeated subpattern containing an
2541        extended class; (b) a recursive reference to a subpattern that followed a
2542        previous extended class. It wasn't skipping over the extended class
2543        correctly when UTF-8 mode was not set.]
2544    
2545     5. A negated single-character class was not being recognized as fixed-length
2546        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2547        compile error "lookbehind assertion is not fixed length".
2548    
2549     6. The RunPerlTest auxiliary script was showing an unexpected difference
2550        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2551        write a Perl script that can interpret lines of an input file either as
2552        byte characters or as UTF-8, which is what "perltest" was being required to
2553        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2554        can't do is switch easily at run time between having the "use utf8;" pragma
2555        or not. In the end, I fudged it by using the RunPerlTest script to insert
2556        "use utf8;" explicitly for the UTF-8 tests.
2557    
2558     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2559        the end of the subject string, contrary to the documentation and to what
2560        Perl does. This was true of both matching functions. Now it matches only at
2561        the start of the subject and immediately after *internal* newlines.
2562    
2563     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2564        a pointer to an int instead of a pointer to an unsigned long int. This
2565        caused problems on 64-bit systems.
2566    
2567     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2568        instance of the 'standard' template library not being so standard".
2569    
2570    10. There was no check on the number of named subpatterns nor the maximum
2571        length of a subpattern name. The product of these values is used to compute
2572        the size of the memory block for a compiled pattern. By supplying a very
2573        long subpattern name and a large number of named subpatterns, the size
2574        computation could be caused to overflow. This is now prevented by limiting
2575        the length of names to 32 characters, and the number of named subpatterns
2576        to 10,000.
2577    
2578    11. Subpatterns that are repeated with specific counts have to be replicated in
2579        the compiled pattern. The size of memory for this was computed from the
2580        length of the subpattern and the repeat count. The latter is limited to
2581        65535, but there was no limit on the former, meaning that integer overflow
2582        could in principle occur. The compiled length of a repeated subpattern is
2583        now limited to 30,000 bytes in order to prevent this.
2584    
2585    12. Added the optional facility to have named substrings with the same name.
2586    
2587    13. Added the ability to use a named substring as a condition, using the
2588        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2589        are numbers (not recommended). Forward references are permitted.
2590    
2591    14. Added forward references in named backreferences (if you see what I mean).
2592    
2593    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2594        pattern could run off the end of the subject. For example, the pattern
2595        "(?s)(.{1,5})"8 did this with the subject "ab".
2596    
2597    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2598        PCRE_CASELESS was set when matching characters that were quantified with ?
2599        or *.
2600    
2601    17. A character class other than a single negated character that had a minimum
2602        but no maximum quantifier - for example [ab]{6,} - was not handled
2603        correctly by pce_dfa_exec(). It would match only one character.
2604    
2605    18. A valid (though odd) pattern that looked like a POSIX character
2606        class but used an invalid character after [ (for example [[,abc,]]) caused
2607        pcre_compile() to give the error "Failed: internal error: code overflow" or
2608        in some cases to crash with a glibc free() error. This could even happen if
2609        the pattern terminated after [[ but there just happened to be a sequence of
2610        letters, a binary zero, and a closing ] in the memory that followed.
2611    
2612    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2613        over the years. Originally (before any Unicode support), just the bottom 8
2614        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2615        output from "man perlunicode" includes this:
2616    
2617          The regular expression compiler produces polymorphic opcodes.  That
2618          is, the pattern adapts to the data and automatically switches to
2619          the Unicode character scheme when presented with Unicode data--or
2620          instead uses a traditional byte scheme when presented with byte
2621          data.
2622    
2623        Sadly, a wide octal escape does not cause a switch, and in a string with
2624        no other multibyte characters, these octal escapes are treated as before.
2625        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2626        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2627        Unicode string.
2628    
2629        I have not perpetrated such confusion in PCRE. Up till now, it took just
2630        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2631        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2632        translate to the appropriate multibyte character.
2633    
2634    29. Applied some refactoring to reduce the number of warnings from Microsoft
2635        and Borland compilers. This has included removing the fudge introduced
2636        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2637        a warning about an unused variable.
2638    
2639    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2640        characters since release 4.0, because Perl (from release 5.004) does not.
2641        [Or at least, is documented not to: some releases seem to be in conflict
2642        with the documentation.] However, when a pattern was studied with
2643        pcre_study() and all its branches started with \s, PCRE still included VT
2644        as a possible starting character. Of course, this did no harm; it just
2645        caused an unnecessary match attempt.
2646    
2647    22. Removed a now-redundant internal flag bit that recorded the fact that case
2648        dependency changed within the pattern. This was once needed for "required
2649        byte" processing, but is no longer used. This recovers a now-scarce options
2650        bit. Also moved the least significant internal flag bit to the most-
2651        significant bit of the word, which was not previously used (hangover from
2652        the days when it was an int rather than a uint) to free up another bit for
2653        the future.
2654    
2655    23. Added support for CRLF line endings as well as CR and LF. As well as the
2656        default being selectable at build time, it can now be changed at runtime
2657        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2658        specify that it is scanning data with non-default line endings.
2659    
2660    24. Changed the definition of CXXLINK to make it agree with the definition of
2661        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2662    
2663    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2664        recursions. This makes a big different to stack usage for some patterns.
2665    
2666    26. If a subpattern containing a named recursion or subroutine reference such
2667        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2668        the space required for the compiled pattern went wrong and gave too small a
2669        value. Depending on the environment, this could lead to "Failed: internal
2670        error: code overflow at offset 49" or "glibc detected double free or
2671        corruption" errors.
2672    
2673    27. Applied patches from Google (a) to support the new newline modes and (b) to
2674        advance over multibyte UTF-8 characters in GlobalReplace.
2675    
2676    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2677        difference for some implementation of PCRE in some Windows version.
2678    
2679    29. Added some extra testing facilities to pcretest:
2680    
2681        \q<number>   in a data line sets the "match limit" value
2682        \Q<number>   in a data line sets the "match recursion limt" value
2683        -S <number>  sets the stack size, where <number> is in megabytes
2684    
2685        The -S option isn't available for Windows.
2686    
2687    
2688    Version 6.6 06-Feb-06
2689    ---------------------
2690    
2691     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2692        in pcreposix.h. I have copied the definition from pcre.h.
2693    
2694     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2695        because pcre.h is no longer a built file.
2696    
2697     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2698        not normally included in the compiled code.
2699    
2700    
2701    Version 6.5 01-Feb-06
2702    ---------------------
2703    
2704     1. When using the partial match feature with pcre_dfa_exec(), it was not
2705        anchoring the second and subsequent partial matches at the new starting
2706        point. This could lead to incorrect results. For example, with the pattern
2707        /1234/, partially matching against "123" and then "a4" gave a match.
2708    
2709     2. Changes to pcregrep:
2710    
2711        (a) All non-match returns from pcre_exec() were being treated as failures
2712            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2713            error message is output. Some extra information is given for the
2714            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2715            probably the only errors that are likely to be caused by users (by
2716            specifying a regex that has nested indefinite repeats, for instance).
2717            If there are more than 20 of these errors, pcregrep is abandoned.
2718    
2719        (b) A binary zero was treated as data while matching, but terminated the
2720            output line if it was written out. This has been fixed: binary zeroes
2721            are now no different to any other data bytes.
2722    
2723        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2724            used to set a locale for matching. The --locale=xxxx long option has
2725            been added (no short equivalent) to specify a locale explicitly on the
2726            pcregrep command, overriding the environment variables.
2727    
2728        (d) When -B was used with -n, some line numbers in the output were one less
2729            than they should have been.
2730    
2731        (e) Added the -o (--only-matching) option.
2732    
2733        (f) If -A or -C was used with -c (count only), some lines of context were
2734            accidentally printed for the final match.
2735    
2736        (g) Added the -H (--with-filename) option.
2737    
2738        (h) The combination of options -rh failed to suppress file names for files
2739            that were found from directory arguments.
2740    
2741        (i) Added the -D (--devices) and -d (--directories) options.
2742    
2743        (j) Added the -F (--fixed-strings) option.
2744    
2745        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2746    
2747        (l) Added the --colo(u)r option.
2748    
2749        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2750            is not present by default.
2751    
2752     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2753        items such as (?R) or (?1), when the recursion could match a number of
2754        alternatives. If it matched one of the alternatives, but subsequently,
2755        outside the recursion, there was a failure, the code tried to back up into
2756        the recursion. However, because of the way PCRE is implemented, this is not
2757        possible, and the result was an incorrect result from the match.
2758    
2759        In order to prevent this happening, the specification of recursion has
2760        been changed so that all such subpatterns are automatically treated as
2761        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2762    
2763     4. I had overlooked the fact that, in some locales, there are characters for
2764        which isalpha() is true but neither isupper() nor islower() are true. In
2765        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2766        and ordfeminine) are like this. This affected the treatment of \w and \W
2767        when they appeared in character classes, but not when they appeared outside
2768        a character class. The bit map for "word" characters is now created
2769        separately from the results of isalnum() instead of just taking it from the
2770        upper, lower, and digit maps. (Plus the underscore character, of course.)
2771    
2772     5. The above bug also affected the handling of POSIX character classes such as
2773        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2774        permanent tables. Instead, the bit maps for such a class were previously
2775        created as the appropriate unions of the upper, lower, and digit bitmaps.
2776        Now they are created by subtraction from the [[:word:]] class, which has
2777        its own bitmap.
2778    
2779     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2780        It is created by subtracting the vertical space characters (\x09, \x0a,
2781        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2782        subtraction was done in the overall bitmap for a character class, meaning
2783        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2784        be recognized. This bug has been fixed.
2785    
2786     7. Patches from the folks at Google:
2787    
2788          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2789          real life, but is still worth protecting against".
2790    
2791          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2792          regular expressions".
2793    
2794          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2795          have it.
2796    
2797          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2798          "configure" and the latter not, in order to fix a problem somebody had
2799          with compiling the Arg class on HP-UX.
2800    
2801          (e) Improve the error-handling of the C++ wrapper a little bit.
2802    
2803          (f) New tests for checking recursion limiting.
2804    
2805     8. The pcre_memmove() function, which is used only if the environment does not
2806        have a standard memmove() function (and is therefore rarely compiled),
2807        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2808        returning a result (though PCRE never actually uses the result).
2809    
2810     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2811        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2812        returned instead of calling malloc() with an overflowing number that would
2813        most likely cause subsequent chaos.
2814    
2815    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2816    
2817    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2818        with this option is matched, the nmatch and pmatch options of regexec() are
2819        ignored.
2820    
2821    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2822        provided in case anyone wants to the the POSIX interface with UTF-8
2823        strings.
2824    
2825    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2826        C++ linking (needed for some HP-UX environments).
2827    
2828    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2829        (unused parameter) and in the pcre_printint() function (omitted "default"
2830        switch label when the default is to do nothing).
2831    
2832    15. Added some code to make it possible, when PCRE is compiled as a C++
2833        library, to replace subject pointers for pcre_exec() with a smart pointer
2834        class, thus making it possible to process discontinuous strings.
2835    
2836    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2837        much the same function. They were added by different people who were trying
2838        to make PCRE easy to compile on non-Unix systems. It has been suggested
2839        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2840        for compiling on Windows systems. I have therefore replaced it with
2841        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2842        defaults to "extern" for C or "extern C" for C++, which works fine on
2843        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2844        SCOPE with something explicit in config.h. In addition:
2845    
2846        (a) pcreposix.h still had just "extern" instead of either of these macros;
2847            I have replaced it with PCRE_DATA_SCOPE.
2848    
2849        (b) Functions such as _pcre_xclass(), which are internal to the library,
2850            but external in the C sense, all had PCRE_EXPORT in their definitions.
2851            This is apparently wrong for the Windows case, so I have removed it.
2852            (It makes no difference on Unix-like systems.)
2853    
2854    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2855        of recursive calls to match(). This is different to MATCH_LIMIT because
2856        that limits the total number of calls to match(), not all of which increase
2857        the depth of recursion. Limiting the recursion depth limits the amount of
2858        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2859        when PCRE is compiled, and changed at run time. A patch from Google adds
2860        this functionality to the C++ interface.
2861    
2862    18. Changes to the handling of Unicode character properties:
2863    
2864        (a) Updated the table to Unicode 4.1.0.
2865    
2866        (b) Recognize characters that are not in the table as "Cn" (undefined).
2867    
2868        (c) I revised the way the table is implemented to a much improved format
2869            which includes recognition of ranges. It now supports the ranges that
2870            are defined in UnicodeData.txt, and it also amalgamates other
2871            characters into ranges. This has reduced the number of entries in the
2872            table from around 16,000 to around 3,000, thus reducing its size
2873            considerably. I realized I did not need to use a tree structure after
2874            all - a binary chop search is just as efficient. Having reduced the
2875            number of entries, I extended their size from 6 bytes to 8 bytes to
2876            allow for more data.
2877    
2878        (d) Added support for Unicode script names via properties such as \p{Han}.
2879    
2880    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2881        matching that character.
2882    
2883    20. When matching a repeated Unicode property with a minimum greater than zero,
2884        (for example \pL{2,}), PCRE could look past the end of the subject if it
2885        reached it while seeking the minimum number of characters. This could
2886        happen only if some of the characters were more than one byte long, because
2887        there is a check for at least the minimum number of bytes.
2888    
2889    21. Refactored the implementation of \p and \P so as to be more general, to
2890        allow for more different types of property in future. This has changed the
2891        compiled form incompatibly. Anybody with saved compiled patterns that use
2892        \p or \P will have to recompile them.
2893    
2894    22. Added "Any" and "L&" to the supported property types.
2895    
2896    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2897        but give a compile time error if the value is greater than 0xff.
2898    
2899    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2900        accidentally not being installed or uninstalled.
2901    
2902    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2903        made were to insert the current release number. This seemed silly, because
2904        it made things harder for people building PCRE on systems that don't run
2905        "configure". I have turned pcre.h into a distributed file, no longer built
2906        by "configure", with the version identification directly included. There is
2907        no longer a pcre.h.in file.
2908    
2909        However, this change necessitated a change to the pcre-config script as
2910        well. It is built from pcre-config.in, and one of the substitutions was the
2911        release number. I have updated configure.ac so that ./configure now finds
2912        the release number by grepping pcre.h.
2913    
2914    26. Added the ability to run the tests under valgrind.
2915    
2916    
2917    Version 6.4 05-Sep-05
2918    ---------------------
2919    
2920     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2921        "--" to be printed when multiple files were scanned, even when none of the
2922        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2923        consider it to be a bug, and have restored the previous behaviour.
2924    
2925     2. A couple of code tidies to get rid of compiler warnings.
2926    
2927     3. The pcretest program used to cheat by referring to symbols in the library
2928        whose names begin with _pcre_. These are internal symbols that are not
2929        really supposed to be visible externally, and in some environments it is
2930        possible to suppress them. The cheating is now confined to including
2931        certain files from the library's source, which is a bit cleaner.
2932    
2933     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2934        file's purpose clearer.
2935    
2936     5. Reorganized pcre_ucp_findchar().
2937    
2938    
2939    Version 6.3 15-Aug-05
2940    ---------------------
2941    
2942     1. The file libpcre.pc.in did not have general read permission in the tarball.
2943    
2944     2. There were some problems when building without C++ support:
2945    
2946        (a) If C++ support was not built, "make install" and "make test" still
2947            tried to test it.
2948    
2949        (b) There were problems when the value of CXX was explicitly set. Some
2950            changes have been made to try to fix these, and ...
2951    
2952        (c) --disable-cpp can now be used to explicitly disable C++ support.
2953    
2954        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2955            backslash in a target when C++ was disabled. This confuses some
2956            versions of "make", apparently. Using an intermediate variable solves
2957            this. (Same for CPP_LOBJ.)
2958    
2959     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2960        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2961        necessary on certain architectures.
2962    
2963     4. Added a setting of -export-symbols-regex to the link command to remove
2964        those symbols that are exported in the C sense, but actually are local
2965        within the library, and not documented. Their names all begin with
2966        "_pcre_". This is not a perfect job, because (a) we have to except some
2967        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2968        available (and never for static libraries). I have made a note to try to
2969        find a way round (a) in the future.
2970    
2971    
2972    Version 6.2 01-Aug-05
2973    ---------------------
2974    
2975     1. There was no test for integer overflow of quantifier values. A construction
2976        such as {1111111111111111} would give undefined results. What is worse, if
2977        a minimum quantifier for a parenthesized subpattern overflowed and became
2978        negative, the calculation of the memory size went wrong. This could have
2979        led to memory overwriting.
2980    
2981     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2982    
2983     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2984        operating environments where this matters.
2985    
2986     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2987        PCRE options from within the C++ wrapper.
2988    
2989     5. Named capturing subpatterns were not being correctly counted when a pattern
2990        was compiled. This caused two problems: (a) If there were more than 100
2991        such subpatterns, the calculation of the memory needed for the whole
2992        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2993        back references of the form \12, where the number was greater than 9, were
2994        not recognized as back references, even though there were sufficient
2995        previous subpatterns.
2996    
2997     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2998        versions of gcc, e.g. 2.95.4.
2999    
3000    
3001    Version 6.1 21-Jun-05
3002    ---------------------
3003    
3004     1. There was one reference to the variable "posix" in pcretest.c that was not
3005        surrounded by "#if !defined NOPOSIX".
3006    
3007     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3008        the cross-check on the old pcre_info() function, for the benefit of the
3009        cut-down version of PCRE that is currently imported into Exim.
3010    
3011     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3012        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3013        patterns that start (?i) but I don't think that matters. The use of (?i) is
3014        just an example; this all applies to the other options as well.
3015    
3016     4. Since libtool seems to echo the compile commands it is issuing, the output
3017        from "make" can be reduced a bit by putting "@" in front of each libtool
3018        compile command.
3019    
3020     5. Patch from the folks at Google for configure.in to be a bit more thorough
3021        in checking for a suitable C++ installation before trying to compile the
3022        C++ stuff. This should fix a reported problem when a compiler was present,
3023        but no suitable headers.
3024    
3025     6. The man pages all had just "PCRE" as their title. I have changed them to
3026        be the relevant file name. I have also arranged that these names are
3027        retained in the file doc/pcre.txt, which is a concatenation in text format
3028        of all the man pages except the little individual ones for each function.
3029    
3030     7. The NON-UNIX-USE file had not been updated for the different set of source
3031        files that come with release 6. I also added a few comments about the C++
3032        wrapper.
3033    
3034    
3035    Version 6.0 07-Jun-05
3036    ---------------------
3037    
3038     1. Some minor internal re-organization to help with my DFA experiments.
3039    
3040     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3041        didn't matter for the library itself when fully configured, but did matter
3042        when compiling without UCP support, or within Exim, where the ucp files are
3043        not imported.
3044    
3045     3. Refactoring of the library code to split up the various functions into
3046        different source modules. The addition of the new DFA matching code (see
3047        below) to a single monolithic source would have made it really too
3048        unwieldy, quite apart from causing all the code to be include in a
3049        statically linked application, when only some functions are used. This is
3050        relevant even without the DFA addition now that patterns can be compiled in
3051        one application and matched in another.
3052    
3053        The downside of splitting up is that there have to be some external
3054        functions and data tables that are used internally in different modules of
3055        the library but which are not part of the API. These have all had their
3056        names changed to start with "_pcre_" so that they are unlikely to clash
3057        with other external names.
3058    
3059     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3060        a different (DFA) algorithm. Although it is slower than the original
3061        function, it does have some advantages for certain types of matching
3062        problem.
3063    
3064     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3065        including restarting after a partial match.
3066    
3067     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3068        defined when compiling for Windows was sent to me. I have put it into the
3069        code, though I have no means of testing or verifying it.
3070    
3071     7. Added the pcre_refcount() auxiliary function.
3072    
3073     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3074        match before or at the first newline in the subject string. In pcretest,
3075        the /f option on a pattern can be used to set this.
3076    
3077     9. A repeated \w when used in UTF-8 mode with characters greater than 256
3078        would behave wrongly. This has been present in PCRE since release 4.0.
3079    
3080    10. A number of changes to the pcregrep command:
3081    
3082        (a) Refactored how -x works; insert ^(...)$ instead of setting
3083            PCRE_ANCHORED and checking the length, in preparation for adding
3084            something similar for -w.
3085    
3086        (b) Added the -w (match as a word) option.
3087    
3088        (c) Refactored the way lines are read and buffered so as to have more
3089            than one at a time available.
3090    
3091        (d) Implemented a pcregrep test script.
3092    
3093        (e) Added the -M (multiline match) option. This allows patterns to match
3094            over several lines of the subject. The buffering ensures that at least
3095            8K, or the rest of the document (whichever is the shorter) is available
3096            for matching (and similarly the previous 8K for lookbehind assertions).
3097    
3098        (f) Changed the --help output so that it now says
3099    
3100              -w, --word-regex(p)
3101    
3102            instead of two lines, one with "regex" and the other with "regexp"
3103            because that confused at least one person since the short forms are the
3104            same. (This required a bit of code, as the output is generated
3105            automatically from a table. It wasn't just a text change.)
3106    
3107        (g) -- can be used to terminate pcregrep options if the next thing isn't an
3108            option but starts with a hyphen. Could be a pattern or a path name
3109            starting with a hyphen, for instance.
3110    
3111        (h) "-" can be given as a file name to represent stdin.
3112    
3113        (i) When file names are being printed, "(standard input)" is used for
3114            the standard input, for compatibility with GNU grep. Previously
3115            "<stdin>" was used.
3116    
3117        (j) The option --label=xxx can be used to supply a name to be used for
3118            stdin when file names are being printed. There is no short form.
3119    
3120        (k) Re-factored the options decoding logic because we are going to add
3121            two more options that take data. Such options can now be given in four
3122            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3123    
3124        (l) Added the -A, -B, and -C options for requesting that lines of context
3125            around matches be printed.
3126    
3127        (m) Added the -L option to print the names of files that do not contain
3128            any matching lines, that is, the complement of -l.
3129    
3130        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3131            continue to scan other files.
3132    
3133        (o) The -s option was incorrectly implemented. For compatibility with other
3134            greps, it now suppresses the error message for a non-existent or non-
3135            accessible file (but not the return code). There is a new option called
3136            -q that suppresses the output of matching lines, which was what -s was
3137            previously doing.
3138    
3139        (p) Added --include and --exclude options to specify files for inclusion
3140            and exclusion when recursing.
3141    
3142    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3143        Hopefully, it now does.
3144    
3145    12. Missing cast in pcre_study().
3146    
3147    13. Added an "uninstall" target to the makefile.
3148    
3149    14. Replaced "extern" in the function prototypes in Makefile.in with
3150        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3151        world, but is set differently for Windows.
3152    
3153    15. Added a second compiling function called pcre_compile2(). The only
3154        difference is that it has an extra argument, which is a pointer to an
3155        integer error code. When there is a compile-time failure, this is set
3156        non-zero, in addition to the error test pointer being set to point to an
3157        error message. The new argument may be NULL if no error number is required
3158        (but then you may as well call pcre_compile(), which is now just a
3159        wrapper). This facility is provided because some applications need a
3160        numeric error indication, but it has also enabled me to tidy up the way
3161        compile-time errors are handled in the POSIX wrapper.
3162    
3163    16. Added VPATH=.libs to the makefile; this should help when building with one
3164        prefix path and installing with another. (Or so I'm told by someone who
3165        knows more about this stuff than I do.)
3166    
3167    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3168        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3169        match everything, including newlines. This is not POSIX-compatible, but
3170        somebody wanted the feature. From pcretest it can be activated by using
3171        both the P and the s flags.
3172    
3173    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3174    
3175    19. libpcre.pc was being incorrectly installed as executable.
3176    
3177    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3178        it now also looks for '\r' so that it will work unmodified on Windows.
3179    
3180    21. Added Google's contributed C++ wrapper to the distribution.
3181    
3182    22. Added some untidy missing memory free() calls in pcretest, to keep
3183        Electric Fence happy when testing.
3184    
3185    
3186    
3187    Version 5.0 13-Sep-04
3188    ---------------------
3189    
3190     1. Internal change: literal characters are no longer packed up into items
3191        containing multiple characters in a single byte-string. Each character
3192        is now matched using a separate opcode. However, there may be more than one
3193        byte in the character in UTF-8 mode.
3194    
3195     2. The pcre_callout_block structure has two new fields: pattern_position and
3196        next_item_length. These contain the offset in the pattern to the next match
3197        item, and its length, respectively.
3198    
3199     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3200        insertion of callouts before each pattern item. Added the /C option to
3201        pcretest to make use of this.
3202    
3203     4. On the advice of a Windows user, the lines
3204    
3205          #if defined(_WIN32) || defined(WIN32)
3206          _setmode( _fileno( stdout ), 0x8000 );
3207          #endif  /* defined(_WIN32) || defined(WIN32) */
3208    
3209        have been added to the source of pcretest. This apparently does useful
3210        magic in relation to line terminators.
3211    
3212     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3213        for the benefit of those environments where the "b" makes a difference.
3214    
3215     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3216        to know about it. I have put a hack into configure.in that adds in code
3217        to set GCC=yes if CC=icc. This seems to end up at a point in the
3218        generated configure script that is early enough to affect the setting of
3219        compiler options, which is what is needed, but I have no means of testing
3220        whether it really works. (The user who reported this had patched the
3221        generated configure script, which of course I cannot do.)
3222    
3223        LATER: After change 22 below (new libtool files), the configure script
3224        seems to know about icc (and also ecc). Therefore, I have commented out
3225        this hack in configure.in.
3226    
3227     7. Added support for pkg-config (2 patches were sent in).
3228    
3229     8. Negated POSIX character classes that used a combination of internal tables
3230        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3231        [[:^ascii]]. Typically, they would match almost any characters. The other
3232        POSIX classes were not broken in this way.
3233    
3234     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3235        to find the match, as PCRE was deluded into thinking that the match had to
3236        start at the start point or following a newline. The same bug applied to
3237        patterns with negative forward assertions or any backward assertions
3238        preceding ".*" at the start, unless the pattern required a fixed first
3239        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3240    
3241    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3242        starting at the last subject character, bytes beyond the end of the subject
3243        string were read.
3244    
3245    11. Renamed the variable "class" as "classbits" to make life easier for C++
3246        users. (Previously there was a macro definition, but it apparently wasn't
3247        enough.)
3248    
3249    12. Added the new field "tables" to the extra data so that tables can be passed
3250        in at exec time, or the internal tables can be re-selected. This allows
3251        a compiled regex to be saved and re-used at a later time by a different
3252        program that might have everything at different addresses.
3253    
3254    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3255        -R library as well as a -L library.
3256    
3257    14. The debugging options of pcretest (-d on the command line or D on a
3258        pattern) showed incorrect output for anything following an extended class
3259        that contained multibyte characters and which was followed by a quantifier.
3260    
3261    15. Added optional support for general category Unicode character properties
3262        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3263        support. It adds about 90K to the size of the library. The meanings of the
3264        inbuilt class escapes such as \d and \s have NOT been changed.
3265    
3266    16. Updated pcredemo.c to include calls to free() to release the memory for the
3267        compiled pattern.
3268    
3269    17. The generated file chartables.c was being created in the source directory
3270        instead of in the building directory. This caused the build to fail if the
3271        source directory was different from the building directory, and was
3272        read-only.
3273    
3274    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3275        file. No doubt somebody will tell me if they don't make sense... Also added
3276        Dan Mooney's comments about building on OpenVMS.
3277    
3278    19. Added support for partial matching via the PCRE_PARTIAL option for
3279        pcre_exec() and the \P data escape in pcretest.
3280    
3281    20. Extended pcretest with 3 new pattern features:
3282    
3283        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3284              write the compiled pattern to the file whose name is "rest-of-line".
3285              This is a straight binary dump of the data, with the saved pointer to
3286              the character tables forced to be NULL. The study data, if any, is
3287              written too. After writing, pcretest reads a new pattern.
3288    
3289        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3290              compiled pattern from the given file. There must not be any
3291              occurrences of "<" in the file name (pretty unlikely); if there are,
3292              pcretest will instead treat the initial "<" as a pattern delimiter.
3293              After reading in the pattern, pcretest goes on to read data lines as
3294              usual.
3295    
3296        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3297              and 16-bit fields in a compiled pattern, to simulate a pattern that
3298              was compiled on a host of opposite endianness.
3299    
3300    21. The pcre-exec() function can now cope with patterns that were compiled on
3301        hosts of opposite endianness, with this restriction:
3302    
3303          As for any compiled expression that is saved and used later, the tables
3304          pointer field cannot be preserved; the extra_data field in the arguments
3305          to pcre_exec() should be used to pass in a tables address if a value
3306          other than the default internal tables were used at compile time.
3307    
3308    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3309        now diagnosed as an error. Previously, most of the time, a negative number
3310        would have been treated as zero, but if in addition "ovector" was passed as
3311        NULL, a crash could occur.
3312    
3313    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3314        new versions from the libtool 1.5 distribution (the last one is a copy of
3315        a file called libtool.m4). This seems to have fixed the need to patch
3316        "configure" to support Darwin 1.3 (which I used to do). However, I still
3317        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3318        workstation).
3319    
3320    24. Changed the PCRE licence to be the more standard "BSD" licence.
3321    
3322    
3323    Version 4.5 01-Dec-03
3324    ---------------------
3325    
3326     1. There has been some re-arrangement of the code for the match() function so
3327        that it can be compiled in a version that does not call itself recursively.
3328        Instead, it keeps those local variables that need separate instances for
3329        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3330        needs to "recurse". Keeping track of where control must go is done by means
3331        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3332        hide most of the details from the main code, and operates only if
3333        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3334        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3335        operating.
3336    
3337        To make it easier for callers to provide specially tailored get/free
3338        functions for this usage, two new functions, pcre_stack_malloc, and
3339        pcre_stack_free, are used. They are always called in strict stacking order,
3340        and the size of block requested is always the same.
3341    
3342        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3343        PCRE has been compiled to use the stack or the heap for recursion. The
3344        -C option of pcretest uses this to show which version is compiled.
3345    
3346        A new data escape \S, is added to pcretest; it causes the amounts of store
3347        obtained and freed by both kinds of malloc/free at match time to be added
3348        to the output.
3349    
3350     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3351        what's available on my current Linux desktop machine.
3352    
3353     3. When matching a UTF-8 string, the test for a valid string at the start has
3354        been extended. If start_offset is not zero, PCRE now checks that it points
3355        to a byte that is the start of a UTF-8 character. If not, it returns
3356        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3357        this is necessary because there may be backward assertions in the pattern.
3358        When matching the same subject several times, it may save resources to use
3359        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3360    
3361     4. The code for checking the validity of UTF-8 strings has been tightened so
3362        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3363        containing "overlong sequences".
3364    
3365     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3366        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3367        should have been "&", but it just so happened that all the cases this let
3368        through by mistake were picked up later in the function.
3369    
3370     6. I had used a variable called "isblank" - this is a C99 function, causing
3371        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3372    
3373     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3374        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3375        so the version is shown; (c) stop "make test" echoing "./RunTest".
3376    
3377     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3378    
3379     9. The prototype for memmove() for systems that don't have it was using
3380        size_t, but the inclusion of the header that defines size_t was later. I've
3381        moved the #includes for the C headers earlier to avoid this.
3382    
3383    10. Added some adjustments to the code to make it easier to compiler on certain
3384        special systems:
3385    
3386          (a) Some "const" qualifiers were missing.
3387          (b) Added the macro EXPORT before all exported functions; by default this
3388              is defined to be empty.
3389          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3390              that it reads its output file name as an argument instead of writing
3391              to the standard output and assuming this can be redirected.
3392    
3393    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3394        class containing characters with values greater than 255, PCRE compilation
3395        went into a loop.
3396    
3397    12. A recursive reference to a subpattern that was within another subpattern
3398        that had a minimum quantifier of zero caused PCRE to crash. For example,
3399        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3400        recursion. If the recursively-called subpattern itself had a zero repeat,
3401        that was OK.
3402    
3403    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3404        buffer into which it was copied (for escape processing) was still set at
3405        1024, so long lines caused crashes.
3406    
3407    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3408        "internal error: code overflow...". This applied to any character class
3409        that was followed by a possessive quantifier.
3410    
3411    15. Modified the Makefile to add libpcre.la as a prerequisite for
3412        libpcreposix.la because I was told this is needed for a parallel build to
3413        work.
3414    
3415    16. If a pattern that contained .* following optional items at the start was
3416        studied, the wrong optimizing data was generated, leading to matching
3417        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3418        matching string must start with a or b or c. The correct conclusion for
3419        this pattern is that a match can start with any character.
3420    
3421    
3422    Version 4.4 13-Aug-03
3423    ---------------------
3424    
3425     1. In UTF-8 mode, a character class containing characters with values between
3426        127 and 255 was not handled correctly if the compiled pattern was studied.
3427        In fixing this, I have also improved the studying algorithm for such
3428        classes (slightly).
3429    
3430     2. Three internal functions had redundant arguments passed to them. Removal
3431        might give a very teeny performance improvement.
3432    
3433     3. Documentation bug: the value of the capture_top field in a callout is *one
3434        more than* the number of the hightest numbered captured substring.
3435    
3436     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3437        in incorrectly linking with a previously installed version. They now link
3438        explicitly with libpcre.la.
3439    
3440     5. configure.in no longer needs to recognize Cygwin specially.
3441    
3442     6. A problem in pcre.in for Windows platforms is fixed.
3443    
3444     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3445        pcretest, it used to include the size of the study block as part of its
3446        output. Unfortunately, the structure contains a field that has a different
3447        size on different hardware architectures. This meant that the tests that
3448        showed this size failed. As the block is currently always of a fixed size,
3449        this information isn't actually particularly useful in pcretest output, so
3450        I have just removed it.
3451    
3452     8. Three pre-processor statements accidentally did not start in column 1.
3453        Sadly, there are *still* compilers around that complain, even though
3454        standard C has not required this for well over a decade. Sigh.
3455    
3456     9. In pcretest, the code for checking callouts passed small integers in the
3457        callout_data field, which is a void * field. However, some picky compilers
3458        complained about the casts involved for this on 64-bit systems. Now
3459        pcretest passes the address of the small integer instead, which should get
3460        rid of the warnings.
3461    
3462    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3463        both compile and run time, and gives an error if an invalid UTF-8 sequence
3464        is found. There is a option for disabling this check in cases where the
3465        string is known to be correct and/or the maximum performance is wanted.
3466    
3467    11. In response to a bug report, I changed one line in Makefile.in from
3468    
3469            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3470        to
3471            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3472    
3473        to look similar to other lines, but I have no way of telling whether this
3474        is the right thing to do, as I do not use Windows. No doubt I'll get told
3475        if it's wrong...
3476    
3477    
3478    Version 4.3 21-May-03
3479    ---------------------
3480    
3481    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3482       Makefile.
3483    
3484    2. Some refactoring to improve the quality of the code:
3485    
3486       (i)   The utf8_table... variables are now declared "const".
3487    
3488       (ii)  The code for \cx, which used the "case flipping" table to upper case
3489             lower case letters, now just substracts 32. This is ASCII-specific,
3490             but the whole concept of \cx is ASCII-specific, so it seems
3491             reasonable.
3492    
3493       (iii) PCRE was using its character types table to recognize decimal and
3494             hexadecimal digits in the pattern. This is silly, because it handles
3495             only 0-9, a-f, and A-F, but the character types table is locale-
3496             specific, which means strange things might happen. A private
3497             table is now used for this - though it costs 256 bytes, a table is
3498             much faster than multiple explicit tests. Of course, the standard
3499             character types table is still used for matching digits in subject
3500             strings against \d.
3501    
3502       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3503             ending in _t are). So I've renamed it as ESC_tee.
3504    
3505    3. The first argument for regexec() in the POSIX wrapper should have been
3506       defined as "const".
3507    
3508    4. Changed pcretest to use malloc() for its buffers so that they can be
3509       Electric Fenced for debugging.
3510    
3511    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3512       to read one or more bytes before the start of the subject string. Often this
3513       had no effect on PCRE's behaviour, but in some circumstances it could
3514       provoke a segmentation fault.
3515    
3516    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3517       to try to read one or more bytes before the start of the subject string.
3518    
3519    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3520       UTF-8 support could misbehave in various ways if the subject string
3521       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3522       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3523       back over UTF-8 characters.)
3524    
3525    
3526    Version 4.2 14-Apr-03
3527    ---------------------
3528    
3529    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3530    
3531    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3532         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3533         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3534         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3535         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3536           and BUILD_EXEEXT
3537         Note: automatic setting of the BUILD variables is not yet working
3538         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3539           compile-time but not at link-time
3540         [LINK]: use for linking executables only
3541         make different versions for Windows and non-Windows
3542         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3543           libraries
3544         [LINK_FOR_BUILD]: new variable
3545         [OBJEXT]: use throughout
3546         [EXEEXT]: use throughout
3547         <winshared>: new target
3548         <wininstall>: new target
3549         <dftables.o>: use native compiler
3550         <dftables>: use native linker
3551         <install>: handle Windows platform correctly
3552         <clean>: ditto
3553         <check>: ditto
3554         copy DLL to top builddir before testing
3555    
3556       As part of these changes, -no-undefined was removed again. This was reported
3557       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3558       in any case.
3559    
3560    3. Some tidies to get rid of compiler warnings:
3561    
3562       . In the match_data structure, match_limit was an unsigned long int, whereas
3563         match_call_count was an int. I've made them both unsigned long ints.
3564    
3565       . In pcretest the fact that a const uschar * doesn't automatically cast to
3566         a void * provoked a warning.
3567    
3568       . Turning on some more compiler warnings threw up some "shadow" variables
3569         and a few more missing casts.
3570    
3571    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3572       option, a class that contained a single character with a value between 128
3573       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3574    
3575    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3576       option, a class that contained several characters, but with at least one
3577       whose value was between 128 and 255 caused PCRE to crash.
3578    
3579    
3580    Version 4.1 12-Mar-03
3581    ---------------------
3582    
3583    1. Compiling with gcc -pedantic found a couple of places where casts were
3584    needed, and a string in dftables.c that was longer than standard compilers are
3585    required to support.
3586    
3587    2. Compiling with Sun's compiler found a few more places where the code could
3588    be tidied up in order to avoid warnings.
3589