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 942 by ph10, Tue Feb 28 14:50:31 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.31
5    -----------------------------
7    1.  Fixing a wrong JIT test case and some compiler warnings.
9    2.  Removed a bashism from the RunTest script.
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)".
15    4.  Partial matching support is added to the JIT compiler.
17    5.  Fixed several bugs concerned with partial matching of items that consist
18        of more than one character:
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.
24        (b) \R did not give a hard partial match if \r was found at the end of the
25            subject.
27        (c) \X did not give a hard partial match after matching one or more
28            characters at the end of the subject.
30        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
31            a partial match for the string "\r".
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.
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.
40    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
41        or /S+[+] with a digit between 1 and 7.
43    8.  OP_NOT now supports any UTF character not just single-byte ones.
45    9.  (*MARK) control verb is now supported by the JIT compiler.
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.)
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.
58    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
59        call.
61    14. As documented, (*COMMIT) is now confined to within a positive assertion.
63    15. It is now possible to link pcretest with libedit as an alternative to
64        libreadline.
66    16. (*COMMIT) control verb is now supported by the JIT compiler.
68    6.  The Unicode data tables have been updated to Unicode 6.1.0.
71    Version 8.30 04-February-2012
72    -----------------------------
74    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
75        name is defined in ctype.h.
77    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
78        only in quite long subpatterns.
80    3.  Removed the function pcre_info(), which has been obsolete and deprecated
81        since it was replaced by pcre_fullinfo() in February 2000.
83    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
84        match a (*MARK), and the match failed at the start of the subject, a
85        reference to memory before the start of the subject could occur. This bug
86        was introduced by fix 17 of release 8.21.
88    5.  A reference to an unset group with zero minimum repetition was giving
89        totally wrong answers (in non-JavaScript-compatibility mode). For example,
90        /(another)?(\1?)test/ matched against "hello world test". This bug was
91        introduced in release 8.13.
93    6.  Add support for 16-bit character strings (a large amount of work involving
94        many changes and refactorings).
96    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
97        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
98        from a file.
100    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
101        rounding is not applied in this particular case).
103    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
104        if they appear, or are escaped, in patterns.
106    10. Get rid of a number of -Wunused-but-set-variable warnings.
108    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
109        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
110        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
111        also returns the mark "x". This bug applied to capturing parentheses,
112        non-capturing parentheses, and atomic parentheses. It also applied to some
113        assertions.
115    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
116        information out of configure.ac instead of relying on pcre.h.generic, which
117        is not stored in the repository.
119    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
120        -lreadline.
122    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
124    15. Applied Graycode's patch to put the top-level frame on the stack rather
125        than the heap when not using the stack for recursion. This gives a
126        performance improvement in many cases when recursion is not deep.
128    16. Experimental code added to "pcretest -C" to output the stack frame size.
131    Version 8.21 12-Dec-2011
132    ------------------------
134    1.  Updating the JIT compiler.
136    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
137        are added as well.
139    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
140        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
141        calling _pcre_jit_exec. Some extra comments are added.
143    4.  (*MARK) settings inside atomic groups that do not contain any capturing
144        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
145        was introduced by change 18 for 8.20.
147    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
148        ECMA-262 standard.
150    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
151        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
152        This bug was probably introduced by change 9 of 8.13.
154    7.  While fixing 6 above, I noticed that a number of other items were being
155        incorrectly rejected as "not fixed length". This arose partly because newer
156        opcodes had not been added to the fixed-length checking code. I have (a)
157        corrected the bug and added tests for these items, and (b) arranged for an
158        error to occur if an unknown opcode is encountered while checking for fixed
159        length instead of just assuming "not fixed length". The items that were
160        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
161        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
162        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
164    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
165        being incorrectly compiled and would have given unpredicatble results.
167    9.  A possessively repeated subpattern with minimum repeat count greater than
168        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
169        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
170        the first (A) could occur when it should not.
172    10. Add a cast and remove a redundant test from the code.
174    11. JIT should use pcre_malloc/pcre_free for allocation.
176    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
177        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
178        is anything other than /usr, -L is still shown).
180    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
182    14. Perl does not support \N without a following name in a [] class; PCRE now
183        also gives an error.
185    15. If a forward reference was repeated with an upper limit of around 2000,
186        it caused the error "internal error: overran compiling workspace". The
187        maximum number of forward references (including repeats) was limited by the
188        internal workspace, and dependent on the LINK_SIZE. The code has been
189        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
190        the default depends on LINK_SIZE. There is a new upper limit (for safety)
191        of around 200,000 forward references. While doing this, I also speeded up
192        the filling in of repeated forward references.
194    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
195        incorrectly expecting the subject to contain another "a" after the start.
197    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
198        in the match, the SKIP should be ignored. This was not happening; instead
199        the SKIP was being treated as NOMATCH. For patterns such as
200        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
201        tested.
203    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
204        now much more compatible with Perl, in particular in cases where the result
205        is a non-match for a non-anchored pattern. For example, if
206        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
207        "m", where previously it did not return a name. A side effect of this
208        change is that for partial matches, the last encountered mark name is
209        returned, as for non matches. A number of tests that were previously not
210        Perl-compatible have been moved into the Perl-compatible test files. The
211        refactoring has had the pleasing side effect of removing one argument from
212        the match() function, thus reducing its stack requirements.
214    19. If the /S+ option was used in pcretest to study a pattern using JIT,
215        subsequent uses of /S (without +) incorrectly behaved like /S+.
217    21. Retrieve executable code size support for the JIT compiler and fixing
218        some warnings.
220    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
221        not work when the shorter character appeared right at the end of the
222        subject string.
224    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
225        systems.
227    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
228        output it when the /M option is used in pcretest.
230    25. The CheckMan script was not being included in the distribution. Also, added
231        an explicit "perl" to run Perl scripts from the PrepareRelease script
232        because this is reportedly needed in Windows.
234    26. If study data was being save in a file and studying had not found a set of
235        "starts with" bytes for the pattern, the data written to the file (though
236        never used) was taken from uninitialized memory and so caused valgrind to
237        complain.
239    27. Updated RunTest.bat as provided by Sheri Pierce.
241    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
243    29. Computation of memory usage for the table of capturing group names was
244        giving an unnecessarily large value.
247    Version 8.20 21-Oct-2011
248    ------------------------
250    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
251        a POSIX class. After further experiments with Perl, which convinced me that
252        Perl has bugs and confusions, a closing square bracket is no longer allowed
253        in a POSIX name. This bug also affected patterns with classes that started
254        with full stops.
256    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
257        captured substring, but while checking the failing first alternative,
258        substring 1 is temporarily captured. If the output vector supplied to
259        pcre_exec() was not big enough for this capture, the yield of the function
260        was still zero ("insufficient space for captured substrings"). This cannot
261        be totally fixed without adding another stack variable, which seems a lot
262        of expense for a edge case. However, I have improved the situation in cases
263        such as /(a)(b)x|abc/ matched against "abc", where the return code
264        indicates that fewer than the maximum number of slots in the ovector have
265        been set.
267    3.  Related to (2) above: when there are more back references in a pattern than
268        slots in the output vector, pcre_exec() uses temporary memory during
269        matching, and copies in the captures as far as possible afterwards. It was
270        using the entire output vector, but this conflicts with the specification
271        that only 2/3 is used for passing back captured substrings. Now it uses
272        only the first 2/3, for compatibility. This is, of course, another edge
273        case.
275    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
276        main code base, and can be used by building with --enable-jit. When this is
277        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
278        runtime --no-jit option is given.
280    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
281        ovector, the return from the function was zero, implying that there were
282        other matches that did not fit. The correct "exactly full" value is now
283        returned.
285    6.  If a subpattern that was called recursively or as a subroutine contained
286        (*PRUNE) or any other control that caused it to give a non-standard return,
287        invalid errors such as "Error -26 (nested recursion at the same subject
288        position)" or even infinite loops could occur.
290    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
291        computing the minimum length on reaching *ACCEPT, and so ended up with the
292        wrong value of 1 rather than 0. Further investigation indicates that
293        computing a minimum subject length in the presence of *ACCEPT is difficult
294        (think back references, subroutine calls), and so I have changed the code
295        so that no minimum is registered for a pattern that contains *ACCEPT.
297    8.  If (*THEN) was present in the first (true) branch of a conditional group,
298        it was not handled as intended. [But see 16 below.]
300    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
301        Sheri Pierce.
303    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
304        the first byte in a match must be "a".
306    11. Change 17 for 8.13 increased the recursion depth for patterns like
307        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
308        pattern contains any instances of (*THEN). If it does not, the old
309        optimizations are restored. It would be nice to do this on a per-group
310        basis, but at the moment that is not feasible.
312    12. In some environments, the output of pcretest -C is CRLF terminated. This
313        broke RunTest's code that checks for the link size. A single white space
314        character after the value is now allowed for.
316    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
317        For "fr", it uses the Windows-specific input and output files.
319    14. If (*THEN) appeared in a group that was called recursively or as a
320        subroutine, it did not work as intended. [But see next item.]
322    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
323        pattern fragments (but not containing any | characters). If A and B are
324        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
325        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
326        D. In other words, Perl considers parentheses that do not contain any |
327        characters to be part of a surrounding alternative, whereas PCRE was
328        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
329        differently. PCRE now behaves in the same way as Perl, except in the case
330        of subroutine/recursion calls such as (?1) which have in any case always
331        been different (but PCRE had them first :-).
333    16. Related to 15 above: Perl does not treat the | in a conditional group as
334        creating alternatives. Such a group is treated in the same way as an
335        ordinary group without any | characters when processing (*THEN). PCRE has
336        been changed to match Perl's behaviour.
338    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
339        RunGrepTest script failed.
341    18. Change 22 for version 13 caused atomic groups to use more stack. This is
342        inevitable for groups that contain captures, but it can lead to a lot of
343        stack use in large patterns. The old behaviour has been restored for atomic
344        groups that do not contain any capturing parentheses.
346    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
347        suppress the check for a minimum subject length at run time. (If it was
348        given to pcre_exec() or pcre_dfa_exec() it did work.)
350    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
351        fail to work when decoding hex characters in data strings in EBCDIC
352        environments.
354    21. It appears that in at least one Mac OS environment, the isxdigit() function
355        is implemented as a macro that evaluates to its argument more than once,
356        contravening the C 90 Standard (I haven't checked a later standard). There
357        was an instance in pcretest which caused it to go wrong when processing
358        \x{...} escapes in subject strings. The has been rewritten to avoid using
359        things like p++ in the argument of isxdigit().
362    Version 8.13 16-Aug-2011
363    ------------------------
365    1.  The Unicode data tables have been updated to Unicode 6.0.0.
367    2.  Two minor typos in pcre_internal.h have been fixed.
369    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
370        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
371        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
373    4.  There were a number of related bugs in the code for matching backrefences
374        caselessly in UTF-8 mode when codes for the characters concerned were
375        different numbers of bytes. For example, U+023A and U+2C65 are an upper
376        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
377        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
378        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
379        2-byte code at the end of the subject (it thought there wasn't enough data
380        left).
382    5.  Comprehensive information about what went wrong is now returned by
383        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
384        as the output vector has at least 2 elements. The offset of the start of
385        the failing character and a reason code are placed in the vector.
387    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
388        now returned is for the first byte of the failing character, instead of the
389        last byte inspected. This is an incompatible change, but I hope it is small
390        enough not to be a problem. It makes the returned offset consistent with
391        pcre_exec() and pcre_dfa_exec().
393    7.  pcretest now gives a text phrase as well as the error number when
394        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
395        failure, the offset and reason code are output.
397    8.  When \R was used with a maximizing quantifier it failed to skip backwards
398        over a \r\n pair if the subsequent match failed. Instead, it just skipped
399        back over a single character (\n). This seems wrong (because it treated the
400        two characters as a single entity when going forwards), conflicts with the
401        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
402        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
403        has been changed.
405    9.  Some internal refactoring has changed the processing so that the handling
406        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
407        time (the PCRE_DOTALL option was changed this way some time ago: version
408        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
409        which was always a bit of a fudge. It also means that there is one less
410        argument for the match() function, which reduces its stack requirements
411        slightly. This change also fixes an incompatibility with Perl: the pattern
412        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
414    10. More internal refactoring has drastically reduced the number of recursive
415        calls to match() for possessively repeated groups such as (abc)++ when
416        using pcre_exec().
418    11. While implementing 10, a number of bugs in the handling of groups were
419        discovered and fixed:
421        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
422        (a|)*(?1) gave a compile-time internal error.
423        ((a|)+)+  did not notice that the outer group could match an empty string.
424        (^a|^)+   was not marked as anchored.
425        (.*a|.*)+ was not marked as matching at start or after a newline.
427    12. Yet more internal refactoring has removed another argument from the match()
428        function. Special calls to this function are now indicated by setting a
429        value in a variable in the "match data" data block.
431    13. Be more explicit in pcre_study() instead of relying on "default" for
432        opcodes that mean there is no starting character; this means that when new
433        ones are added and accidentally left out of pcre_study(), testing should
434        pick them up.
436    14. The -s option of pcretest has been documented for ages as being an old
437        synonym of -m (show memory usage). I have changed it to mean "force study
438        for every regex", that is, assume /S for every regex. This is similar to -i
439        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
440        using it. It makes it easier to run collections of tests with and without
441        study enabled, and thereby test pcre_study() more easily. All the standard
442        tests are now run with and without -s (but some patterns can be marked as
443        "never study" - see 20 below).
445    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
446        restoration of the capturing data to the outer values was not happening
447        correctly.
449    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
450        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
451        pattern had matched an empty string, and so incorrectly returned a no
452        match.
454    17. There was optimizing code for the last branch of non-capturing parentheses,
455        and also for the obeyed branch of a conditional subexpression, which used
456        tail recursion to cut down on stack usage. Unfortunately, now that there is
457        the possibility of (*THEN) occurring in these branches, tail recursion is
458        no longer possible because the return has to be checked for (*THEN). These
459        two optimizations have therefore been removed. [But see 8.20/11 above.]
461    18. If a pattern containing \R was studied, it was assumed that \R always
462        matched two bytes, thus causing the minimum subject length to be
463        incorrectly computed because \R can also match just one byte.
465    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
466        was incorrectly computed.
468    20. If /S is present twice on a test pattern in pcretest input, it now
469        *disables* studying, thereby overriding the use of -s on the command line
470        (see 14 above). This is necessary for one or two tests to keep the output
471        identical in both cases.
473    21. When (*ACCEPT) was used in an assertion that matched an empty string and
474        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
476    22. When an atomic group that contained a capturing parenthesis was
477        successfully matched, but the branch in which it appeared failed, the
478        capturing was not being forgotten if a higher numbered group was later
479        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
480        group 1 to "a", when in fact it should be unset. This applied to multi-
481        branched capturing and non-capturing groups, repeated or not, and also to
482        positive assertions (capturing in negative assertions does not happen
483        in PCRE) and also to nested atomic groups.
485    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
486        subject after a captured substring, to make it easier to tell which of a
487        number of identical substrings has been captured.
489    24. The way atomic groups are processed by pcre_exec() has been changed so that
490        if they are repeated, backtracking one repetition now resets captured
491        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
492        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
493        "aaa". Previously, it would have been "a". As part of this code
494        refactoring, the way recursive calls are handled has also been changed.
496    25. If an assertion condition captured any substrings, they were not passed
497        back unless some other capturing happened later. For example, if
498        (?(?=(a))a) was matched against "a", no capturing was returned.
500    26. When studying a pattern that contained subroutine calls or assertions,
501        the code for finding the minimum length of a possible match was handling
502        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
503        group 1 called group 2 while simultaneously a separate group 2 called group
504        1). A stack overflow occurred in this case. I have fixed this by limiting
505        the recursion depth to 10.
507    27. Updated RunTest.bat in the distribution to the version supplied by Tom
508        Fortmann. This supports explicit test numbers on the command line, and has
509        argument validation and error reporting.
511    28. An instance of \X with an unlimited repeat could fail if at any point the
512        first character it looked at was a mark character.
514    29. Some minor code refactoring concerning Unicode properties and scripts
515        should reduce the stack requirement of match() slightly.
517    30. Added the '=' option to pcretest to check the setting of unused capturing
518        slots at the end of the pattern, which are documented as being -1, but are
519        not included in the return count.
521    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
522        compiled something random. Now it gives a compile-time error (as does
523        Perl).
525    32. A *MARK encountered during the processing of a positive assertion is now
526        recorded and passed back (compatible with Perl).
528    33. If --only-matching or --colour was set on a pcregrep call whose pattern
529        had alternative anchored branches, the search for a second match in a line
530        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
531        matched the line "0102" twice. The same bug affected patterns that started
532        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
533        twice.
535    34. Previously, PCRE did not allow quantification of assertions. However, Perl
536        does, and because of capturing effects, quantifying parenthesized
537        assertions may at times be useful. Quantifiers are now allowed for
538        parenthesized assertions.
540    35. A minor code tidy in pcre_compile() when checking options for \R usage.
542    36. \g was being checked for fancy things in a character class, when it should
543        just be a literal "g".
545    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
546        appearance of a nested POSIX class supersedes an apparent external class.
547        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
548        unescaped square brackets may also appear as part of class names. For
549        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
550        more like Perl. (But see 8.20/1 above.)
552    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
553        was because it thought it was \N{name}, which is not supported).
555    39. Add minix to OS list not supporting the -S option in pcretest.
557    40. PCRE tries to detect cases of infinite recursion at compile time, but it
558        cannot analyze patterns in sufficient detail to catch mutual recursions
559        such as ((?1))((?2)). There is now a runtime test that gives an error if a
560        subgroup is called recursively as a subpattern for a second time at the
561        same position in the subject string. In previous releases this might have
562        been caught by the recursion limit, or it might have run out of stack.
564    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
565        happen only once. PCRE was, however incorrectly giving a compile time error
566        "recursive call could loop indefinitely" because it cannot analyze the
567        pattern in sufficient detail. The compile time test no longer happens when
568        PCRE is compiling a conditional subpattern, but actual runaway loops are
569        now caught at runtime (see 40 above).
571    42. It seems that Perl allows any characters other than a closing parenthesis
572        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
573        has been changed to be the same.
575    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
576        as not to get warnings when autogen.sh is called. Also changed
577        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
579    44. To help people who use pcregrep to scan files containing exceedingly long
580        lines, the following changes have been made:
582        (a) The default value of the buffer size parameter has been increased from
583            8K to 20K. (The actual buffer used is three times this size.)
585        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
586            PCRE is built.
588        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
589            to be set at run time.
591        (d) Numerical values in pcregrep options can be followed by K or M, for
592            example --buffer-size=50K.
594        (e) If a line being scanned overflows pcregrep's buffer, an error is now
595            given and the return code is set to 2.
597    45. Add a pointer to the latest mark to the callout data block.
599    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
600        partial match of an empty string instead of no match. This was specific to
601        the use of ".".
603    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
604        complete match instead of a partial match. This bug was dependent on both
605        the PCRE_UTF8 and PCRE_DOTALL options being set.
607    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
608        starting byte set, because \b was not being ignored.
611    Version 8.12 15-Jan-2011
612    ------------------------
614    1.  Fixed some typos in the markup of the man pages, and wrote a script that
615        checks for such things as part of the documentation building process.
617    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
618        --match-limit and --recursion-limit options (added for 8.11). In
619        particular, this made one of the standard tests fail. (The integer value
620        went into the wrong half of a long int.)
622    3.  If the --colour option was given to pcregrep with -v (invert match), it
623        did strange things, either producing crazy output, or crashing. It should,
624        of course, ignore a request for colour when reporting lines that do not
625        match.
627    4.  Another pcregrep bug caused similar problems if --colour was specified with
628        -M (multiline) and the pattern match finished with a line ending.
630    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
631        matched in multiline mode, the following line was shown as part of the
632        match. This seems wrong, so I have changed it.
634    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
635        the check for further matches in the same line (so they could be coloured)
636        to overrun the end of the current line. If another match was found, it was
637        incorrectly shown (and then shown again when found in the next line).
639    7.  If pcregrep was compiled under Windows, there was a reference to the
640        function pcregrep_exit() before it was defined. I am assuming this was
641        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
642        reported by a user. I've moved the definition above the reference.
645    Version 8.11 10-Dec-2010
646    ------------------------
648    1.  (*THEN) was not working properly if there were untried alternatives prior
649        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
650        backtracked to try for "b" instead of moving to the next alternative branch
651        at the same level (in this case, to look for "c"). The Perl documentation
652        is clear that when (*THEN) is backtracked onto, it goes to the "next
653        alternative in the innermost enclosing group".
655    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
656        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
657        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
658        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
659        (*THEN).
661    3.  If \s appeared in a character class, it removed the VT character from
662        the class, even if it had been included by some previous item, for example
663        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
664        of \s, but is part of the POSIX "space" class.)
666    4.  A partial match never returns an empty string (because you can always
667        match an empty string at the end of the subject); however the checking for
668        an empty string was starting at the "start of match" point. This has been
669        changed to the "earliest inspected character" point, because the returned
670        data for a partial match starts at this character. This means that, for
671        example, /(?<=abc)def/ gives a partial match for the subject "abc"
672        (previously it gave "no match").
674    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
675        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
676        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
677        has an implication that the given string is incomplete (because a partial
678        match is preferred over a full match). For this reason, these items now
679        give a partial match in this situation. [Aside: previously, the one case
680        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
681        match rather than a full match, which was wrong by the old rules, but is
682        now correct.]
684    6.  There was a bug in the handling of #-introduced comments, recognized when
685        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
686        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
687        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
688        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
689        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
690        places in pcre_compile().
692    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
693        comments when looking ahead for named forward references to subpatterns,
694        the only newline sequence it recognized was NL. It now handles newlines
695        according to the set newline convention.
697    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
698        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
699        cater for a lack of strerror(). These oversights have been fixed.
701    9.  Added --match-limit and --recursion-limit to pcregrep.
703    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
705    11. When the -o option was used, pcregrep was setting a return code of 1, even
706        when matches were found, and --line-buffered was not being honoured.
708    12. Added an optional parentheses number to the -o and --only-matching options
709        of pcregrep.
711    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
712        can match an empty string. The code to do it in pcretest and pcredemo
713        needed fixing:
715        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
716            only one byte after an empty string match just before CRLF (this case
717            just got forgotten; "any" and "anycrlf" were OK).
719        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
720            mode when an empty string match preceded an ASCII character followed by
721            a non-ASCII character. (The code for advancing by one character rather
722            than one byte was nonsense.)
724        (c) The pcredemo.c sample program did not have any code at all to handle
725            the cases when CRLF is a valid newline sequence.
727    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
728        as a starting offset was within the subject string. There is now a new
729        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
730        negative or greater than the length of the string. In order to test this,
731        pcretest is extended to allow the setting of negative starting offsets.
733    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
734        starting offset points to the beginning of a UTF-8 character was
735        unnecessarily clumsy. I tidied it up.
737    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
738        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
740    17. Nobody had reported that the --include_dir option, which was added in
741        release 7.7 should have been called --include-dir (hyphen, not underscore)
742        for compatibility with GNU grep. I have changed it to --include-dir, but
743        left --include_dir as an undocumented synonym, and the same for
744        --exclude-dir, though that is not available in GNU grep, at least as of
745        release 2.5.4.
747    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
748        characters from a string of bytes) have been redefined so as not to use
749        loops, in order to improve performance in some environments. At the same
750        time, I abstracted some of the common code into auxiliary macros to save
751        repetition (this should not affect the compiled code).
753    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
754        compile-time error is now given if \c is not followed by an ASCII
755        character, that is, a byte less than 128. (In EBCDIC mode, the code is
756        different, and any byte value is allowed.)
758    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
759        START_OPTIMIZE option, which is now allowed at compile time - but just
760        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
761        to pcregrep and other applications that have no direct access to PCRE
762        options. The new /Y option in pcretest sets this option when calling
763        pcre_compile().
765    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
766        back references. Groups containing recursive back references were forced to
767        be atomic by that change, but in the case of named groups, the amount of
768        memory required was incorrectly computed, leading to "Failed: internal
769        error: code overflow". This has been fixed.
771    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
772        pcretest.c, to avoid build problems in some Borland environments.
775    Version 8.10 25-Jun-2010
776    ------------------------
778    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
779        THEN.
781    2.  (*ACCEPT) was not working when inside an atomic group.
783    3.  Inside a character class, \B is treated as a literal by default, but
784        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
785        causes the error). The code is unchanged, but I tidied the documentation.
787    4.  Inside a character class, PCRE always treated \R and \X as literals,
788        whereas Perl faults them if its -w option is set. I have changed PCRE so
789        that it faults them when PCRE_EXTRA is set.
791    5.  Added support for \N, which always matches any character other than
792        newline. (It is the same as "." when PCRE_DOTALL is not set.)
794    6.  When compiling pcregrep with newer versions of gcc which may have
795        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
796        declared with attribute warn_unused_result" were given. Just casting the
797        result to (void) does not stop the warnings; a more elaborate fudge is
798        needed. I've used a macro to implement this.
800    7.  Minor change to pcretest.c to avoid a compiler warning.
802    8.  Added four artifical Unicode properties to help with an option to make
803        \s etc use properties (see next item). The new properties are: Xan
804        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
806    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
807        use Unicode properties. (*UCP) at the start of a pattern can be used to set
808        this option. Modified pcretest to add /W to test this facility. Added
809        REG_UCP to make it available via the POSIX interface.
811    10. Added --line-buffered to pcregrep.
813    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
814        studied, and the match started with a letter with a code point greater than
815        127 whose first byte was different to the first byte of the other case of
816        the letter, the other case of this starting letter was not recognized
817        (#976).
819    12. If a pattern that was studied started with a repeated Unicode property
820        test, for example, \p{Nd}+, there was the theoretical possibility of
821        setting up an incorrect bitmap of starting bytes, but fortunately it could
822        not have actually happened in practice until change 8 above was made (it
823        added property types that matched character-matching opcodes).
825    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
826        possible starting bytes for non-anchored patterns.
828    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
829        \R, and also a number of cases that involve Unicode properties, both
830        explicit and implicit when PCRE_UCP is set.
832    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
833        input, it could crash or give wrong results if characters with values
834        greater than 0xc0 were present in the subject string. (Detail: it assumed
835        UTF-8 input when processing these items.)
837    16. Added a lot of (int) casts to avoid compiler warnings in systems where
838        size_t is 64-bit (#991).
840    17. Added a check for running out of memory when PCRE is compiled with
841        --disable-stack-for-recursion (#990).
843    18. If the last data line in a file for pcretest does not have a newline on
844        the end, a newline was missing in the output.
846    19. The default pcre_chartables.c file recognizes only ASCII characters (values
847        less than 128) in its various bitmaps. However, there is a facility for
848        generating tables according to the current locale when PCRE is compiled. It
849        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
850        characters, are recognized by isspace() and therefore were getting set in
851        these tables, and indeed these tables seem to approximate to ISO 8859. This
852        caused a problem in UTF-8 mode when pcre_study() was used to create a list
853        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
854        which of course cannot start UTF-8 characters. I have changed the code so
855        that only real ASCII characters (less than 128) and the correct starting
856        bytes for UTF-8 encodings are set for characters greater than 127 when in
857        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
858        altogether.)
860    20. Added the /T option to pcretest so as to be able to run tests with non-
861        standard character tables, thus making it possible to include the tests
862        used for 19 above in the standard set of tests.
864    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
865        reference to a subpattern the other side of a comment that contains an
866        opening parenthesis caused either an internal compiling error, or a
867        reference to the wrong subpattern.
870    Version 8.02 19-Mar-2010
871    ------------------------
873    1.  The Unicode data tables have been updated to Unicode 5.2.0.
875    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
876        configured.
878    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
879        original author of that file, following a query about its status.
881    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
882        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
884    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
885        quantifier applied to a forward-referencing subroutine call, could compile
886        incorrect code or give the error "internal error: previously-checked
887        referenced subpattern not found".
889    6.  Both MS Visual Studio and Symbian OS have problems with initializing
890        variables to point to external functions. For these systems, therefore,
891        pcre_malloc etc. are now initialized to local functions that call the
892        relevant global functions.
894    7.  There were two entries missing in the vectors called coptable and poptable
895        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
896        I've fixed the data, and added a kludgy way of testing at compile time that
897        the lengths are correct (equal to the number of opcodes).
899    8.  Following on from 7, I added a similar kludge to check the length of the
900        eint vector in pcreposix.c.
902    9.  Error texts for pcre_compile() are held as one long string to avoid too
903        much relocation at load time. To find a text, the string is searched,
904        counting zeros. There was no check for running off the end of the string,
905        which could happen if a new error number was added without updating the
906        string.
908    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
910    11. \K was not working if it appeared in an atomic group or in a group that
911        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
912        \K is "not well defined" if used in an assertion. PCRE now accepts it if
913        the assertion is positive, but not if it is negative.
915    12. Change 11 fortuitously reduced the size of the stack frame used in the
916        "match()" function of pcre_exec.c by one pointer. Forthcoming
917        implementation of support for (*MARK) will need an extra pointer on the
918        stack; I have reserved it now, so that the stack frame size does not
919        decrease.
921    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
922        item in branch that calls a recursion is a subroutine call - as in the
923        second branch in the above example - was incorrectly given the compile-
924        time error "recursive call could loop indefinitely" because pcre_compile()
925        was not correctly checking the subroutine for matching a non-empty string.
927    14. The checks for overrunning compiling workspace could trigger after an
928        overrun had occurred. This is a "should never occur" error, but it can be
929        triggered by pathological patterns such as hundreds of nested parentheses.
930        The checks now trigger 100 bytes before the end of the workspace.
932    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
935    Version 8.01 19-Jan-2010
936    ------------------------
938    1.  If a pattern contained a conditional subpattern with only one branch (in
939        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
940        computed the wrong minimum data length (which is of course zero for such
941        subpatterns). This could cause incorrect "no match" results.
943    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
944        the pattern is reset in the first branch, pcre_compile() failed with
945        "internal error: code overflow at offset...". This happened only when
946        the reset was to the original external option setting. (An optimization
947        abstracts leading options settings into an external setting, which was the
948        cause of this.)
950    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
951        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
952        assertion pattern did not match (meaning that the assertion was true), it
953        was incorrectly treated as false if the SKIP had been reached during the
954        matching. This also applied to assertions used as conditions.
956    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
957        assertion subpattern, including such a pattern used as a condition,
958        unpredictable results occurred, instead of the error return
961    5.  The C++ GlobalReplace function was not working like Perl for the special
962        situation when an empty string is matched. It now does the fancy magic
963        stuff that is necessary.
965    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
966        removed. (These were left over from very, very early versions of PCRE.)
968    7.  Some cosmetic changes to the code to make life easier when compiling it
969        as part of something else:
971        (a) Change DEBUG to PCRE_DEBUG.
973        (b) In pcre_compile(), rename the member of the "branch_chain" structure
974            called "current" as "current_branch", to prevent a collision with the
975            Linux macro when compiled as a kernel module.
977        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
978            prevent a collision with the Linux macro when compiled as a kernel
979            module.
981    8.  In pcre_compile() there are some checks for integer overflows that used to
982        cast potentially large values to (double). This has been changed to that
983        when building, a check for int64_t is made, and if it is found, it is used
984        instead, thus avoiding the use of floating point arithmetic. (There is no
985        other use of FP in PCRE.) If int64_t is not found, the fallback is to
986        double.
988    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
989        2005 (difference between two addresses compared to an unsigned value).
991    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
992        custom one, because of the following reported problem in Windows:
994          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
995              under Win32.
996          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
997              therefore missing the function definition.
998          - The compiler thus generates a "C" signature for the test function.
999          - The linker fails to find the "C" function.
1000          - PCRE fails to configure if asked to do so against libbz2.
1002    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1003        messages were output:
1005          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1006          rerunning libtoolize, to keep the correct libtool macros in-tree.
1007          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1009        I have done both of these things.
1011    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1012        most of the time, it *can* run out if it is given a pattern that contains a
1013        runaway infinite recursion. I updated the discussion in the pcrestack man
1014        page.
1016    13. Now that we have gone to the x.xx style of version numbers, the minor
1017        version may start with zero. Using 08 or 09 is a bad idea because users
1018        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1019        interpreted as invalid octal numbers. I've updated the previous comment in
1020        configure.ac, and also added a check that gives an error if 08 or 09 are
1021        used.
1023    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1024        causing partial matching to fail when the end of the subject matched \W
1025        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1027    15. There were some discrepancies between the declarations in pcre_internal.h
1028        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1029        their definitions. The declarations used "const uschar *" and the
1030        definitions used USPTR. Even though USPTR is normally defined as "const
1031        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1032        reported that: "This difference in casting confuses some C++ compilers, for
1033        example, SunCC recognizes above declarations as different functions and
1034        generates broken code for hbpcre." I have changed the declarations to use
1035        USPTR.
1037    16. GNU libtool is named differently on some systems. The autogen.sh script now
1038        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1039        (FreeBSD).
1041    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1042        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1043        comment: "Figure out how to create a longlong from a string: strtoll and
1044        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1045        instance, but it only takes 2 args instead of 3!"
1047    18. A subtle bug concerned with back references has been fixed by a change of
1048        specification, with a corresponding code fix. A pattern such as
1049        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1050        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1051        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1052        same bug. Such groups have to be quantified to be useful, or contained
1053        inside another quantified group. (If there's no repetition, the reference
1054        can never match.) The problem arises because, having left the group and
1055        moved on to the rest of the pattern, a later failure that backtracks into
1056        the group uses the captured value from the final iteration of the group
1057        rather than the correct earlier one. I have fixed this in PCRE by forcing
1058        any group that contains a reference to itself to be an atomic group; that
1059        is, there cannot be any backtracking into it once it has completed. This is
1060        similar to recursive and subroutine calls.
1063    Version 8.00 19-Oct-09
1064    ----------------------
1066    1.  The table for translating pcre_compile() error codes into POSIX error codes
1067        was out-of-date, and there was no check on the pcre_compile() error code
1068        being within the table. This could lead to an OK return being given in
1069        error.
1071    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1072        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1073        in a Windows environment.
1075    3.  The pcregrep --count option prints the count for each file even when it is
1076        zero, as does GNU grep. However, pcregrep was also printing all files when
1077        --files-with-matches was added. Now, when both options are given, it prints
1078        counts only for those files that have at least one match. (GNU grep just
1079        prints the file name in this circumstance, but including the count seems
1080        more useful - otherwise, why use --count?) Also ensured that the
1081        combination -clh just lists non-zero counts, with no names.
1083    4.  The long form of the pcregrep -F option was incorrectly implemented as
1084        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1085        but it seems right to fix it, and I didn't think it was worth preserving
1086        the old behaviour.
1088    5.  The command line items --regex=pattern and --regexp=pattern were not
1089        recognized by pcregrep, which required --regex pattern or --regexp pattern
1090        (with a space rather than an '='). The man page documented the '=' forms,
1091        which are compatible with GNU grep; these now work.
1093    6.  No libpcreposix.pc file was created for pkg-config; there was just
1094        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1096    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1097        when UCP support is not needed, by modifying the Python script that
1098        generates it from Unicode data files. This should not matter if the module
1099        is correctly used as a library, but I received one complaint about 50K of
1100        unwanted data. My guess is that the person linked everything into his
1101        program rather than using a library. Anyway, it does no harm.
1103    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1104        was a minimum greater than 1 for a wide character in a possessive
1105        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1106        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1107        character. Chaos in the form of incorrect output or a compiling loop could
1108        result.
1110    9.  The restrictions on what a pattern can contain when partial matching is
1111        requested for pcre_exec() have been removed. All patterns can now be
1112        partially matched by this function. In addition, if there are at least two
1113        slots in the offset vector, the offset of the earliest inspected character
1114        for the match and the offset of the end of the subject are set in them when
1115        PCRE_ERROR_PARTIAL is returned.
1117    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1118        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1119        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1120        and may be more useful for multi-segment matching.
1122    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1123        used to be given if ever the end of the subject was reached; now it is
1124        given only if matching could not proceed because another character was
1125        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1126        string "Z", which now yields "no match" instead of "partial match". In the
1127        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1128        final character ended with (*FAIL).
1130    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1131        if the pattern had a "must contain" character that was already found in the
1132        earlier partial match, unless partial matching was again requested. For
1133        example, with the pattern /dog.(body)?/, the "must contain" character is
1134        "g". If the first part-match was for the string "dog", restarting with
1135        "sbody" failed. This bug has been fixed.
1137    13. The string returned by pcre_dfa_exec() after a partial match has been
1138        changed so that it starts at the first inspected character rather than the
1139        first character of the match. This makes a difference only if the pattern
1140        starts with a lookbehind assertion or \b or \B (\K is not supported by
1141        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1142        matching functions compatible, and I think it's the right thing to do.
1144    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1145        so that the demonstration program is easily available in environments where
1146        PCRE has not been installed from source.
1148    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1149        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1150        library.
1152    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1153        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1154        is not the first non-POSIX option to be added. Clearly some people find
1155        these options useful.
1157    17. If a caller to the POSIX matching function regexec() passes a non-zero
1158        value for nmatch with a NULL value for pmatch, the value of
1159        nmatch is forced to zero.
1161    18. RunGrepTest did not have a test for the availability of the -u option of
1162        the diff command, as RunTest does. It now checks in the same way as
1163        RunTest, and also checks for the -b option.
1165    19. If an odd number of negated classes containing just a single character
1166        interposed, within parentheses, between a forward reference to a named
1167        subpattern and the definition of the subpattern, compilation crashed with
1168        an internal error, complaining that it could not find the referenced
1169        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1170        [The bug was that it was starting one character too far in when skipping
1171        over the character class, thus treating the ] as data rather than
1172        terminating the class. This meant it could skip too much.]
1174    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1175        /g option in pcretest when the pattern contains \K, which makes it possible
1176        to have an empty string match not at the start, even when the pattern is
1177        anchored. Updated pcretest and pcredemo to use this option.
1179    21. If the maximum number of capturing subpatterns in a recursion was greater
1180        than the maximum at the outer level, the higher number was returned, but
1181        with unset values at the outer level. The correct (outer level) value is
1182        now given.
1184    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1185        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1186        make it do so. The string so far is captured, making this feature
1187        compatible with Perl.
1189    23. The tests have been re-organized, adding tests 11 and 12, to make it
1190        possible to check the Perl 5.10 features against Perl 5.10.
1192    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1193        pattern matches a fixed length string. PCRE did not allow this; now it
1194        does. Neither allows recursion.
1196    25. I finally figured out how to implement a request to provide the minimum
1197        length of subject string that was needed in order to match a given pattern.
1198        (It was back references and recursion that I had previously got hung up
1199        on.) This code has now been added to pcre_study(); it finds a lower bound
1200        to the length of subject needed. It is not necessarily the greatest lower
1201        bound, but using it to avoid searching strings that are too short does give
1202        some useful speed-ups. The value is available to calling programs via
1203        pcre_fullinfo().
1205    26. While implementing 25, I discovered to my embarrassment that pcretest had
1206        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1207        study optimizations had never been tested with that matching function.
1208        Oops. What is worse, even when it was passed study data, there was a bug in
1209        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1210        were also very few tests of studied patterns with pcre_dfa_exec().
1212    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1213        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1214        on the other side of the coin, they are no longer allowed to have different
1215        names, because these cannot be distinguished in PCRE, and this has caused
1216        confusion. (This is a difference from Perl.)
1218    28. When duplicate subpattern names are present (necessarily with different
1219        numbers, as required by 27 above), and a test is made by name in a
1220        conditional pattern, either for a subpattern having been matched, or for
1221        recursion in such a pattern, all the associated numbered subpatterns are
1222        tested, and the overall condition is true if the condition is true for any
1223        one of them. This is the way Perl works, and is also more like the way
1224        testing by number works.
1227    Version 7.9 11-Apr-09
1228    ---------------------
1230    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1231        (pcretest), all targets were linked against these libraries. This included
1232        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1233        libraries. This caused unwanted dependencies to be created. This problem
1234        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1235        pcretest is linked with readline.
1237    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1238        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1239        moved outside it again, because FALSE and TRUE are already defined in AIX,
1240        but BOOL is not.
1242    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1243        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1245    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1246        hyphens) following file names and line numbers when outputting matching
1247        lines. This is not true; no spaces are inserted. I have also clarified the
1248        wording for the --colour (or --color) option.
1250    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1251        was not coloured; this is different to GNU grep, so I have changed it to be
1252        the same.
1254    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1255        each matching line was coloured. Now it goes on to look for further matches
1256        of any of the test patterns, which is the same behaviour as GNU grep.
1258    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1259        doesn't make sense to accept an empty string match in pcregrep, so I have
1260        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1261        seems to be how GNU grep behaves.
1263    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1264        start or after a newline", because the conditional assertion was not being
1265        correctly handled. The rule now is that both the assertion and what follows
1266        in the first alternative must satisfy the test.
1268    9.  If auto-callout was enabled in a pattern with a conditional group whose
1269        condition was an assertion, PCRE could crash during matching, both with
1270        pcre_exec() and pcre_dfa_exec().
1272    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1273        used for matching.
1275    11. Unicode property support in character classes was not working for
1276        characters (bytes) greater than 127 when not in UTF-8 mode.
1278    12. Added the -M command line option to pcretest.
1280    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1282    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1284    16. Added comments and documentation about mis-use of no_arg in the C++
1285        wrapper.
1287    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1288        from Martin Jerabek that uses macro names for all relevant character and
1289        string constants.
1291    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1292        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1293        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1294        these, but not everybody uses configure.
1296    19. A conditional group that had only one branch was not being correctly
1297        recognized as an item that could match an empty string. This meant that an
1298        enclosing group might also not be so recognized, causing infinite looping
1299        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1300        with the subject "ab", where knowledge that the repeated group can match
1301        nothing is needed in order to break the loop.
1303    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1304        exec(), but without supplying a callout function, matching went wrong.
1306    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1307        leak if the size of the offset vector was greater than 30. When the vector
1308        is smaller, the saved offsets during recursion go onto a local stack
1309        vector, but for larger vectors malloc() is used. It was failing to free
1310        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1311        error, in fact).
1313    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1314        heapframe that is used only when UTF-8 support is enabled. This caused no
1315        problem, but was untidy.
1317    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1318        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1319        included within another project.
1321    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1322        slightly modified by me:
1324          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1325              not building pcregrep.
1327          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1328              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1330    25. Forward references, both numeric and by name, in patterns that made use of
1331        duplicate group numbers, could behave incorrectly or give incorrect errors,
1332        because when scanning forward to find the reference group, PCRE was not
1333        taking into account the duplicate group numbers. A pattern such as
1334        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1336    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1337        the feature of a custom pointer more persuasive (as requested by a user).
1339    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1340        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1341        pre-defined.
1343    28. Added support for (*UTF8) at the start of a pattern.
1345    29. Arrange for flags added by the "release type" setting in CMake to be shown
1346        in the configuration summary.
1349    Version 7.8 05-Sep-08
1350    ---------------------
1352    1.  Replaced UCP searching code with optimized version as implemented for Ad
1353        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1354        stage table and inline lookup instead of a function, giving speed ups of 2
1355        to 5 times on some simple patterns that I tested. Permission was given to
1356        distribute the MultiStage2.py script that generates the tables (it's not in
1357        the tarball, but is in the Subversion repository).
1359    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1360        scripts.
1362    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1363        a group with a zero qualifier. The result of the study could be incorrect,
1364        or the function might crash, depending on the pattern.
1366    4.  Caseless matching was not working for non-ASCII characters in back
1367        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1368        It now works when Unicode Property Support is available.
1370    5.  In pcretest, an escape such as \x{de} in the data was always generating
1371        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1372        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1373        truncation.
1375    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1377    7.  Added two (int) casts to pcregrep when printing the difference of two
1378        pointers, in case they are 64-bit values.
1380    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1381        test 2 if it fails.
1383    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1384        and a #define of that name to empty if it is not externally set. This is to
1385        allow users of MSVC to set it if necessary.
1387    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1388        the convenience functions in the pcre_get.c source file.
1390    11. An option change at the start of a pattern that had top-level alternatives
1391        could cause overwriting and/or a crash. This command provoked a crash in
1392        some environments:
1394          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1396        This potential security problem was recorded as CVE-2008-2371.
1398    12. For a pattern where the match had to start at the beginning or immediately
1399        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1400        pcre_dfa_exec() could read past the end of the passed subject if there was
1401        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1402        pcretest so that it places the subject at the end of its malloc-ed buffer.
1404    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1405        exec() might read past the end of the data buffer in UTF-8 mode.
1407    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1408        the data contained the byte 0x85 as part of a UTF-8 character within its
1409        first line. This applied both to normal and DFA matching.
1411    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1412        /^[^d]*?$/8 failed to match "abc".
1414    16. Added a missing copyright notice to pcrecpp_internal.h.
1416    17. Make it more clear in the documentation that values returned from
1417        pcre_exec() in ovector are byte offsets, not character counts.
1419    18. Tidied a few places to stop certain compilers from issuing warnings.
1421    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1422        supplied by Stefan Weber. I made a further small update for 7.8 because
1423        there is a change of source arrangements: the pcre_searchfuncs.c module is
1424        replaced by pcre_ucd.c.
1427    Version 7.7 07-May-08
1428    ---------------------
1430    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1431        a string to a long long, pretend we don't even have a long long." This is
1432        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1434    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1435        pre-7.6 versions, which defined a global no_arg variable instead of putting
1436        it in the RE class. (See also #8 below.)
1438    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1439        Lopes.
1441    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1443        (1) The include/exclude patterns were being applied to the whole pathnames
1444            of files, instead of just to the final components.
1446        (2) If there was more than one level of directory, the subdirectories were
1447            skipped unless they satisfied the include/exclude conditions. This is
1448            inconsistent with GNU grep (and could even be seen as contrary to the
1449            pcregrep specification - which I improved to make it absolutely clear).
1450            The action now is always to scan all levels of directory, and just
1451            apply the include/exclude patterns to regular files.
1453    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1454        --exclude_dir in the tests to avoid scanning .svn directories.
1456    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1457        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1458        doesn't support NULs in patterns.
1460    7.  Added some missing "const"s to declarations of static tables in
1461        pcre_compile.c and pcre_dfa_exec.c.
1463    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1464        caused by fix #2  above. (Subsequently also a second patch to fix the
1465        first patch. And a third patch - this was a messy problem.)
1467    9.  Applied Craig's patch to remove the use of push_back().
1469    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1470        matching function regexec().
1472    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1473        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1474        references. PCRE supports relative numbers with this syntax (I don't think
1475        Oniguruma does).
1477    12. Previously, a group with a zero repeat such as (...){0} was completely
1478        omitted from the compiled regex. However, this means that if the group
1479        was called as a subroutine from elsewhere in the pattern, things went wrong
1480        (an internal error was given). Such groups are now left in the compiled
1481        pattern, with a new opcode that causes them to be skipped at execution
1482        time.
1484    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1485        to the way PCRE behaves:
1487        (a) A lone ] character is dis-allowed (Perl treats it as data).
1489        (b) A back reference to an unmatched subpattern matches an empty string
1490            (Perl fails the current match path).
1492        (c) A data ] in a character class must be notated as \] because if the
1493            first data character in a class is ], it defines an empty class. (In
1494            Perl it is not possible to have an empty class.) The empty class []
1495            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1496            The negative empty class [^] matches any one character, independently
1497            of the DOTALL setting.
1499    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1500        non-existent subpattern following a character class starting with ']' and
1501        containing () gave an internal compiling error instead of "reference to
1502        non-existent subpattern". Fortunately, when the pattern did exist, the
1503        compiled code was correct. (When scanning forwards to check for the
1504        existencd of the subpattern, it was treating the data ']' as terminating
1505        the class, so got the count wrong. When actually compiling, the reference
1506        was subsequently set up correctly.)
1508    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1509        it was being rejected as not supported by pcre_dfa_exec(), even though
1510        other assertions are supported. I have made pcre_dfa_exec() support
1511        (*FAIL).
1513    16. The implementation of 13c above involved the invention of a new opcode,
1514        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1515        cannot be changed at match time, I realized I could make a small
1516        improvement to matching performance by compiling OP_ALLANY instead of
1517        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1518        on the OP_ANY path.
1520    17. Compiling pcretest on Windows with readline support failed without the
1521        following two fixes: (1) Make the unistd.h include conditional on
1522        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1524    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1525        ncurses library to be included for pcretest when ReadLine support is
1526        requested, but also to allow for it to be overridden. This patch came from
1527        Daniel Bergström.
1529    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1530        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1531        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1532        spotting this.
1535    Version 7.6 28-Jan-08
1536    ---------------------
1538    1.  A character class containing a very large number of characters with
1539        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1540        overflow.
1542    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1543        HAVE_LONG_LONG is not defined.
1545    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1546        bring them up to date and include new features. This patch includes:
1548        - Fixed PH's badly added libz and libbz2 support.
1549        - Fixed a problem with static linking.
1550        - Added pcredemo. [But later removed - see 7 below.]
1551        - Fixed dftables problem and added an option.
1552        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1553            HAVE_LONG_LONG.
1554        - Added readline support for pcretest.
1555        - Added an listing of the option settings after cmake has run.
1557    4.  A user submitted a patch to Makefile that makes it easy to create
1558        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1559        Makefile.am that cause it to include this special target, without
1560        affecting anything else. Note that the same mingw target plus all
1561        the other distribution libraries and programs are now supported
1562        when configuring with CMake (see 6 below) instead of with
1563        Configure/Make.
1565    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1566        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1567        exported in the Windows port". It has not yet been confirmed that the patch
1568        solves the problem, but it does no harm.
1570    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1571        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1572        with CMake, and also correct the comment about stack recursion.
1574    7.  Remove the automatic building of pcredemo from the ./configure system and
1575        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1576        of a program that users should build themselves after PCRE is installed, so
1577        building it automatically is not really right. What is more, it gave
1578        trouble in some build environments.
1580    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1583    Version 7.5 10-Jan-08
1584    ---------------------
1586    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1587        values in parens when parsing an RE using the C++ wrapper."
1589    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1590        Characters greater than 255 were excluded from the class instead of being
1591        included.
1593    3.  The same bug as (2) above applied to negated POSIX classes such as
1594        [:^space:].
1596    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1597        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1598        I have changed it.
1600    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1601        first named subpattern) and a construct such as (?&a) would reference the
1602        first named subpattern whose name started with "a" (in other words, the
1603        length check was missing). Both these problems are fixed. "Subpattern name
1604        expected" is now given for (?&) (a zero-length name), and this patch also
1605        makes it give the same error for \k'' (previously it complained that that
1606        was a reference to a non-existent subpattern).
1608    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1609        this is right because (?- can be followed by option settings as well as by
1610        digits. I have, however, made the messages clearer.
1612    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1613        than the number used in the conditional) now cause a compile-time error.
1614        This is actually not compatible with Perl, which accepts such patterns, but
1615        treats the conditional as always being FALSE (as PCRE used to), but it
1616        seems to me that giving a diagnostic is better.
1618    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1619        and messages.
1621    9.  Fix two occurrences of "backslash" in comments that should have been
1622        "backspace".
1624    10. Remove two redundant lines of code that can never be obeyed (their function
1625        was moved elsewhere).
1627    11. The program that makes PCRE's Unicode character property table had a bug
1628        which caused it to generate incorrect table entries for sequences of
1629        characters that have the same character type, but are in different scripts.
1630        It amalgamated them into a single range, with the script of the first of
1631        them. In other words, some characters were in the wrong script. There were
1632        thirteen such cases, affecting characters in the following ranges:
1634          U+002b0 - U+002c1
1635          U+0060c - U+0060d
1636          U+0061e - U+00612
1637          U+0064b - U+0065e
1638          U+0074d - U+0076d
1639          U+01800 - U+01805
1640          U+01d00 - U+01d77
1641          U+01d9b - U+01dbf
1642          U+0200b - U+0200f
1643          U+030fc - U+030fe
1644          U+03260 - U+0327f
1645          U+0fb46 - U+0fbb1
1646          U+10450 - U+1049d
1648    12. The -o option (show only the matching part of a line) for pcregrep was not
1649        compatible with GNU grep in that, if there was more than one match in a
1650        line, it showed only the first of them. It now behaves in the same way as
1651        GNU grep.
1653    13. If the -o and -v options were combined for pcregrep, it printed a blank
1654        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1655        does the same. The return code can be used to tell if there were any
1656        non-matching lines.
1658    14. Added --file-offsets and --line-offsets to pcregrep.
1660    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1661        infinitely looping recursion. The bug was that positive lookaheads were not
1662        being skipped when checking for a possible empty match (negative lookaheads
1663        and both kinds of lookbehind were skipped).
1665    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1666        inclusion of <windows.h> to before rather than after the definition of
1667        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1669    17. Specifying a possessive quantifier with a specific limit for a Unicode
1670        character property caused pcre_compile() to compile bad code, which led at
1671        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1672        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1673        caused the error; without that there was no problem.
1675    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1677    19. Added --enable-pcretest-libreadline.
1679    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1680        RE::GlobalReplace(). As a result, the number of replacements returned was
1681        double what it should be. I removed one of the increments, but Craig sent a
1682        later patch that removed the other one (the right fix) and added unit tests
1683        that check the return values (which was not done before).
1685    21. Several CMake things:
1687        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1688            the names libpcre and libpcreposix, not just pcre and pcreposix.
1690        (2) The above change means that pcretest and pcregrep are now correctly
1691            linked with the newly-built libraries, not previously installed ones.
1695    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1696        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1697        UTF-8 newline character). The key issue is that the pattern starts .*;
1698        this means that the match must be either at the beginning, or after a
1699        newline. The bug was in the code for advancing after a failed match and
1700        checking that the new position followed a newline. It was not taking
1701        account of UTF-8 characters correctly.
1703    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1704        character classes. PCRE was not treating the sequence [:...:] as a
1705        character class unless the ... were all letters. Perl, however, seems to
1706        allow any characters between [: and :], though of course it rejects as
1707        unknown any "names" that contain non-letters, because all the known class
1708        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1709        for example, whereas PCRE did not - it did not recognize a POSIX character
1710        class. This seemed a bit dangerous, so the code has been changed to be
1711        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1712        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1713        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1714        Perl does, and where it didn't before.
1716    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1717        Windows environments %n is disabled by default.
1720    Version 7.4 21-Sep-07
1721    ---------------------
1723    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1724        means that a class such as [\s] counted as "explicit reference to CR or
1725        LF". That isn't really right - the whole point of the change was to try to
1726        help when there was an actual mention of one of the two characters. So now
1727        the change happens only if \r or \n (or a literal CR or LF) character is
1728        encountered.
1730    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1731        of both had grown to the point where there were only 3 bits left.
1732        Fortunately, there was spare space in the data structure, and so I have
1733        moved the internal flags into a new 16-bit field to free up more option
1734        bits.
1736    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1737        but did not set the internal JCHANGED flag - either of these is enough to
1738        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1739        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1740        start sets both bits.
1742    4.  Added options (at build time, compile time, exec time) to change \R from
1743        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1745    5.  doc/pcresyntax.html was missing from the distribution.
1747    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1748        compatibility, even though it is no longer used.
1750    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1751        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1752        windows.h file is present (where different names are used). [This was
1753        reversed later after testing - see 16 below.]
1755    8.  Changed all #include <config.h> to #include "config.h". There were also
1756        some further <pcre.h> cases that I changed to "pcre.h".
1758    9.  When pcregrep was used with the --colour option, it missed the line ending
1759        sequence off the lines that it output.
1761    10. It was pointed out to me that arrays of string pointers cause lots of
1762        relocations when a shared library is dynamically loaded. A technique of
1763        using a single long string with a table of offsets can drastically reduce
1764        these. I have refactored PCRE in four places to do this. The result is
1765        dramatic:
1767          Originally:                          290
1768          After changing UCP table:            187
1769          After changing error message table:   43
1770          After changing table of "verbs"       36
1771          After changing table of Posix names   22
1773        Thanks to the folks working on Gregex for glib for this insight.
1775    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1776        unicode-properties was also set.
1778    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1780    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1781        checked only for CRLF.
1783    14. Added casts to pcretest.c to avoid compiler warnings.
1785    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1787    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1788        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1789        entirely. This removes changes made in 7 above.
1791    17. The CMake files have been updated, and there is now more information about
1792        building with CMake in the NON-UNIX-USE document.
1795    Version 7.3 28-Aug-07
1796    ---------------------
1798     1. In the rejigging of the build system that eventually resulted in 7.1, the
1799        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1800        brackets there is not right, since it causes compilers to look for an
1801        installed pcre.h, not the version that is in the source that is being
1802        compiled (which of course may be different). I have changed it back to:
1804          #include "pcre.h"
1806        I have a vague recollection that the change was concerned with compiling in
1807        different directories, but in the new build system, that is taken care of
1808        by the VPATH setting the Makefile.
1810     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1811        when the subject happened to end in the byte 0x85 (e.g. if the last
1812        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1813        characters but of course it shouldn't be taken as a newline when it is part
1814        of another character. The bug was that, for an unlimited repeat of . in
1815        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1816        characters when looking for a newline.
1818     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1820     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1821        in debug output.
1823     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1824        long printing in the pcrecpp unittest when running under MinGW.
1826     6. ESC_K was left out of the EBCDIC table.
1828     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1829        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1830        limit also applies to "virtual nesting" when a pattern is recursive, and in
1831        this case 1000 isn't so big. I have been able to remove this limit at the
1832        expense of backing off one optimization in certain circumstances. Normally,
1833        when pcre_exec() would call its internal match() function recursively and
1834        immediately return the result unconditionally, it uses a "tail recursion"
1835        feature to save stack. However, when a subpattern that can match an empty
1836        string has an unlimited repetition quantifier, it no longer makes this
1837        optimization. That gives it a stack frame in which to save the data for
1838        checking that an empty string has been matched. Previously this was taken
1839        from the 1000-entry workspace that had been reserved. So now there is no
1840        explicit limit, but more stack is used.
1842     8. Applied Daniel's patches to solve problems with the import/export magic
1843        syntax that is required for Windows, and which was going wrong for the
1844        pcreposix and pcrecpp parts of the library. These were overlooked when this
1845        problem was solved for the main library.
1847     9. There were some crude static tests to avoid integer overflow when computing
1848        the size of patterns that contain repeated groups with explicit upper
1849        limits. As the maximum quantifier is 65535, the maximum group length was
1850        set at 30,000 so that the product of these two numbers did not overflow a
1851        32-bit integer. However, it turns out that people want to use groups that
1852        are longer than 30,000 bytes (though not repeat them that many times).
1853        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1854        made it possible to implement the integer overflow checks in a much more
1855        dynamic way, which I have now done. The artificial limitation on group
1856        length has been removed - we now have only the limit on the total length of
1857        the compiled pattern, which depends on the LINK_SIZE setting.
1859    10. Fixed a bug in the documentation for get/copy named substring when
1860        duplicate names are permitted. If none of the named substrings are set, the
1861        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1862        empty string.
1864    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1865        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1866        because the ] is interpreted as the first data character and the
1867        terminating ] is not found. PCRE has been made compatible with Perl in this
1868        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1869        cause memory overwriting.
1871    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1872        string has been matched (to stop an infinite loop). It was not recognizing
1873        a conditional subpattern that could match an empty string if that
1874        subpattern was within another subpattern. For example, it looped when
1875        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1876        condition was not nested. This bug has been fixed.
1878    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1879        past the start of the subject in the presence of bytes with the top bit
1880        set, for example "\x8aBCD".
1882    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1883        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1885    14. Optimized (?!) to (*FAIL).
1887    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1888        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1889        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1890        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1891        does: it's just the validity check that is more restrictive.
1893    16. Inserted checks for integer overflows during escape sequence (backslash)
1894        processing, and also fixed erroneous offset values for syntax errors during
1895        backslash processing.
1897    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1898        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1900    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1901        caused an overrun.
1903    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1904        something other than just ASCII characters) inside a group that had an
1905        unlimited repeat caused a loop at compile time (while checking to see
1906        whether the group could match an empty string).
1908    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1909        [\P{Any}] did so. (Error in the code for printing property names.)
1911    21. An orphan \E inside a character class could cause a crash.
1913    22. A repeated capturing bracket such as (A)? could cause a wild memory
1914        reference during compilation.
1916    23. There are several functions in pcre_compile() that scan along a compiled
1917        expression for various reasons (e.g. to see if it's fixed length for look
1918        behind). There were bugs in these functions when a repeated \p or \P was
1919        present in the pattern. These operators have additional parameters compared
1920        with \d, etc, and these were not being taken into account when moving along
1921        the compiled data. Specifically:
1923        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1924            length.
1926        (b) An item such as \pL+ within a repeated group could cause crashes or
1927            loops.
1929        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1930            "reference to non-existent subpattern" error.
1932        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1934    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1935        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1937    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1939    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1940        character were causing crashes (broken optimization).
1942    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1943        \p or \P) caused a compile-time loop.
1945    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1946        break. For example, the unstudied pattern [\r\n]A does not match the string
1947        "\r\nA" because change 7.0/46 below moves the current point on by two
1948        characters after failing to match at the start. However, the pattern \nA
1949        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1950        the same is true. There doesn't seem any very clean way out of this, but
1951        what I have chosen to do makes the common cases work: PCRE now takes note
1952        of whether there can be an explicit match for \r or \n anywhere in the
1953        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1954        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1955        pattern has explicit CR or LF references.
1957    29. Added (*CR) etc for changing newline setting at start of pattern.
1960    Version 7.2 19-Jun-07
1961    ---------------------
1963     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1964        which is apparently normally available under Windows.
1966     2. Re-jig the pcregrep tests with different newline settings in an attempt
1967        to make them independent of the local environment's newline setting.
1969     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1971     4. Some of the "internals" tests were previously cut out when the link size
1972        was not 2, because the output contained actual offsets. The recent new
1973        "Z" feature of pcretest means that these can be cut out, making the tests
1974        usable with all link sizes.
1976     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1977        stack recursion. This gives a massive performance boost under BSD, but just
1978        a small improvement under Linux. However, it saves one field in the frame
1979        in all cases.
1981     6. Added more features from the forthcoming Perl 5.10:
1983        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1984            recursion call. It refers to the nth most recently opened parentheses.
1986        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1987            to be opened parentheses.
1989        (c) Conditions that refer to capturing parentheses can be specified
1990            relatively, for example, (?(-2)... or (?(+3)...
1992        (d) \K resets the start of the current match so that everything before
1993            is not part of it.
1995        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1997        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1998            reference syntax.
2000        (g) (?| introduces a group in which the numbering of parentheses in each
2001            alternative starts with the same number.
2003        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2005     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2008     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2009        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2010        for detecting groups that can match an empty string.
2012     9. A pattern with a very large number of alternatives (more than several
2013        hundred) was running out of internal workspace during the pre-compile
2014        phase, where pcre_compile() figures out how much memory will be needed. A
2015        bit of new cunning has reduced the workspace needed for groups with
2016        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2017        workspace instead of running out of the 4096 that are available.
2019    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2021    11. Applied patch from Google to remove an optimization that didn't quite work.
2022        The report of the bug said:
2024          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2025          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2026          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2028    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2029        it matched the wrong number of bytes.
2032    Version 7.1 24-Apr-07
2033    ---------------------
2035     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2036        that is more "standard", making use of automake and other Autotools. There
2037        is some re-arrangement of the files and adjustment of comments consequent
2038        on this.
2040     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2041        for recursive directory scanning broke on some systems because the files
2042        are not scanned in any specific order and on different systems the order
2043        was different. A call to "sort" has been inserted into RunGrepTest for the
2044        approprate test as a short-term fix. In the longer term there may be an
2045        alternative.
2047     3. I had an email from Eric Raymond about problems translating some of PCRE's
2048        man pages to HTML (despite the fact that I distribute HTML pages, some
2049        people do their own conversions for various reasons). The problems
2050        concerned the use of low-level troff macros .br and .in. I have therefore
2051        removed all such uses from the man pages (some were redundant, some could
2052        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2053        HTML has been updated to handle .nf/.fi and to complain if it encounters
2054        .br or .in.
2056     4. Updated comments in configure.ac that get placed in config.h.in and also
2057        arranged for config.h to be included in the distribution, with the name
2058        config.h.generic, for the benefit of those who have to compile without
2059        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2061     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2062        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2063        makevp.bat for the current PCRE, using the additional files
2064        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2066     6. A Windows user reported a minor discrepancy with test 2, which turned out
2067        to be caused by a trailing space on an input line that had got lost in his
2068        copy. The trailing space was an accident, so I've just removed it.
2070     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2071        that is needed.
2073     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2074        as "const" (a) because they are and (b) because it helps the PHP
2075        maintainers who have recently made a script to detect big data structures
2076        in the php code that should be moved to the .rodata section. I remembered
2077        to update Builducptable as well, so it won't revert if ucptable.h is ever
2078        re-created.
2080     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2081        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2082        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2083        support is not required. This saves 1.5-2K of code, which is important in
2084        some applications.
2086        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2087        so as not to refer to the tables, even though these functions will never be
2088        called when UTF-8 support is disabled. Otherwise there are problems with a
2089        shared library.
2091    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2093        (a) It was defining its arguments as char * instead of void *.
2095        (b) It was assuming that all moves were upwards in memory; this was true
2096            a long time ago when I wrote it, but is no longer the case.
2098        The emulated memove() is provided for those environments that have neither
2099        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2100        is clearly not the case, as these two bugs were recently reported.
2102    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2103        and Detrail to create the HTML documentation, the .txt form of the man
2104        pages, and it removes trailing spaces from listed files. It also creates
2105        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2106        case, it wraps all the #defines with #ifndefs. This script should be run
2107        before "make dist".
2109    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2110        with Unicode property support.
2112        (a) For a maximizing quantifier, if the two different cases of the
2113            character were of different lengths in their UTF-8 codings (there are
2114            some cases like this - I found 11), and the matching function had to
2115            back up over a mixture of the two cases, it incorrectly assumed they
2116            were both the same length.
2118        (b) When PCRE was configured to use the heap rather than the stack for
2119            recursion during matching, it was not correctly preserving the data for
2120            the other case of a UTF-8 character when checking ahead for a match
2121            while processing a minimizing repeat. If the check also involved
2122            matching a wide character, but failed, corruption could cause an
2123            erroneous result when trying to check for a repeat of the original
2124            character.
2126    13. Some tidying changes to the testing mechanism:
2128        (a) The RunTest script now detects the internal link size and whether there
2129            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2130            values substituted by "configure". (The RunGrepTest script already did
2131            this for UTF-8.) The configure.ac script no longer substitutes the
2132            relevant variables.
2134        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2135            with length and offset values. This means that the output is different
2136            for different internal link sizes. Test 2 is skipped for link sizes
2137            other than 2 because of this, bypassing the problem. Unfortunately,
2138            there was also a test in test 3 (the locale tests) that used /B and
2139            failed for link sizes other than 2. Rather than cut the whole test out,
2140            I have added a new /Z option to pcretest that replaces the length and
2141            offset values with spaces. This is now used to make test 3 independent
2142            of link size. (Test 2 will be tidied up later.)
2144    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2145        segmentation fault instead of returning the appropriate error message.
2147    15. In multiline mode when the newline sequence was set to "any", the pattern
2148        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2149        This doesn't seem right; it now treats the CRLF combination as the line
2150        ending, and so does not match in that case. It's only a pattern such as ^$
2151        that would hit this one: something like ^ABC$ would have failed after \r
2152        and then tried again after \r\n.
2154    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2155        in an attempt to make files that differ only in their line terminators
2156        compare equal. This works on Linux.
2158    17. Under certain error circumstances pcregrep might try to free random memory
2159        as it exited. This is now fixed, thanks to valgrind.
2161    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2162        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2163        was because its rules for how to advance for /g after matching an empty
2164        string at the end of a line did not allow for this case. They now check for
2165        it specially.
2167    20. pcretest is supposed to handle patterns and data of any length, by
2168        extending its buffers when necessary. It was getting this wrong when the
2169        buffer for a data line had to be extended.
2171    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2172        CRLF as a newline sequence.
2174    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2175        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2176        I have nevertheless tidied it up.
2178    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2180    24. Added a man page for pcre-config.
2183    Version 7.0 19-Dec-06
2184    ---------------------
2186     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2187        moving to gcc 4.1.1.
2189     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2190        sys/time.h, which is documented as needed for this function. It doesn't
2191        seem to matter on Linux, but it showed up on some releases of OS X.
2193     3. It seems that there are systems where bytes whose values are greater than
2194        127 match isprint() in the "C" locale. The "C" locale should be the
2195        default when a C program starts up. In most systems, only ASCII printing
2196        characters match isprint(). This difference caused the output from pcretest
2197        to vary, making some of the tests fail. I have changed pcretest so that:
2199        (a) When it is outputting text in the compiled version of a pattern, bytes
2200            other than 32-126 are always shown as hex escapes.
2202        (b) When it is outputting text that is a matched part of a subject string,
2203            it does the same, unless a different locale has been set for the match
2204            (using the /L modifier). In this case, it uses isprint() to decide.
2206     4. Fixed a major bug that caused incorrect computation of the amount of memory
2207        required for a compiled pattern when options that changed within the
2208        pattern affected the logic of the preliminary scan that determines the
2209        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2210        that the computed length was too small. The symptoms of this bug were
2211        either the PCRE error "internal error: code overflow" from pcre_compile(),
2212        or a glibc crash with a message such as "pcretest: free(): invalid next
2213        size (fast)". Examples of patterns that provoked this bug (shown in
2214        pcretest format) are:
2216          /(?-x: )/x
2217          /(?x)(?-x: \s*#\s*)/
2218          /((?i)[\x{c0}])/8
2219          /(?i:[\x{c0}])/8
2221        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2222        is now done differently.
2224     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2225        wrapper classes; (b) implement a new function in the C++ scanner that is
2226        more efficient than the old way of doing things because it avoids levels of
2227        recursion in the regex matching; (c) add a paragraph to the documentation
2228        for the FullMatch() function.
2230     6. The escape sequence \n was being treated as whatever was defined as
2231        "newline". Not only was this contrary to the documentation, which states
2232        that \n is character 10 (hex 0A), but it also went horribly wrong when
2233        "newline" was defined as CRLF. This has been fixed.
2235     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2236        was being set to -1 for the "end of line" case (supposedly a value that no
2237        character can have). Though this value is never used (the check for end of
2238        line is "zero bytes in current character"), it caused compiler complaints.
2239        I've changed it to 0xffffffff.
2241     8. In pcre_version.c, the version string was being built by a sequence of
2242        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2243        string (as it is for production releases) called a macro with an empty
2244        argument. The C standard says the result of this is undefined. The gcc
2245        compiler treats it as an empty string (which was what was wanted) but it is
2246        reported that Visual C gives an error. The source has been hacked around to
2247        avoid this problem.
2249     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2250        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2251        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2252        of them did).
2254    10. Originally, pcretest opened its input and output without "b"; then I was
2255        told that "b" was needed in some environments, so it was added for release
2256        5.0 to both the input and output. (It makes no difference on Unix-like
2257        systems.) Later I was told that it is wrong for the input on Windows. I've
2258        now abstracted the modes into two macros, to make it easier to fiddle with
2259        them, and removed "b" from the input mode under Windows.
2261    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2263    12. Added -help and --help to pcretest as an official way of being reminded
2264        of the options.
2266    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2267        and pcrecpp.cc because they annoy compilers at high warning levels.
2269    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2271    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2272        scripts are not C programs :-) and which was not noticed because it works
2273        on Linux.
2275    16. pcretest is supposed to handle any length of pattern and data line (as one
2276        line or as a continued sequence of lines) by extending its input buffer if
2277        necessary. This feature was broken for very long pattern lines, leading to
2278        a string of junk being passed to pcre_compile() if the pattern was longer
2279        than about 50K.
2281    17. I have done a major re-factoring of the way pcre_compile() computes the
2282        amount of memory needed for a compiled pattern. Previously, there was code
2283        that made a preliminary scan of the pattern in order to do this. That was
2284        OK when PCRE was new, but as the facilities have expanded, it has become
2285        harder and harder to keep it in step with the real compile phase, and there
2286        have been a number of bugs (see for example, 4 above). I have now found a
2287        cunning way of running the real compile function in a "fake" mode that
2288        enables it to compute how much memory it would need, while actually only
2289        ever using a few hundred bytes of working memory and without too many
2290        tests of the mode. This should make future maintenance and development
2291        easier. A side effect of this work is that the limit of 200 on the nesting
2292        depth of parentheses has been removed (though this was never a serious
2293        limitation, I suspect). However, there is a downside: pcre_compile() now
2294        runs more slowly than before (30% or more, depending on the pattern). I
2295        hope this isn't a big issue. There is no effect on runtime performance.
2297    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2298        newline (only possible for the last line of a file) and it was a
2299        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2301    19. Added additional timing features to pcretest. (1) The -tm option now times
2302        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2303        separate command line item, by a number that specifies the number of
2304        repeats to use when timing. The default is 50000; this gives better
2305        precision, but takes uncomfortably long for very large patterns.
2307    20. Extended pcre_study() to be more clever in cases where a branch of a
2308        subpattern has no definite first character. For example, (a*|b*)[cd] would
2309        previously give no result from pcre_study(). Now it recognizes that the
2310        first character must be a, b, c, or d.
2312    21. There was an incorrect error "recursive call could loop indefinitely" if
2313        a subpattern (or the entire pattern) that was being tested for matching an
2314        empty string contained only one non-empty item after a nested subpattern.
2315        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2316        incorrectly, because the \d was being skipped in the check.
2318    22. The pcretest program now has a new pattern option /B and a command line
2319        option -b, which is equivalent to adding /B to every pattern. This causes
2320        it to show the compiled bytecode, without the additional information that
2321        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2322        is the same as /B/I).
2324    23. A new optimization is now able automatically to treat some sequences such
2325        as a*b as a*+b. More specifically, if something simple (such as a character
2326        or a simple class like \d) has an unlimited quantifier, and is followed by
2327        something that cannot possibly match the quantified thing, the quantifier
2328        is automatically "possessified".
2330    24. A recursive reference to a subpattern whose number was greater than 39
2331        went wrong under certain circumstances in UTF-8 mode. This bug could also
2332        have affected the operation of pcre_study().
2334    25. Realized that a little bit of performance could be had by replacing
2335        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2337    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2339    27. Possessive quantifiers such as a++ were previously implemented by turning
2340        them into atomic groups such as ($>a+). Now they have their own opcodes,
2341        which improves performance. This includes the automatically created ones
2342        from 23 above.
2344    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2345        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2346        the first matched character to be a colon. This applied both to named and
2347        numbered groups.
2349    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2351    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2352        building PCRE on Windows easier, so I have included it in the distribution.
2354    31. There is now a check in pcretest against a ridiculously large number being
2355        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2356        loop, the loop is abandoned.
2358    32. Forward references to subpatterns in conditions such as (?(2)...) where
2359        subpattern 2 is defined later cause pcre_compile() to search forwards in
2360        the pattern for the relevant set of parentheses. This search went wrong
2361        when there were unescaped parentheses in a character class, parentheses
2362        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2364    33. "Subroutine" calls and backreferences were previously restricted to
2365        referencing subpatterns earlier in the regex. This restriction has now
2366        been removed.
2368    34. Added a number of extra features that are going to be in Perl 5.10. On the
2369        whole, these are just syntactic alternatives for features that PCRE had
2370        previously implemented using the Python syntax or my own invention. The
2371        other formats are all retained for compatibility.
2373        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2374            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2375            also .NET compatible.
2377        (b) A recursion or subroutine call to a named group can now be defined as
2378            (?&name) as well as (?P>name).
2380        (c) A backreference to a named group can now be defined as \k<name> or
2381            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2382            5.10, are also .NET compatible.
2384        (d) A conditional reference to a named group can now use the syntax
2385            (?(<name>) or (?('name') as well as (?(name).
2387        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2388            groups (named and numbered) that are never evaluated inline, but can be
2389            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2390            is always false. There may be only one alternative in such a group.
2392        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2393            as the simple (?(R). The condition is true only if the most recent
2394            recursion is that of the given number or name. It does not search out
2395            through the entire recursion stack.
2397        (g) The escape \gN or \g{N} has been added, where N is a positive or
2398            negative number, specifying an absolute or relative reference.
2400    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2401        some "unreachable code" warnings.
2403    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2404        things, this adds five new scripts.
2406    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2407        There were also incompatibilities regarding the handling of \Q..\E inside
2408        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2409        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2411    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2412        matches an empty string, and forcibly breaks the loop. There were bugs in
2413        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2414        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2415        separate and independent bugs (that affected different cases) have been
2416        fixed.
2418    39. Refactored the code to abolish the use of different opcodes for small
2419        capturing bracket numbers. This is a tidy that I avoided doing when I
2420        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2421        The new approach is not only tidier, it makes it possible to reduce the
2422        memory needed to fix the previous bug (38).
2424    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2425        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2426        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2427        mode.
2429    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2430        report.
2432    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2433        copy construction and assignment in the C++ wrapper.
2435    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2436        couple of bugs that could have given wrong results in the "--newline=crlf"
2437        case.
2439    44. Added a number of casts and did some reorganization of signed/unsigned int
2440        variables following suggestions from Dair Grant. Also renamed the variable
2441        "this" as "item" because it is a C++ keyword.
2443    45. Arranged for dftables to add
2445          #include "pcre_internal.h"
2447        to pcre_chartables.c because without it, gcc 4.x may remove the array
2448        definition from the final binary if PCRE is built into a static library and
2449        dead code stripping is activated.
2451    46. For an unanchored pattern, if a match attempt fails at the start of a
2452        newline sequence, and the newline setting is CRLF or ANY, and the next two
2453        characters are CRLF, advance by two characters instead of one.
2456    Version 6.7 04-Jul-06
2457    ---------------------
2459     1. In order to handle tests when input lines are enormously long, pcretest has
2460        been re-factored so that it automatically extends its buffers when
2461        necessary. The code is crude, but this _is_ just a test program. The
2462        default size has been increased from 32K to 50K.
2464     2. The code in pcre_study() was using the value of the re argument before
2465        testing it for NULL. (Of course, in any sensible call of the function, it
2466        won't be NULL.)
2468     3. The memmove() emulation function in pcre_internal.h, which is used on
2469        systems that lack both memmove() and bcopy() - that is, hardly ever -
2470        was missing a "static" storage class specifier.
2472     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2473        containing an extended class (one that cannot be represented by a bitmap
2474        because it contains high-valued characters or Unicode property items, e.g.
2475        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2476        pattern, but PCRE should not loop if you do not (it no longer does).
2477        [Detail: two cases were found: (a) a repeated subpattern containing an
2478        extended class; (b) a recursive reference to a subpattern that followed a
2479        previous extended class. It wasn't skipping over the extended class
2480        correctly when UTF-8 mode was not set.]
2482     5. A negated single-character class was not being recognized as fixed-length
2483        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2484        compile error "lookbehind assertion is not fixed length".
2486     6. The RunPerlTest auxiliary script was showing an unexpected difference
2487        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2488        write a Perl script that can interpret lines of an input file either as
2489        byte characters or as UTF-8, which is what "perltest" was being required to
2490        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2491        can't do is switch easily at run time between having the "use utf8;" pragma
2492        or not. In the end, I fudged it by using the RunPerlTest script to insert
2493        "use utf8;" explicitly for the UTF-8 tests.
2495     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2496        the end of the subject string, contrary to the documentation and to what
2497        Perl does. This was true of both matching functions. Now it matches only at
2498        the start of the subject and immediately after *internal* newlines.
2500     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2501        a pointer to an int instead of a pointer to an unsigned long int. This
2502        caused problems on 64-bit systems.
2504     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2505        instance of the 'standard' template library not being so standard".
2507    10. There was no check on the number of named subpatterns nor the maximum
2508        length of a subpattern name. The product of these values is used to compute
2509        the size of the memory block for a compiled pattern. By supplying a very
2510        long subpattern name and a large number of named subpatterns, the size
2511        computation could be caused to overflow. This is now prevented by limiting
2512        the length of names to 32 characters, and the number of named subpatterns
2513        to 10,000.
2515    11. Subpatterns that are repeated with specific counts have to be replicated in
2516        the compiled pattern. The size of memory for this was computed from the
2517        length of the subpattern and the repeat count. The latter is limited to
2518        65535, but there was no limit on the former, meaning that integer overflow
2519        could in principle occur. The compiled length of a repeated subpattern is
2520        now limited to 30,000 bytes in order to prevent this.
2522    12. Added the optional facility to have named substrings with the same name.
2524    13. Added the ability to use a named substring as a condition, using the
2525        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2526        are numbers (not recommended). Forward references are permitted.
2528    14. Added forward references in named backreferences (if you see what I mean).
2530    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2531        pattern could run off the end of the subject. For example, the pattern
2532        "(?s)(.{1,5})"8 did this with the subject "ab".
2534    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2535        PCRE_CASELESS was set when matching characters that were quantified with ?
2536        or *.
2538    17. A character class other than a single negated character that had a minimum
2539        but no maximum quantifier - for example [ab]{6,} - was not handled
2540        correctly by pce_dfa_exec(). It would match only one character.
2542    18. A valid (though odd) pattern that looked like a POSIX character
2543        class but used an invalid character after [ (for example [[,abc,]]) caused
2544        pcre_compile() to give the error "Failed: internal error: code overflow" or
2545        in some cases to crash with a glibc free() error. This could even happen if
2546        the pattern terminated after [[ but there just happened to be a sequence of
2547        letters, a binary zero, and a closing ] in the memory that followed.
2549    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2550        over the years. Originally (before any Unicode support), just the bottom 8
2551        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2552        output from "man perlunicode" includes this:
2554          The regular expression compiler produces polymorphic opcodes.  That
2555          is, the pattern adapts to the data and automatically switches to
2556          the Unicode character scheme when presented with Unicode data--or
2557          instead uses a traditional byte scheme when presented with byte
2558          data.
2560        Sadly, a wide octal escape does not cause a switch, and in a string with
2561        no other multibyte characters, these octal escapes are treated as before.
2562        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2563        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2564        Unicode string.
2566        I have not perpetrated such confusion in PCRE. Up till now, it took just
2567        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2568        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2569        translate to the appropriate multibyte character.
2571    29. Applied some refactoring to reduce the number of warnings from Microsoft
2572        and Borland compilers. This has included removing the fudge introduced
2573        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2574        a warning about an unused variable.
2576    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2577        characters since release 4.0, because Perl (from release 5.004) does not.
2578        [Or at least, is documented not to: some releases seem to be in conflict
2579        with the documentation.] However, when a pattern was studied with
2580        pcre_study() and all its branches started with \s, PCRE still included VT
2581        as a possible starting character. Of course, this did no harm; it just
2582        caused an unnecessary match attempt.
2584    22. Removed a now-redundant internal flag bit that recorded the fact that case
2585        dependency changed within the pattern. This was once needed for "required
2586        byte" processing, but is no longer used. This recovers a now-scarce options
2587        bit. Also moved the least significant internal flag bit to the most-
2588        significant bit of the word, which was not previously used (hangover from
2589        the days when it was an int rather than a uint) to free up another bit for
2590        the future.
2592    23. Added support for CRLF line endings as well as CR and LF. As well as the
2593        default being selectable at build time, it can now be changed at runtime
2594        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2595        specify that it is scanning data with non-default line endings.
2597    24. Changed the definition of CXXLINK to make it agree with the definition of
2598        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2600    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2601        recursions. This makes a big different to stack usage for some patterns.
2603    26. If a subpattern containing a named recursion or subroutine reference such
2604        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2605        the space required for the compiled pattern went wrong and gave too small a
2606        value. Depending on the environment, this could lead to "Failed: internal
2607        error: code overflow at offset 49" or "glibc detected double free or
2608        corruption" errors.
2610    27. Applied patches from Google (a) to support the new newline modes and (b) to
2611        advance over multibyte UTF-8 characters in GlobalReplace.
2613    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2614        difference for some implementation of PCRE in some Windows version.
2616    29. Added some extra testing facilities to pcretest:
2618        \q<number>   in a data line sets the "match limit" value
2619        \Q<number>   in a data line sets the "match recursion limt" value
2620        -S <number>  sets the stack size, where <number> is in megabytes
2622        The -S option isn't available for Windows.
2625    Version 6.6 06-Feb-06
2626    ---------------------
2628     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2629        in pcreposix.h. I have copied the definition from pcre.h.
2631     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2632        because pcre.h is no longer a built file.
2634     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2635        not normally included in the compiled code.
2638    Version 6.5 01-Feb-06
2639    ---------------------
2641     1. When using the partial match feature with pcre_dfa_exec(), it was not
2642        anchoring the second and subsequent partial matches at the new starting
2643        point. This could lead to incorrect results. For example, with the pattern
2644        /1234/, partially matching against "123" and then "a4" gave a match.
2646     2. Changes to pcregrep:
2648        (a) All non-match returns from pcre_exec() were being treated as failures
2649            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2650            error message is output. Some extra information is given for the
2651            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2652            probably the only errors that are likely to be caused by users (by
2653            specifying a regex that has nested indefinite repeats, for instance).
2654            If there are more than 20 of these errors, pcregrep is abandoned.
2656        (b) A binary zero was treated as data while matching, but terminated the
2657            output line if it was written out. This has been fixed: binary zeroes
2658            are now no different to any other data bytes.
2660        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2661            used to set a locale for matching. The --locale=xxxx long option has
2662            been added (no short equivalent) to specify a locale explicitly on the
2663            pcregrep command, overriding the environment variables.
2665        (d) When -B was used with -n, some line numbers in the output were one less
2666            than they should have been.
2668        (e) Added the -o (--only-matching) option.
2670        (f) If -A or -C was used with -c (count only), some lines of context were
2671            accidentally printed for the final match.
2673        (g) Added the -H (--with-filename) option.
2675        (h) The combination of options -rh failed to suppress file names for files
2676            that were found from directory arguments.
2678        (i) Added the -D (--devices) and -d (--directories) options.
2680        (j) Added the -F (--fixed-strings) option.
2682        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2684        (l) Added the --colo(u)r option.
2686        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2687            is not present by default.
2689     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2690        items such as (?R) or (?1), when the recursion could match a number of
2691        alternatives. If it matched one of the alternatives, but subsequently,
2692        outside the recursion, there was a failure, the code tried to back up into
2693        the recursion. However, because of the way PCRE is implemented, this is not
2694        possible, and the result was an incorrect result from the match.
2696        In order to prevent this happening, the specification of recursion has
2697        been changed so that all such subpatterns are automatically treated as
2698        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2700     4. I had overlooked the fact that, in some locales, there are characters for
2701        which isalpha() is true but neither isupper() nor islower() are true. In
2702        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2703        and ordfeminine) are like this. This affected the treatment of \w and \W
2704        when they appeared in character classes, but not when they appeared outside
2705        a character class. The bit map for "word" characters is now created
2706        separately from the results of isalnum() instead of just taking it from the
2707        upper, lower, and digit maps. (Plus the underscore character, of course.)
2709     5. The above bug also affected the handling of POSIX character classes such as
2710        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2711        permanent tables. Instead, the bit maps for such a class were previously
2712        created as the appropriate unions of the upper, lower, and digit bitmaps.
2713        Now they are created by subtraction from the [[:word:]] class, which has
2714        its own bitmap.
2716     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2717        It is created by subtracting the vertical space characters (\x09, \x0a,
2718        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2719        subtraction was done in the overall bitmap for a character class, meaning
2720        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2721        be recognized. This bug has been fixed.
2723     7. Patches from the folks at Google:
2725          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2726          real life, but is still worth protecting against".
2728          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2729          regular expressions".
2731          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2732          have it.
2734          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2735          "configure" and the latter not, in order to fix a problem somebody had
2736          with compiling the Arg class on HP-UX.
2738          (e) Improve the error-handling of the C++ wrapper a little bit.
2740          (f) New tests for checking recursion limiting.
2742     8. The pcre_memmove() function, which is used only if the environment does not
2743        have a standard memmove() function (and is therefore rarely compiled),
2744        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2745        returning a result (though PCRE never actually uses the result).
2747     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2748        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2749        returned instead of calling malloc() with an overflowing number that would
2750        most likely cause subsequent chaos.
2752    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2754    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2755        with this option is matched, the nmatch and pmatch options of regexec() are
2756        ignored.
2758    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2759        provided in case anyone wants to the the POSIX interface with UTF-8
2760        strings.
2762    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2763        C++ linking (needed for some HP-UX environments).
2765    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2766        (unused parameter) and in the pcre_printint() function (omitted "default"
2767        switch label when the default is to do nothing).
2769    15. Added some code to make it possible, when PCRE is compiled as a C++
2770        library, to replace subject pointers for pcre_exec() with a smart pointer
2771        class, thus making it possible to process discontinuous strings.
2773    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2774        much the same function. They were added by different people who were trying
2775        to make PCRE easy to compile on non-Unix systems. It has been suggested
2776        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2777        for compiling on Windows systems. I have therefore replaced it with
2778        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2779        defaults to "extern" for C or "extern C" for C++, which works fine on
2780        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2781        SCOPE with something explicit in config.h. In addition:
2783        (a) pcreposix.h still had just "extern" instead of either of these macros;
2784            I have replaced it with PCRE_DATA_SCOPE.
2786        (b) Functions such as _pcre_xclass(), which are internal to the library,
2787            but external in the C sense, all had PCRE_EXPORT in their definitions.
2788            This is apparently wrong for the Windows case, so I have removed it.
2789            (It makes no difference on Unix-like systems.)
2791    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2792        of recursive calls to match(). This is different to MATCH_LIMIT because
2793        that limits the total number of calls to match(), not all of which increase
2794        the depth of recursion. Limiting the recursion depth limits the amount of
2795        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2796        when PCRE is compiled, and changed at run time. A patch from Google adds
2797        this functionality to the C++ interface.
2799    18. Changes to the handling of Unicode character properties:
2801        (a) Updated the table to Unicode 4.1.0.
2803        (b) Recognize characters that are not in the table as "Cn" (undefined).
2805        (c) I revised the way the table is implemented to a much improved format
2806            which includes recognition of ranges. It now supports the ranges that
2807            are defined in UnicodeData.txt, and it also amalgamates other
2808            characters into ranges. This has reduced the number of entries in the
2809            table from around 16,000 to around 3,000, thus reducing its size
2810            considerably. I realized I did not need to use a tree structure after
2811            all - a binary chop search is just as efficient. Having reduced the
2812            number of entries, I extended their size from 6 bytes to 8 bytes to
2813            allow for more data.
2815        (d) Added support for Unicode script names via properties such as \p{Han}.
2817    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2818        matching that character.
2820    20. When matching a repeated Unicode property with a minimum greater than zero,
2821        (for example \pL{2,}), PCRE could look past the end of the subject if it
2822        reached it while seeking the minimum number of characters. This could
2823        happen only if some of the characters were more than one byte long, because
2824        there is a check for at least the minimum number of bytes.
2826    21. Refactored the implementation of \p and \P so as to be more general, to
2827        allow for more different types of property in future. This has changed the
2828        compiled form incompatibly. Anybody with saved compiled patterns that use
2829        \p or \P will have to recompile them.
2831    22. Added "Any" and "L&" to the supported property types.
2833    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2834        but give a compile time error if the value is greater than 0xff.
2836    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2837        accidentally not being installed or uninstalled.
2839    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2840        made were to insert the current release number. This seemed silly, because
2841        it made things harder for people building PCRE on systems that don't run
2842        "configure". I have turned pcre.h into a distributed file, no longer built
2843        by "configure", with the version identification directly included. There is
2844        no longer a pcre.h.in file.
2846        However, this change necessitated a change to the pcre-config script as
2847        well. It is built from pcre-config.in, and one of the substitutions was the
2848        release number. I have updated configure.ac so that ./configure now finds
2849        the release number by grepping pcre.h.
2851    26. Added the ability to run the tests under valgrind.
2854    Version 6.4 05-Sep-05
2855    ---------------------
2857     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2858        "--" to be printed when multiple files were scanned, even when none of the
2859        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2860        consider it to be a bug, and have restored the previous behaviour.
2862     2. A couple of code tidies to get rid of compiler warnings.
2864     3. The pcretest program used to cheat by referring to symbols in the library
2865        whose names begin with _pcre_. These are internal symbols that are not
2866        really supposed to be visible externally, and in some environments it is
2867        possible to suppress them. The cheating is now confined to including
2868        certain files from the library's source, which is a bit cleaner.
2870     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2871        file's purpose clearer.
2873     5. Reorganized pcre_ucp_findchar().
2876    Version 6.3 15-Aug-05
2877    ---------------------
2879     1. The file libpcre.pc.in did not have general read permission in the tarball.
2881     2. There were some problems when building without C++ support:
2883        (a) If C++ support was not built, "make install" and "make test" still
2884            tried to test it.
2886        (b) There were problems when the value of CXX was explicitly set. Some
2887            changes have been made to try to fix these, and ...
2889        (c) --disable-cpp can now be used to explicitly disable C++ support.
2891        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2892            backslash in a target when C++ was disabled. This confuses some
2893            versions of "make", apparently. Using an intermediate variable solves
2894            this. (Same for CPP_LOBJ.)
2896     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2897        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2898        necessary on certain architectures.
2900     4. Added a setting of -export-symbols-regex to the link command to remove
2901        those symbols that are exported in the C sense, but actually are local
2902        within the library, and not documented. Their names all begin with
2903        "_pcre_". This is not a perfect job, because (a) we have to except some
2904        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2905        available (and never for static libraries). I have made a note to try to
2906        find a way round (a) in the future.
2909    Version 6.2 01-Aug-05
2910    ---------------------
2912     1. There was no test for integer overflow of quantifier values. A construction
2913        such as {1111111111111111} would give undefined results. What is worse, if
2914        a minimum quantifier for a parenthesized subpattern overflowed and became
2915        negative, the calculation of the memory size went wrong. This could have
2916        led to memory overwriting.
2918     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2920     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2921        operating environments where this matters.
2923     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2924        PCRE options from within the C++ wrapper.
2926     5. Named capturing subpatterns were not being correctly counted when a pattern
2927        was compiled. This caused two problems: (a) If there were more than 100
2928        such subpatterns, the calculation of the memory needed for the whole
2929        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2930        back references of the form \12, where the number was greater than 9, were
2931        not recognized as back references, even though there were sufficient
2932        previous subpatterns.
2934     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2935        versions of gcc, e.g. 2.95.4.
2938    Version 6.1 21-Jun-05
2939    ---------------------
2941     1. There was one reference to the variable "posix" in pcretest.c that was not
2942        surrounded by "#if !defined NOPOSIX".
2944     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2945        the cross-check on the old pcre_info() function, for the benefit of the
2946        cut-down version of PCRE that is currently imported into Exim.
2948     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2949        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2950        patterns that start (?i) but I don't think that matters. The use of (?i) is
2951        just an example; this all applies to the other options as well.
2953     4. Since libtool seems to echo the compile commands it is issuing, the output
2954        from "make" can be reduced a bit by putting "@" in front of each libtool
2955        compile command.
2957     5. Patch from the folks at Google for configure.in to be a bit more thorough
2958        in checking for a suitable C++ installation before trying to compile the
2959        C++ stuff. This should fix a reported problem when a compiler was present,
2960        but no suitable headers.
2962     6. The man pages all had just "PCRE" as their title. I have changed them to
2963        be the relevant file name. I have also arranged that these names are
2964        retained in the file doc/pcre.txt, which is a concatenation in text format
2965        of all the man pages except the little individual ones for each function.
2967     7. The NON-UNIX-USE file had not been updated for the different set of source
2968        files that come with release 6. I also added a few comments about the C++
2969        wrapper.
2972    Version 6.0 07-Jun-05
2973    ---------------------
2975     1. Some minor internal re-organization to help with my DFA experiments.
2977     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2978        didn't matter for the library itself when fully configured, but did matter
2979        when compiling without UCP support, or within Exim, where the ucp files are
2980        not imported.
2982     3. Refactoring of the library code to split up the various functions into
2983        different source modules. The addition of the new DFA matching code (see
2984        below) to a single monolithic source would have made it really too
2985        unwieldy, quite apart from causing all the code to be include in a
2986        statically linked application, when only some functions are used. This is
2987        relevant even without the DFA addition now that patterns can be compiled in
2988        one application and matched in another.
2990        The downside of splitting up is that there have to be some external
2991        functions and data tables that are used internally in different modules of
2992        the library but which are not part of the API. These have all had their
2993        names changed to start with "_pcre_" so that they are unlikely to clash
2994        with other external names.
2996     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2997        a different (DFA) algorithm. Although it is slower than the original
2998        function, it does have some advantages for certain types of matching
2999        problem.
3001     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3002        including restarting after a partial match.
3004     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3005        defined when compiling for Windows was sent to me. I have put it into the
3006        code, though I have no means of testing or verifying it.
3008     7. Added the pcre_refcount() auxiliary function.
3010     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3011        match before or at the first newline in the subject string. In pcretest,
3012        the /f option on a pattern can be used to set this.
3014     9. A repeated \w when used in UTF-8 mode with characters greater than 256
3015        would behave wrongly. This has been present in PCRE since release 4.0.
3017    10. A number of changes to the pcregrep command:
3019        (a) Refactored how -x works; insert ^(...)$ instead of setting
3020            PCRE_ANCHORED and checking the length, in preparation for adding
3021            something similar for -w.
3023        (b) Added the -w (match as a word) option.
3025        (c) Refactored the way lines are read and buffered so as to have more
3026            than one at a time available.
3028        (d) Implemented a pcregrep test script.
3030        (e) Added the -M (multiline match) option. This allows patterns to match
3031            over several lines of the subject. The buffering ensures that at least
3032            8K, or the rest of the document (whichever is the shorter) is available
3033            for matching (and similarly the previous 8K for lookbehind assertions).
3035        (f) Changed the --help output so that it now says
3037              -w, --word-regex(p)
3039            instead of two lines, one with "regex" and the other with "regexp"
3040            because that confused at least one person since the short forms are the
3041            same. (This required a bit of code, as the output is generated
3042            automatically from a table. It wasn't just a text change.)
3044        (g) -- can be used to terminate pcregrep options if the next thing isn't an
3045            option but starts with a hyphen. Could be a pattern or a path name
3046            starting with a hyphen, for instance.
3048        (h) "-" can be given as a file name to represent stdin.
3050        (i) When file names are being printed, "(standard input)" is used for
3051            the standard input, for compatibility with GNU grep. Previously
3052            "<stdin>" was used.
3054        (j) The option --label=xxx can be used to supply a name to be used for
3055            stdin when file names are being printed. There is no short form.
3057        (k) Re-factored the options decoding logic because we are going to add
3058            two more options that take data. Such options can now be given in four
3059            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3061        (l) Added the -A, -B, and -C options for requesting that lines of context
3062            around matches be printed.
3064        (m) Added the -L option to print the names of files that do not contain
3065            any matching lines, that is, the complement of -l.
3067        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3068            continue to scan other files.
3070        (o) The -s option was incorrectly implemented. For compatibility with other
3071            greps, it now suppresses the error message for a non-existent or non-
3072            accessible file (but not the return code). There is a new option called
3073            -q that suppresses the output of matching lines, which was what -s was
3074            previously doing.
3076        (p) Added --include and --exclude options to specify files for inclusion
3077            and exclusion when recursing.
3079    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3080        Hopefully, it now does.
3082    12. Missing cast in pcre_study().
3084    13. Added an "uninstall" target to the makefile.
3086    14. Replaced "extern" in the function prototypes in Makefile.in with
3087        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3088        world, but is set differently for Windows.
3090    15. Added a second compiling function called pcre_compile2(). The only
3091        difference is that it has an extra argument, which is a pointer to an
3092        integer error code. When there is a compile-time failure, this is set
3093        non-zero, in addition to the error test pointer being set to point to an
3094        error message. The new argument may be NULL if no error number is required
3095        (but then you may as well call pcre_compile(), which is now just a
3096        wrapper). This facility is provided because some applications need a
3097        numeric error indication, but it has also enabled me to tidy up the way
3098        compile-time errors are handled in the POSIX wrapper.
3100    16. Added VPATH=.libs to the makefile; this should help when building with one
3101        prefix path and installing with another. (Or so I'm told by someone who
3102        knows more about this stuff than I do.)
3104    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3105        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3106        match everything, including newlines. This is not POSIX-compatible, but
3107        somebody wanted the feature. From pcretest it can be activated by using
3108        both the P and the s flags.
3110    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3112    19. libpcre.pc was being incorrectly installed as executable.
3114    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3115        it now also looks for '\r' so that it will work unmodified on Windows.
3117    21. Added Google's contributed C++ wrapper to the distribution.
3119    22. Added some untidy missing memory free() calls in pcretest, to keep
3120        Electric Fence happy when testing.
3124    Version 5.0 13-Sep-04
3125    ---------------------
3127     1. Internal change: literal characters are no longer packed up into items
3128        containing multiple characters in a single byte-string. Each character
3129        is now matched using a separate opcode. However, there may be more than one
3130        byte in the character in UTF-8 mode.
3132     2. The pcre_callout_block structure has two new fields: pattern_position and
3133        next_item_length. These contain the offset in the pattern to the next match
3134        item, and its length, respectively.
3136     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3137        insertion of callouts before each pattern item. Added the /C option to
3138        pcretest to make use of this.
3140     4. On the advice of a Windows user, the lines
3142          #if defined(_WIN32) || defined(WIN32)
3143          _setmode( _fileno( stdout ), 0x8000 );
3144          #endif  /* defined(_WIN32) || defined(WIN32) */
3146        have been added to the source of pcretest. This apparently does useful
3147        magic in relation to line terminators.
3149     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3150        for the benefit of those environments where the "b" makes a difference.
3152     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3153        to know about it. I have put a hack into configure.in that adds in code
3154        to set GCC=yes if CC=icc. This seems to end up at a point in the
3155        generated configure script that is early enough to affect the setting of
3156        compiler options, which is what is needed, but I have no means of testing
3157        whether it really works. (The user who reported this had patched the
3158        generated configure script, which of course I cannot do.)
3160        LATER: After change 22 below (new libtool files), the configure script
3161        seems to know about icc (and also ecc). Therefore, I have commented out
3162        this hack in configure.in.
3164     7. Added support for pkg-config (2 patches were sent in).
3166     8. Negated POSIX character classes that used a combination of internal tables
3167        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3168        [[:^ascii]]. Typically, they would match almost any characters. The other
3169        POSIX classes were not broken in this way.
3171     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3172        to find the match, as PCRE was deluded into thinking that the match had to
3173        start at the start point or following a newline. The same bug applied to
3174        patterns with negative forward assertions or any backward assertions
3175        preceding ".*" at the start, unless the pattern required a fixed first
3176        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3178    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3179        starting at the last subject character, bytes beyond the end of the subject
3180        string were read.
3182    11. Renamed the variable "class" as "classbits" to make life easier for C++
3183        users. (Previously there was a macro definition, but it apparently wasn't
3184        enough.)
3186    12. Added the new field "tables" to the extra data so that tables can be passed
3187        in at exec time, or the internal tables can be re-selected. This allows
3188        a compiled regex to be saved and re-used at a later time by a different
3189        program that might have everything at different addresses.
3191    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3192        -R library as well as a -L library.
3194    14. The debugging options of pcretest (-d on the command line or D on a
3195        pattern) showed incorrect output for anything following an extended class
3196        that contained multibyte characters and which was followed by a quantifier.
3198    15. Added optional support for general category Unicode character properties
3199        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3200        support. It adds about 90K to the size of the library. The meanings of the
3201        inbuilt class escapes such as \d and \s have NOT been changed.
3203    16. Updated pcredemo.c to include calls to free() to release the memory for the
3204        compiled pattern.
3206    17. The generated file chartables.c was being created in the source directory
3207        instead of in the building directory. This caused the build to fail if the
3208        source directory was different from the building directory, and was
3209        read-only.
3211    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3212        file. No doubt somebody will tell me if they don't make sense... Also added
3213        Dan Mooney's comments about building on OpenVMS.
3215    19. Added support for partial matching via the PCRE_PARTIAL option for
3216        pcre_exec() and the \P data escape in pcretest.
3218    20. Extended pcretest with 3 new pattern features:
3220        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3221              write the compiled pattern to the file whose name is "rest-of-line".
3222              This is a straight binary dump of the data, with the saved pointer to
3223              the character tables forced to be NULL. The study data, if any, is
3224              written too. After writing, pcretest reads a new pattern.
3226        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3227              compiled pattern from the given file. There must not be any
3228              occurrences of "<" in the file name (pretty unlikely); if there are,
3229              pcretest will instead treat the initial "<" as a pattern delimiter.
3230              After reading in the pattern, pcretest goes on to read data lines as
3231              usual.
3233        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3234              and 16-bit fields in a compiled pattern, to simulate a pattern that
3235              was compiled on a host of opposite endianness.
3237    21. The pcre-exec() function can now cope with patterns that were compiled on
3238        hosts of opposite endianness, with this restriction:
3240          As for any compiled expression that is saved and used later, the tables
3241          pointer field cannot be preserved; the extra_data field in the arguments
3242          to pcre_exec() should be used to pass in a tables address if a value
3243          other than the default internal tables were used at compile time.
3245    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3246        now diagnosed as an error. Previously, most of the time, a negative number
3247        would have been treated as zero, but if in addition "ovector" was passed as
3248        NULL, a crash could occur.
3250    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3251        new versions from the libtool 1.5 distribution (the last one is a copy of
3252        a file called libtool.m4). This seems to have fixed the need to patch
3253        "configure" to support Darwin 1.3 (which I used to do). However, I still
3254        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3255        workstation).
3257    24. Changed the PCRE licence to be the more standard "BSD" licence.
3260    Version 4.5 01-Dec-03
3261    ---------------------
3263     1. There has been some re-arrangement of the code for the match() function so
3264        that it can be compiled in a version that does not call itself recursively.
3265        Instead, it keeps those local variables that need separate instances for
3266        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3267        needs to "recurse". Keeping track of where control must go is done by means
3268        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3269        hide most of the details from the main code, and operates only if
3270        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3271        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3272        operating.
3274        To make it easier for callers to provide specially tailored get/free
3275        functions for this usage, two new functions, pcre_stack_malloc, and
3276        pcre_stack_free, are used. They are always called in strict stacking order,
3277        and the size of block requested is always the same.
3279        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3280        PCRE has been compiled to use the stack or the heap for recursion. The
3281        -C option of pcretest uses this to show which version is compiled.
3283        A new data escape \S, is added to pcretest; it causes the amounts of store
3284        obtained and freed by both kinds of malloc/free at match time to be added
3285        to the output.
3287     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3288        what's available on my current Linux desktop machine.
3290     3. When matching a UTF-8 string, the test for a valid string at the start has
3291        been extended. If start_offset is not zero, PCRE now checks that it points
3292        to a byte that is the start of a UTF-8 character. If not, it returns
3293        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3294        this is necessary because there may be backward assertions in the pattern.
3295        When matching the same subject several times, it may save resources to use
3296        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3298     4. The code for checking the validity of UTF-8 strings has been tightened so
3299        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3300        containing "overlong sequences".
3302     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3303        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3304        should have been "&", but it just so happened that all the cases this let
3305        through by mistake were picked up later in the function.
3307     6. I had used a variable called "isblank" - this is a C99 function, causing
3308        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3310     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3311        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3312        so the version is shown; (c) stop "make test" echoing "./RunTest".
3314     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3316     9. The prototype for memmove() for systems that don't have it was using
3317        size_t, but the inclusion of the header that defines size_t was later. I've
3318        moved the #includes for the C headers earlier to avoid this.
3320    10. Added some adjustments to the code to make it easier to compiler on certain
3321        special systems:
3323          (a) Some "const" qualifiers were missing.
3324          (b) Added the macro EXPORT before all exported functions; by default this
3325              is defined to be empty.
3326          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3327              that it reads its output file name as an argument instead of writing
3328              to the standard output and assuming this can be redirected.
3330    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3331        class containing characters with values greater than 255, PCRE compilation
3332        went into a loop.
3334    12. A recursive reference to a subpattern that was within another subpattern
3335        that had a minimum quantifier of zero caused PCRE to crash. For example,
3336        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3337        recursion. If the recursively-called subpattern itself had a zero repeat,
3338        that was OK.
3340    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3341        buffer into which it was copied (for escape processing) was still set at
3342        1024, so long lines caused crashes.
3344    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3345        "internal error: code overflow...". This applied to any character class
3346        that was followed by a possessive quantifier.
3348    15. Modified the Makefile to add libpcre.la as a prerequisite for
3349        libpcreposix.la because I was told this is needed for a parallel build to
3350        work.
3352    16. If a pattern that contained .* following optional items at the start was
3353        studied, the wrong optimizing data was generated, leading to matching
3354        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3355        matching string must start with a or b or c. The correct conclusion for
3356        this pattern is that a match can start with any character.
3359    Version 4.4 13-Aug-03
3360    ---------------------
3362     1. In UTF-8 mode, a character class containing characters with values between
3363        127 and 255 was not handled correctly if the compiled pattern was studied.
3364        In fixing this, I have also improved the studying algorithm for such
3365        classes (slightly).
3367     2. Three internal functions had redundant arguments passed to them. Removal
3368        might give a very teeny performance improvement.
3370     3. Documentation bug: the value of the capture_top field in a callout is *one
3371        more than* the number of the hightest numbered captured substring.
3373     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3374        in incorrectly linking with a previously installed version. They now link
3375        explicitly with libpcre.la.
3377     5. configure.in no longer needs to recognize Cygwin specially.
3379     6. A problem in pcre.in for Windows platforms is fixed.
3381     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3382        pcretest, it used to include the size of the study block as part of its
3383        output. Unfortunately, the structure contains a field that has a different
3384        size on different hardware architectures. This meant that the tests that
3385        showed this size failed. As the block is currently always of a fixed size,
3386        this information isn't actually particularly useful in pcretest output, so
3387        I have just removed it.
3389     8. Three pre-processor statements accidentally did not start in column 1.
3390        Sadly, there are *still* compilers around that complain, even though
3391        standard C has not required this for well over a decade. Sigh.
3393     9. In pcretest, the code for checking callouts passed small integers in the
3394        callout_data field, which is a void * field. However, some picky compilers
3395        complained about the casts involved for this on 64-bit systems. Now
3396        pcretest passes the address of the small integer instead, which should get
3397        rid of the warnings.
3399    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3400        both compile and run time, and gives an error if an invalid UTF-8 sequence
3401        is found. There is a option for disabling this check in cases where the
3402        string is known to be correct and/or the maximum performance is wanted.
3404    11. In response to a bug report, I changed one line in Makefile.in from
3406            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3407        to
3408            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3410        to look similar to other lines, but I have no way of telling whether this
3411        is the right thing to do, as I do not use Windows. No doubt I'll get told
3412        if it's wrong...
3415    Version 4.3 21-May-03
3416    ---------------------
3418    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3419       Makefile.
3421    2. Some refactoring to improve the quality of the code:
3423       (i)   The utf8_table... variables are now declared "const".
3425       (ii)  The code for \cx, which used the "case flipping" table to upper case
3426             lower case letters, now just substracts 32. This is ASCII-specific,
3427             but the whole concept of \cx is ASCII-specific, so it seems
3428             reasonable.
3430       (iii) PCRE was using its character types table to recognize decimal and
3431             hexadecimal digits in the pattern. This is silly, because it handles
3432             only 0-9, a-f, and A-F, but the character types table is locale-
3433             specific, which means strange things might happen. A private
3434             table is now used for this - though it costs 256 bytes, a table is
3435             much faster than multiple explicit tests. Of course, the standard
3436             character types table is still used for matching digits in subject
3437             strings against \d.
3439       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3440             ending in _t are). So I've renamed it as ESC_tee.
3442    3. The first argument for regexec() in the POSIX wrapper should have been
3443       defined as "const".
3445    4. Changed pcretest to use malloc() for its buffers so that they can be
3446       Electric Fenced for debugging.
3448    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3449       to read one or more bytes before the start of the subject string. Often this
3450       had no effect on PCRE's behaviour, but in some circumstances it could
3451       provoke a segmentation fault.
3453    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3454       to try to read one or more bytes before the start of the subject string.
3456    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3457       UTF-8 support could misbehave in various ways if the subject string
3458       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3459       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3460       back over UTF-8 characters.)
3463    Version 4.2 14-Apr-03
3464    ---------------------
3466    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3468    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3469         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3470         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3471         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3472         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3473           and BUILD_EXEEXT
3474         Note: automatic setting of the BUILD variables is not yet working
3475         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3476           compile-time but not at link-time
3477         [LINK]: use for linking executables only
3478         make different versions for Windows and non-Windows
3479         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3480           libraries
3481         [LINK_FOR_BUILD]: new variable
3482         [OBJEXT]: use throughout
3483         [EXEEXT]: use throughout
3484         <winshared>: new target
3485         <wininstall>: new target
3486         <dftables.o>: use native compiler
3487         <dftables>: use native linker
3488         <install>: handle Windows platform correctly
3489         <clean>: ditto
3490         <check>: ditto
3491         copy DLL to top builddir before testing
3493       As part of these changes, -no-undefined was removed again. This was reported
3494       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3495       in any case.
3497    3. Some tidies to get rid of compiler warnings:
3499       . In the match_data structure, match_limit was an unsigned long int, whereas
3500         match_call_count was an int. I've made them both unsigned long ints.
3502       . In pcretest the fact that a const uschar * doesn't automatically cast to
3503         a void * provoked a warning.
3505       . Turning on some more compiler warnings threw up some "shadow" variables
3506         and a few more missing casts.
3508    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3509       option, a class that contained a single character with a value between 128
3510       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3512    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3513       option, a class that contained several characters, but with at least one
3514       whose value was between 128 and 255 caused PCRE to crash.
3517    Version 4.1 12-Mar-03
3518    ---------------------
3520    1. Compiling with gcc -pedantic found a couple of places where casts were
3521    needed, and a string in dftables.c that was longer than standard compilers are
3522    required to support.
3524    2. Compiling with Sun's compiler found a few more places where the code could
3525    be tidied up in order to avoid warnings.
3527    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3528    first of these names is deprecated in the latest Autoconf in favour of the name
3529    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3530    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3531    analogy I have changed it to CFLAGS_FOR_BUILD.
3533    4. Added -no-undefined to the linking command in the Makefile, because this is
3534    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3535    linking step for the pcreposix library.
3537    5. PCRE was failing to diagnose the case of two named groups with the same
3538    name.
3540    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3541    literal character that is needed in the subject for a match, and scans along to
3542    ensure that it is present before embarking on the full matching process. This
3543    saves time in cases of nested unlimited repeats that are never going to match.
3544    Problem: the scan can take a lot of time if the subject is very long (e.g.
3545    megabytes), thus penalizing straightforward matches. It is now done only if the
3546    amount of subject to be scanned is less than 1000 bytes.
3548    7. A lesser problem with the same optimization is that it was recording the
3549    first character of an anchored pattern as "needed", thus provoking a search
3550    right along the subject, even when the first match of the pattern was going to
3551    fail. The "needed" character is now not set for anchored patterns, unless it
3552    follows something in the pattern that is of non-fixed length. Thus, it still
3553    fulfils its original purpose of finding quick non-matches in cases of nested
3554    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3557    Version 4.0 17-Feb-03
3558    ---------------------
3560    1. If a comment in an extended regex that started immediately after a meta-item
3561    extended to the end of string, PCRE compiled incorrect data. This could lead to
3562    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3564    2. Moved to autoconf 2.53 and libtool 1.4.2.
3566    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3567    the special perltest8 script is no longer needed - all the tests can be run
3568    from a single perltest script.
3570    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3571    by \s. It has now been removed in PCRE. This means it isn't recognized as
3572    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3573    class [:space:] *does* include VT, thereby creating a mess.
3575    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3576    space and tab.
3578    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3579    its new features into the main test script, reducing the number of scripts.
3581    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3582    were backward compatible, and made the (?i) apply to the whole pattern, as if
3583    /i were given. Now it behaves more logically, and applies the option setting
3584    only to what follows. PCRE has been changed to follow suit. However, if it
3585    finds options settings right at the start of the pattern, it extracts them into
3586    the global options, as before. Thus, they show up in the info data.
3588    8. Added support for the \Q...\E escape sequence. Characters in between are
3589    treated as literals. This is slightly different from Perl in that $ and @ are
3590    also handled as literals inside the quotes. In Perl, they will cause variable
3591    interpolation. Note the following examples:
3593        Pattern            PCRE matches      Perl matches
3595        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3596        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3597        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3599    For compatibility with Perl, \Q...\E sequences are recognized inside character
3600    classes as well as outside them.
3602    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3603    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3604    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3605    signed/unsigned warnings.
3607    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3608    option for pcretest, so I've replaced it by a simple function that does just
3609    that job.
3611    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3612    "pcregrep -".
3614    12. Added &quo