ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 61 by nigel, Sat Feb 24 21:39:58 2007 UTC revision 933 by ph10, Sat Feb 25 12:18:23 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4  Version 3.0 02-Jan-02  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.
59    Version 8.30 04-February-2012
60    -----------------------------
62    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
63        name is defined in ctype.h.
65    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
66        only in quite long subpatterns.
68    3.  Removed the function pcre_info(), which has been obsolete and deprecated
69        since it was replaced by pcre_fullinfo() in February 2000.
71    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
72        match a (*MARK), and the match failed at the start of the subject, a
73        reference to memory before the start of the subject could occur. This bug
74        was introduced by fix 17 of release 8.21.
76    5.  A reference to an unset group with zero minimum repetition was giving
77        totally wrong answers (in non-JavaScript-compatibility mode). For example,
78        /(another)?(\1?)test/ matched against "hello world test". This bug was
79        introduced in release 8.13.
81    6.  Add support for 16-bit character strings (a large amount of work involving
82        many changes and refactorings).
84    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
85        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
86        from a file.
88    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
89        rounding is not applied in this particular case).
91    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
92        if they appear, or are escaped, in patterns.
94    10. Get rid of a number of -Wunused-but-set-variable warnings.
96    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
97        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
98        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
99        also returns the mark "x". This bug applied to capturing parentheses,
100        non-capturing parentheses, and atomic parentheses. It also applied to some
101        assertions.
103    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
104        information out of configure.ac instead of relying on pcre.h.generic, which
105        is not stored in the repository.
107    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
108        -lreadline.
110    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
112    15. Applied Graycode's patch to put the top-level frame on the stack rather
113        than the heap when not using the stack for recursion. This gives a
114        performance improvement in many cases when recursion is not deep.
116    16. Experimental code added to "pcretest -C" to output the stack frame size.
119    Version 8.21 12-Dec-2011
120    ------------------------
122    1.  Updating the JIT compiler.
124    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
125        are added as well.
127    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
128        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
129        calling _pcre_jit_exec. Some extra comments are added.
131    4.  (*MARK) settings inside atomic groups that do not contain any capturing
132        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
133        was introduced by change 18 for 8.20.
135    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
136        ECMA-262 standard.
138    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
139        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
140        This bug was probably introduced by change 9 of 8.13.
142    7.  While fixing 6 above, I noticed that a number of other items were being
143        incorrectly rejected as "not fixed length". This arose partly because newer
144        opcodes had not been added to the fixed-length checking code. I have (a)
145        corrected the bug and added tests for these items, and (b) arranged for an
146        error to occur if an unknown opcode is encountered while checking for fixed
147        length instead of just assuming "not fixed length". The items that were
148        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
149        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
150        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
152    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
153        being incorrectly compiled and would have given unpredicatble results.
155    9.  A possessively repeated subpattern with minimum repeat count greater than
156        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
157        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
158        the first (A) could occur when it should not.
160    10. Add a cast and remove a redundant test from the code.
162    11. JIT should use pcre_malloc/pcre_free for allocation.
164    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
165        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
166        is anything other than /usr, -L is still shown).
168    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
170    14. Perl does not support \N without a following name in a [] class; PCRE now
171        also gives an error.
173    15. If a forward reference was repeated with an upper limit of around 2000,
174        it caused the error "internal error: overran compiling workspace". The
175        maximum number of forward references (including repeats) was limited by the
176        internal workspace, and dependent on the LINK_SIZE. The code has been
177        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
178        the default depends on LINK_SIZE. There is a new upper limit (for safety)
179        of around 200,000 forward references. While doing this, I also speeded up
180        the filling in of repeated forward references.
182    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
183        incorrectly expecting the subject to contain another "a" after the start.
185    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
186        in the match, the SKIP should be ignored. This was not happening; instead
187        the SKIP was being treated as NOMATCH. For patterns such as
188        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
189        tested.
191    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
192        now much more compatible with Perl, in particular in cases where the result
193        is a non-match for a non-anchored pattern. For example, if
194        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
195        "m", where previously it did not return a name. A side effect of this
196        change is that for partial matches, the last encountered mark name is
197        returned, as for non matches. A number of tests that were previously not
198        Perl-compatible have been moved into the Perl-compatible test files. The
199        refactoring has had the pleasing side effect of removing one argument from
200        the match() function, thus reducing its stack requirements.
202    19. If the /S+ option was used in pcretest to study a pattern using JIT,
203        subsequent uses of /S (without +) incorrectly behaved like /S+.
205    21. Retrieve executable code size support for the JIT compiler and fixing
206        some warnings.
208    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
209        not work when the shorter character appeared right at the end of the
210        subject string.
212    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
213        systems.
215    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
216        output it when the /M option is used in pcretest.
218    25. The CheckMan script was not being included in the distribution. Also, added
219        an explicit "perl" to run Perl scripts from the PrepareRelease script
220        because this is reportedly needed in Windows.
222    26. If study data was being save in a file and studying had not found a set of
223        "starts with" bytes for the pattern, the data written to the file (though
224        never used) was taken from uninitialized memory and so caused valgrind to
225        complain.
227    27. Updated RunTest.bat as provided by Sheri Pierce.
229    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
231    29. Computation of memory usage for the table of capturing group names was
232        giving an unnecessarily large value.
235    Version 8.20 21-Oct-2011
236    ------------------------
238    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
239        a POSIX class. After further experiments with Perl, which convinced me that
240        Perl has bugs and confusions, a closing square bracket is no longer allowed
241        in a POSIX name. This bug also affected patterns with classes that started
242        with full stops.
244    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
245        captured substring, but while checking the failing first alternative,
246        substring 1 is temporarily captured. If the output vector supplied to
247        pcre_exec() was not big enough for this capture, the yield of the function
248        was still zero ("insufficient space for captured substrings"). This cannot
249        be totally fixed without adding another stack variable, which seems a lot
250        of expense for a edge case. However, I have improved the situation in cases
251        such as /(a)(b)x|abc/ matched against "abc", where the return code
252        indicates that fewer than the maximum number of slots in the ovector have
253        been set.
255    3.  Related to (2) above: when there are more back references in a pattern than
256        slots in the output vector, pcre_exec() uses temporary memory during
257        matching, and copies in the captures as far as possible afterwards. It was
258        using the entire output vector, but this conflicts with the specification
259        that only 2/3 is used for passing back captured substrings. Now it uses
260        only the first 2/3, for compatibility. This is, of course, another edge
261        case.
263    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
264        main code base, and can be used by building with --enable-jit. When this is
265        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
266        runtime --no-jit option is given.
268    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
269        ovector, the return from the function was zero, implying that there were
270        other matches that did not fit. The correct "exactly full" value is now
271        returned.
273    6.  If a subpattern that was called recursively or as a subroutine contained
274        (*PRUNE) or any other control that caused it to give a non-standard return,
275        invalid errors such as "Error -26 (nested recursion at the same subject
276        position)" or even infinite loops could occur.
278    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
279        computing the minimum length on reaching *ACCEPT, and so ended up with the
280        wrong value of 1 rather than 0. Further investigation indicates that
281        computing a minimum subject length in the presence of *ACCEPT is difficult
282        (think back references, subroutine calls), and so I have changed the code
283        so that no minimum is registered for a pattern that contains *ACCEPT.
285    8.  If (*THEN) was present in the first (true) branch of a conditional group,
286        it was not handled as intended. [But see 16 below.]
288    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
289        Sheri Pierce.
291    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
292        the first byte in a match must be "a".
294    11. Change 17 for 8.13 increased the recursion depth for patterns like
295        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
296        pattern contains any instances of (*THEN). If it does not, the old
297        optimizations are restored. It would be nice to do this on a per-group
298        basis, but at the moment that is not feasible.
300    12. In some environments, the output of pcretest -C is CRLF terminated. This
301        broke RunTest's code that checks for the link size. A single white space
302        character after the value is now allowed for.
304    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
305        For "fr", it uses the Windows-specific input and output files.
307    14. If (*THEN) appeared in a group that was called recursively or as a
308        subroutine, it did not work as intended. [But see next item.]
310    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
311        pattern fragments (but not containing any | characters). If A and B are
312        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
313        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
314        D. In other words, Perl considers parentheses that do not contain any |
315        characters to be part of a surrounding alternative, whereas PCRE was
316        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
317        differently. PCRE now behaves in the same way as Perl, except in the case
318        of subroutine/recursion calls such as (?1) which have in any case always
319        been different (but PCRE had them first :-).
321    16. Related to 15 above: Perl does not treat the | in a conditional group as
322        creating alternatives. Such a group is treated in the same way as an
323        ordinary group without any | characters when processing (*THEN). PCRE has
324        been changed to match Perl's behaviour.
326    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
327        RunGrepTest script failed.
329    18. Change 22 for version 13 caused atomic groups to use more stack. This is
330        inevitable for groups that contain captures, but it can lead to a lot of
331        stack use in large patterns. The old behaviour has been restored for atomic
332        groups that do not contain any capturing parentheses.
334    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
335        suppress the check for a minimum subject length at run time. (If it was
336        given to pcre_exec() or pcre_dfa_exec() it did work.)
338    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
339        fail to work when decoding hex characters in data strings in EBCDIC
340        environments.
342    21. It appears that in at least one Mac OS environment, the isxdigit() function
343        is implemented as a macro that evaluates to its argument more than once,
344        contravening the C 90 Standard (I haven't checked a later standard). There
345        was an instance in pcretest which caused it to go wrong when processing
346        \x{...} escapes in subject strings. The has been rewritten to avoid using
347        things like p++ in the argument of isxdigit().
350    Version 8.13 16-Aug-2011
351    ------------------------
353    1.  The Unicode data tables have been updated to Unicode 6.0.0.
355    2.  Two minor typos in pcre_internal.h have been fixed.
357    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
358        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
359        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
361    4.  There were a number of related bugs in the code for matching backrefences
362        caselessly in UTF-8 mode when codes for the characters concerned were
363        different numbers of bytes. For example, U+023A and U+2C65 are an upper
364        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
365        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
366        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
367        2-byte code at the end of the subject (it thought there wasn't enough data
368        left).
370    5.  Comprehensive information about what went wrong is now returned by
371        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
372        as the output vector has at least 2 elements. The offset of the start of
373        the failing character and a reason code are placed in the vector.
375    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
376        now returned is for the first byte of the failing character, instead of the
377        last byte inspected. This is an incompatible change, but I hope it is small
378        enough not to be a problem. It makes the returned offset consistent with
379        pcre_exec() and pcre_dfa_exec().
381    7.  pcretest now gives a text phrase as well as the error number when
382        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
383        failure, the offset and reason code are output.
385    8.  When \R was used with a maximizing quantifier it failed to skip backwards
386        over a \r\n pair if the subsequent match failed. Instead, it just skipped
387        back over a single character (\n). This seems wrong (because it treated the
388        two characters as a single entity when going forwards), conflicts with the
389        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
390        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
391        has been changed.
393    9.  Some internal refactoring has changed the processing so that the handling
394        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
395        time (the PCRE_DOTALL option was changed this way some time ago: version
396        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
397        which was always a bit of a fudge. It also means that there is one less
398        argument for the match() function, which reduces its stack requirements
399        slightly. This change also fixes an incompatibility with Perl: the pattern
400        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
402    10. More internal refactoring has drastically reduced the number of recursive
403        calls to match() for possessively repeated groups such as (abc)++ when
404        using pcre_exec().
406    11. While implementing 10, a number of bugs in the handling of groups were
407        discovered and fixed:
409        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
410        (a|)*(?1) gave a compile-time internal error.
411        ((a|)+)+  did not notice that the outer group could match an empty string.
412        (^a|^)+   was not marked as anchored.
413        (.*a|.*)+ was not marked as matching at start or after a newline.
415    12. Yet more internal refactoring has removed another argument from the match()
416        function. Special calls to this function are now indicated by setting a
417        value in a variable in the "match data" data block.
419    13. Be more explicit in pcre_study() instead of relying on "default" for
420        opcodes that mean there is no starting character; this means that when new
421        ones are added and accidentally left out of pcre_study(), testing should
422        pick them up.
424    14. The -s option of pcretest has been documented for ages as being an old
425        synonym of -m (show memory usage). I have changed it to mean "force study
426        for every regex", that is, assume /S for every regex. This is similar to -i
427        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
428        using it. It makes it easier to run collections of tests with and without
429        study enabled, and thereby test pcre_study() more easily. All the standard
430        tests are now run with and without -s (but some patterns can be marked as
431        "never study" - see 20 below).
433    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
434        restoration of the capturing data to the outer values was not happening
435        correctly.
437    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
438        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
439        pattern had matched an empty string, and so incorrectly returned a no
440        match.
442    17. There was optimizing code for the last branch of non-capturing parentheses,
443        and also for the obeyed branch of a conditional subexpression, which used
444        tail recursion to cut down on stack usage. Unfortunately, now that there is
445        the possibility of (*THEN) occurring in these branches, tail recursion is
446        no longer possible because the return has to be checked for (*THEN). These
447        two optimizations have therefore been removed. [But see 8.20/11 above.]
449    18. If a pattern containing \R was studied, it was assumed that \R always
450        matched two bytes, thus causing the minimum subject length to be
451        incorrectly computed because \R can also match just one byte.
453    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
454        was incorrectly computed.
456    20. If /S is present twice on a test pattern in pcretest input, it now
457        *disables* studying, thereby overriding the use of -s on the command line
458        (see 14 above). This is necessary for one or two tests to keep the output
459        identical in both cases.
461    21. When (*ACCEPT) was used in an assertion that matched an empty string and
462        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
464    22. When an atomic group that contained a capturing parenthesis was
465        successfully matched, but the branch in which it appeared failed, the
466        capturing was not being forgotten if a higher numbered group was later
467        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
468        group 1 to "a", when in fact it should be unset. This applied to multi-
469        branched capturing and non-capturing groups, repeated or not, and also to
470        positive assertions (capturing in negative assertions does not happen
471        in PCRE) and also to nested atomic groups.
473    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
474        subject after a captured substring, to make it easier to tell which of a
475        number of identical substrings has been captured.
477    24. The way atomic groups are processed by pcre_exec() has been changed so that
478        if they are repeated, backtracking one repetition now resets captured
479        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
480        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
481        "aaa". Previously, it would have been "a". As part of this code
482        refactoring, the way recursive calls are handled has also been changed.
484    25. If an assertion condition captured any substrings, they were not passed
485        back unless some other capturing happened later. For example, if
486        (?(?=(a))a) was matched against "a", no capturing was returned.
488    26. When studying a pattern that contained subroutine calls or assertions,
489        the code for finding the minimum length of a possible match was handling
490        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
491        group 1 called group 2 while simultaneously a separate group 2 called group
492        1). A stack overflow occurred in this case. I have fixed this by limiting
493        the recursion depth to 10.
495    27. Updated RunTest.bat in the distribution to the version supplied by Tom
496        Fortmann. This supports explicit test numbers on the command line, and has
497        argument validation and error reporting.
499    28. An instance of \X with an unlimited repeat could fail if at any point the
500        first character it looked at was a mark character.
502    29. Some minor code refactoring concerning Unicode properties and scripts
503        should reduce the stack requirement of match() slightly.
505    30. Added the '=' option to pcretest to check the setting of unused capturing
506        slots at the end of the pattern, which are documented as being -1, but are
507        not included in the return count.
509    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
510        compiled something random. Now it gives a compile-time error (as does
511        Perl).
513    32. A *MARK encountered during the processing of a positive assertion is now
514        recorded and passed back (compatible with Perl).
516    33. If --only-matching or --colour was set on a pcregrep call whose pattern
517        had alternative anchored branches, the search for a second match in a line
518        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
519        matched the line "0102" twice. The same bug affected patterns that started
520        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
521        twice.
523    34. Previously, PCRE did not allow quantification of assertions. However, Perl
524        does, and because of capturing effects, quantifying parenthesized
525        assertions may at times be useful. Quantifiers are now allowed for
526        parenthesized assertions.
528    35. A minor code tidy in pcre_compile() when checking options for \R usage.
530    36. \g was being checked for fancy things in a character class, when it should
531        just be a literal "g".
533    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
534        appearance of a nested POSIX class supersedes an apparent external class.
535        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
536        unescaped square brackets may also appear as part of class names. For
537        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
538        more like Perl. (But see 8.20/1 above.)
540    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
541        was because it thought it was \N{name}, which is not supported).
543    39. Add minix to OS list not supporting the -S option in pcretest.
545    40. PCRE tries to detect cases of infinite recursion at compile time, but it
546        cannot analyze patterns in sufficient detail to catch mutual recursions
547        such as ((?1))((?2)). There is now a runtime test that gives an error if a
548        subgroup is called recursively as a subpattern for a second time at the
549        same position in the subject string. In previous releases this might have
550        been caught by the recursion limit, or it might have run out of stack.
552    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
553        happen only once. PCRE was, however incorrectly giving a compile time error
554        "recursive call could loop indefinitely" because it cannot analyze the
555        pattern in sufficient detail. The compile time test no longer happens when
556        PCRE is compiling a conditional subpattern, but actual runaway loops are
557        now caught at runtime (see 40 above).
559    42. It seems that Perl allows any characters other than a closing parenthesis
560        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
561        has been changed to be the same.
563    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
564        as not to get warnings when autogen.sh is called. Also changed
565        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
567    44. To help people who use pcregrep to scan files containing exceedingly long
568        lines, the following changes have been made:
570        (a) The default value of the buffer size parameter has been increased from
571            8K to 20K. (The actual buffer used is three times this size.)
573        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
574            PCRE is built.
576        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
577            to be set at run time.
579        (d) Numerical values in pcregrep options can be followed by K or M, for
580            example --buffer-size=50K.
582        (e) If a line being scanned overflows pcregrep's buffer, an error is now
583            given and the return code is set to 2.
585    45. Add a pointer to the latest mark to the callout data block.
587    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
588        partial match of an empty string instead of no match. This was specific to
589        the use of ".".
591    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
592        complete match instead of a partial match. This bug was dependent on both
593        the PCRE_UTF8 and PCRE_DOTALL options being set.
595    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
596        starting byte set, because \b was not being ignored.
599    Version 8.12 15-Jan-2011
600    ------------------------
602    1.  Fixed some typos in the markup of the man pages, and wrote a script that
603        checks for such things as part of the documentation building process.
605    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
606        --match-limit and --recursion-limit options (added for 8.11). In
607        particular, this made one of the standard tests fail. (The integer value
608        went into the wrong half of a long int.)
610    3.  If the --colour option was given to pcregrep with -v (invert match), it
611        did strange things, either producing crazy output, or crashing. It should,
612        of course, ignore a request for colour when reporting lines that do not
613        match.
615    4.  Another pcregrep bug caused similar problems if --colour was specified with
616        -M (multiline) and the pattern match finished with a line ending.
618    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
619        matched in multiline mode, the following line was shown as part of the
620        match. This seems wrong, so I have changed it.
622    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
623        the check for further matches in the same line (so they could be coloured)
624        to overrun the end of the current line. If another match was found, it was
625        incorrectly shown (and then shown again when found in the next line).
627    7.  If pcregrep was compiled under Windows, there was a reference to the
628        function pcregrep_exit() before it was defined. I am assuming this was
629        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
630        reported by a user. I've moved the definition above the reference.
633    Version 8.11 10-Dec-2010
634    ------------------------
636    1.  (*THEN) was not working properly if there were untried alternatives prior
637        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
638        backtracked to try for "b" instead of moving to the next alternative branch
639        at the same level (in this case, to look for "c"). The Perl documentation
640        is clear that when (*THEN) is backtracked onto, it goes to the "next
641        alternative in the innermost enclosing group".
643    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
644        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
645        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
646        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
647        (*THEN).
649    3.  If \s appeared in a character class, it removed the VT character from
650        the class, even if it had been included by some previous item, for example
651        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
652        of \s, but is part of the POSIX "space" class.)
654    4.  A partial match never returns an empty string (because you can always
655        match an empty string at the end of the subject); however the checking for
656        an empty string was starting at the "start of match" point. This has been
657        changed to the "earliest inspected character" point, because the returned
658        data for a partial match starts at this character. This means that, for
659        example, /(?<=abc)def/ gives a partial match for the subject "abc"
660        (previously it gave "no match").
662    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
663        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
664        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
665        has an implication that the given string is incomplete (because a partial
666        match is preferred over a full match). For this reason, these items now
667        give a partial match in this situation. [Aside: previously, the one case
668        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
669        match rather than a full match, which was wrong by the old rules, but is
670        now correct.]
672    6.  There was a bug in the handling of #-introduced comments, recognized when
673        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
674        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
675        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
676        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
677        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
678        places in pcre_compile().
680    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
681        comments when looking ahead for named forward references to subpatterns,
682        the only newline sequence it recognized was NL. It now handles newlines
683        according to the set newline convention.
685    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
686        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
687        cater for a lack of strerror(). These oversights have been fixed.
689    9.  Added --match-limit and --recursion-limit to pcregrep.
691    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
693    11. When the -o option was used, pcregrep was setting a return code of 1, even
694        when matches were found, and --line-buffered was not being honoured.
696    12. Added an optional parentheses number to the -o and --only-matching options
697        of pcregrep.
699    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
700        can match an empty string. The code to do it in pcretest and pcredemo
701        needed fixing:
703        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
704            only one byte after an empty string match just before CRLF (this case
705            just got forgotten; "any" and "anycrlf" were OK).
707        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
708            mode when an empty string match preceded an ASCII character followed by
709            a non-ASCII character. (The code for advancing by one character rather
710            than one byte was nonsense.)
712        (c) The pcredemo.c sample program did not have any code at all to handle
713            the cases when CRLF is a valid newline sequence.
715    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
716        as a starting offset was within the subject string. There is now a new
717        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
718        negative or greater than the length of the string. In order to test this,
719        pcretest is extended to allow the setting of negative starting offsets.
721    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
722        starting offset points to the beginning of a UTF-8 character was
723        unnecessarily clumsy. I tidied it up.
725    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
726        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
728    17. Nobody had reported that the --include_dir option, which was added in
729        release 7.7 should have been called --include-dir (hyphen, not underscore)
730        for compatibility with GNU grep. I have changed it to --include-dir, but
731        left --include_dir as an undocumented synonym, and the same for
732        --exclude-dir, though that is not available in GNU grep, at least as of
733        release 2.5.4.
735    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
736        characters from a string of bytes) have been redefined so as not to use
737        loops, in order to improve performance in some environments. At the same
738        time, I abstracted some of the common code into auxiliary macros to save
739        repetition (this should not affect the compiled code).
741    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
742        compile-time error is now given if \c is not followed by an ASCII
743        character, that is, a byte less than 128. (In EBCDIC mode, the code is
744        different, and any byte value is allowed.)
746    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
747        START_OPTIMIZE option, which is now allowed at compile time - but just
748        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
749        to pcregrep and other applications that have no direct access to PCRE
750        options. The new /Y option in pcretest sets this option when calling
751        pcre_compile().
753    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
754        back references. Groups containing recursive back references were forced to
755        be atomic by that change, but in the case of named groups, the amount of
756        memory required was incorrectly computed, leading to "Failed: internal
757        error: code overflow". This has been fixed.
759    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
760        pcretest.c, to avoid build problems in some Borland environments.
763    Version 8.10 25-Jun-2010
764    ------------------------
766    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
767        THEN.
769    2.  (*ACCEPT) was not working when inside an atomic group.
771    3.  Inside a character class, \B is treated as a literal by default, but
772        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
773        causes the error). The code is unchanged, but I tidied the documentation.
775    4.  Inside a character class, PCRE always treated \R and \X as literals,
776        whereas Perl faults them if its -w option is set. I have changed PCRE so
777        that it faults them when PCRE_EXTRA is set.
779    5.  Added support for \N, which always matches any character other than
780        newline. (It is the same as "." when PCRE_DOTALL is not set.)
782    6.  When compiling pcregrep with newer versions of gcc which may have
783        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
784        declared with attribute warn_unused_result" were given. Just casting the
785        result to (void) does not stop the warnings; a more elaborate fudge is
786        needed. I've used a macro to implement this.
788    7.  Minor change to pcretest.c to avoid a compiler warning.
790    8.  Added four artifical Unicode properties to help with an option to make
791        \s etc use properties (see next item). The new properties are: Xan
792        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
794    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
795        use Unicode properties. (*UCP) at the start of a pattern can be used to set
796        this option. Modified pcretest to add /W to test this facility. Added
797        REG_UCP to make it available via the POSIX interface.
799    10. Added --line-buffered to pcregrep.
801    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
802        studied, and the match started with a letter with a code point greater than
803        127 whose first byte was different to the first byte of the other case of
804        the letter, the other case of this starting letter was not recognized
805        (#976).
807    12. If a pattern that was studied started with a repeated Unicode property
808        test, for example, \p{Nd}+, there was the theoretical possibility of
809        setting up an incorrect bitmap of starting bytes, but fortunately it could
810        not have actually happened in practice until change 8 above was made (it
811        added property types that matched character-matching opcodes).
813    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
814        possible starting bytes for non-anchored patterns.
816    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
817        \R, and also a number of cases that involve Unicode properties, both
818        explicit and implicit when PCRE_UCP is set.
820    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
821        input, it could crash or give wrong results if characters with values
822        greater than 0xc0 were present in the subject string. (Detail: it assumed
823        UTF-8 input when processing these items.)
825    16. Added a lot of (int) casts to avoid compiler warnings in systems where
826        size_t is 64-bit (#991).
828    17. Added a check for running out of memory when PCRE is compiled with
829        --disable-stack-for-recursion (#990).
831    18. If the last data line in a file for pcretest does not have a newline on
832        the end, a newline was missing in the output.
834    19. The default pcre_chartables.c file recognizes only ASCII characters (values
835        less than 128) in its various bitmaps. However, there is a facility for
836        generating tables according to the current locale when PCRE is compiled. It
837        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
838        characters, are recognized by isspace() and therefore were getting set in
839        these tables, and indeed these tables seem to approximate to ISO 8859. This
840        caused a problem in UTF-8 mode when pcre_study() was used to create a list
841        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
842        which of course cannot start UTF-8 characters. I have changed the code so
843        that only real ASCII characters (less than 128) and the correct starting
844        bytes for UTF-8 encodings are set for characters greater than 127 when in
845        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
846        altogether.)
848    20. Added the /T option to pcretest so as to be able to run tests with non-
849        standard character tables, thus making it possible to include the tests
850        used for 19 above in the standard set of tests.
852    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
853        reference to a subpattern the other side of a comment that contains an
854        opening parenthesis caused either an internal compiling error, or a
855        reference to the wrong subpattern.
858    Version 8.02 19-Mar-2010
859    ------------------------
861    1.  The Unicode data tables have been updated to Unicode 5.2.0.
863    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
864        configured.
866    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
867        original author of that file, following a query about its status.
869    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
870        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
872    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
873        quantifier applied to a forward-referencing subroutine call, could compile
874        incorrect code or give the error "internal error: previously-checked
875        referenced subpattern not found".
877    6.  Both MS Visual Studio and Symbian OS have problems with initializing
878        variables to point to external functions. For these systems, therefore,
879        pcre_malloc etc. are now initialized to local functions that call the
880        relevant global functions.
882    7.  There were two entries missing in the vectors called coptable and poptable
883        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
884        I've fixed the data, and added a kludgy way of testing at compile time that
885        the lengths are correct (equal to the number of opcodes).
887    8.  Following on from 7, I added a similar kludge to check the length of the
888        eint vector in pcreposix.c.
890    9.  Error texts for pcre_compile() are held as one long string to avoid too
891        much relocation at load time. To find a text, the string is searched,
892        counting zeros. There was no check for running off the end of the string,
893        which could happen if a new error number was added without updating the
894        string.
896    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
898    11. \K was not working if it appeared in an atomic group or in a group that
899        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
900        \K is "not well defined" if used in an assertion. PCRE now accepts it if
901        the assertion is positive, but not if it is negative.
903    12. Change 11 fortuitously reduced the size of the stack frame used in the
904        "match()" function of pcre_exec.c by one pointer. Forthcoming
905        implementation of support for (*MARK) will need an extra pointer on the
906        stack; I have reserved it now, so that the stack frame size does not
907        decrease.
909    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
910        item in branch that calls a recursion is a subroutine call - as in the
911        second branch in the above example - was incorrectly given the compile-
912        time error "recursive call could loop indefinitely" because pcre_compile()
913        was not correctly checking the subroutine for matching a non-empty string.
915    14. The checks for overrunning compiling workspace could trigger after an
916        overrun had occurred. This is a "should never occur" error, but it can be
917        triggered by pathological patterns such as hundreds of nested parentheses.
918        The checks now trigger 100 bytes before the end of the workspace.
920    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
923    Version 8.01 19-Jan-2010
924    ------------------------
926    1.  If a pattern contained a conditional subpattern with only one branch (in
927        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
928        computed the wrong minimum data length (which is of course zero for such
929        subpatterns). This could cause incorrect "no match" results.
931    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
932        the pattern is reset in the first branch, pcre_compile() failed with
933        "internal error: code overflow at offset...". This happened only when
934        the reset was to the original external option setting. (An optimization
935        abstracts leading options settings into an external setting, which was the
936        cause of this.)
938    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
939        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
940        assertion pattern did not match (meaning that the assertion was true), it
941        was incorrectly treated as false if the SKIP had been reached during the
942        matching. This also applied to assertions used as conditions.
944    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
945        assertion subpattern, including such a pattern used as a condition,
946        unpredictable results occurred, instead of the error return
949    5.  The C++ GlobalReplace function was not working like Perl for the special
950        situation when an empty string is matched. It now does the fancy magic
951        stuff that is necessary.
953    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
954        removed. (These were left over from very, very early versions of PCRE.)
956    7.  Some cosmetic changes to the code to make life easier when compiling it
957        as part of something else:
959        (a) Change DEBUG to PCRE_DEBUG.
961        (b) In pcre_compile(), rename the member of the "branch_chain" structure
962            called "current" as "current_branch", to prevent a collision with the
963            Linux macro when compiled as a kernel module.
965        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
966            prevent a collision with the Linux macro when compiled as a kernel
967            module.
969    8.  In pcre_compile() there are some checks for integer overflows that used to
970        cast potentially large values to (double). This has been changed to that
971        when building, a check for int64_t is made, and if it is found, it is used
972        instead, thus avoiding the use of floating point arithmetic. (There is no
973        other use of FP in PCRE.) If int64_t is not found, the fallback is to
974        double.
976    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
977        2005 (difference between two addresses compared to an unsigned value).
979    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
980        custom one, because of the following reported problem in Windows:
982          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
983              under Win32.
984          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
985              therefore missing the function definition.
986          - The compiler thus generates a "C" signature for the test function.
987          - The linker fails to find the "C" function.
988          - PCRE fails to configure if asked to do so against libbz2.
990    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
991        messages were output:
993          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
994          rerunning libtoolize, to keep the correct libtool macros in-tree.
995          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
997        I have done both of these things.
999    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1000        most of the time, it *can* run out if it is given a pattern that contains a
1001        runaway infinite recursion. I updated the discussion in the pcrestack man
1002        page.
1004    13. Now that we have gone to the x.xx style of version numbers, the minor
1005        version may start with zero. Using 08 or 09 is a bad idea because users
1006        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1007        interpreted as invalid octal numbers. I've updated the previous comment in
1008        configure.ac, and also added a check that gives an error if 08 or 09 are
1009        used.
1011    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1012        causing partial matching to fail when the end of the subject matched \W
1013        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1015    15. There were some discrepancies between the declarations in pcre_internal.h
1016        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1017        their definitions. The declarations used "const uschar *" and the
1018        definitions used USPTR. Even though USPTR is normally defined as "const
1019        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1020        reported that: "This difference in casting confuses some C++ compilers, for
1021        example, SunCC recognizes above declarations as different functions and
1022        generates broken code for hbpcre." I have changed the declarations to use
1023        USPTR.
1025    16. GNU libtool is named differently on some systems. The autogen.sh script now
1026        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1027        (FreeBSD).
1029    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1030        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1031        comment: "Figure out how to create a longlong from a string: strtoll and
1032        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1033        instance, but it only takes 2 args instead of 3!"
1035    18. A subtle bug concerned with back references has been fixed by a change of
1036        specification, with a corresponding code fix. A pattern such as
1037        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1038        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1039        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1040        same bug. Such groups have to be quantified to be useful, or contained
1041        inside another quantified group. (If there's no repetition, the reference
1042        can never match.) The problem arises because, having left the group and
1043        moved on to the rest of the pattern, a later failure that backtracks into
1044        the group uses the captured value from the final iteration of the group
1045        rather than the correct earlier one. I have fixed this in PCRE by forcing
1046        any group that contains a reference to itself to be an atomic group; that
1047        is, there cannot be any backtracking into it once it has completed. This is
1048        similar to recursive and subroutine calls.
1051    Version 8.00 19-Oct-09
1052    ----------------------
1054    1.  The table for translating pcre_compile() error codes into POSIX error codes
1055        was out-of-date, and there was no check on the pcre_compile() error code
1056        being within the table. This could lead to an OK return being given in
1057        error.
1059    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1060        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1061        in a Windows environment.
1063    3.  The pcregrep --count option prints the count for each file even when it is
1064        zero, as does GNU grep. However, pcregrep was also printing all files when
1065        --files-with-matches was added. Now, when both options are given, it prints
1066        counts only for those files that have at least one match. (GNU grep just
1067        prints the file name in this circumstance, but including the count seems
1068        more useful - otherwise, why use --count?) Also ensured that the
1069        combination -clh just lists non-zero counts, with no names.
1071    4.  The long form of the pcregrep -F option was incorrectly implemented as
1072        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1073        but it seems right to fix it, and I didn't think it was worth preserving
1074        the old behaviour.
1076    5.  The command line items --regex=pattern and --regexp=pattern were not
1077        recognized by pcregrep, which required --regex pattern or --regexp pattern
1078        (with a space rather than an '='). The man page documented the '=' forms,
1079        which are compatible with GNU grep; these now work.
1081    6.  No libpcreposix.pc file was created for pkg-config; there was just
1082        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1084    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1085        when UCP support is not needed, by modifying the Python script that
1086        generates it from Unicode data files. This should not matter if the module
1087        is correctly used as a library, but I received one complaint about 50K of
1088        unwanted data. My guess is that the person linked everything into his
1089        program rather than using a library. Anyway, it does no harm.
1091    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1092        was a minimum greater than 1 for a wide character in a possessive
1093        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1094        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1095        character. Chaos in the form of incorrect output or a compiling loop could
1096        result.
1098    9.  The restrictions on what a pattern can contain when partial matching is
1099        requested for pcre_exec() have been removed. All patterns can now be
1100        partially matched by this function. In addition, if there are at least two
1101        slots in the offset vector, the offset of the earliest inspected character
1102        for the match and the offset of the end of the subject are set in them when
1103        PCRE_ERROR_PARTIAL is returned.
1105    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1106        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1107        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1108        and may be more useful for multi-segment matching.
1110    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1111        used to be given if ever the end of the subject was reached; now it is
1112        given only if matching could not proceed because another character was
1113        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1114        string "Z", which now yields "no match" instead of "partial match". In the
1115        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1116        final character ended with (*FAIL).
1118    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1119        if the pattern had a "must contain" character that was already found in the
1120        earlier partial match, unless partial matching was again requested. For
1121        example, with the pattern /dog.(body)?/, the "must contain" character is
1122        "g". If the first part-match was for the string "dog", restarting with
1123        "sbody" failed. This bug has been fixed.
1125    13. The string returned by pcre_dfa_exec() after a partial match has been
1126        changed so that it starts at the first inspected character rather than the
1127        first character of the match. This makes a difference only if the pattern
1128        starts with a lookbehind assertion or \b or \B (\K is not supported by
1129        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1130        matching functions compatible, and I think it's the right thing to do.
1132    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1133        so that the demonstration program is easily available in environments where
1134        PCRE has not been installed from source.
1136    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1137        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1138        library.
1140    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1141        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1142        is not the first non-POSIX option to be added. Clearly some people find
1143        these options useful.
1145    17. If a caller to the POSIX matching function regexec() passes a non-zero
1146        value for nmatch with a NULL value for pmatch, the value of
1147        nmatch is forced to zero.
1149    18. RunGrepTest did not have a test for the availability of the -u option of
1150        the diff command, as RunTest does. It now checks in the same way as
1151        RunTest, and also checks for the -b option.
1153    19. If an odd number of negated classes containing just a single character
1154        interposed, within parentheses, between a forward reference to a named
1155        subpattern and the definition of the subpattern, compilation crashed with
1156        an internal error, complaining that it could not find the referenced
1157        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1158        [The bug was that it was starting one character too far in when skipping
1159        over the character class, thus treating the ] as data rather than
1160        terminating the class. This meant it could skip too much.]
1162    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1163        /g option in pcretest when the pattern contains \K, which makes it possible
1164        to have an empty string match not at the start, even when the pattern is
1165        anchored. Updated pcretest and pcredemo to use this option.
1167    21. If the maximum number of capturing subpatterns in a recursion was greater
1168        than the maximum at the outer level, the higher number was returned, but
1169        with unset values at the outer level. The correct (outer level) value is
1170        now given.
1172    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1173        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1174        make it do so. The string so far is captured, making this feature
1175        compatible with Perl.
1177    23. The tests have been re-organized, adding tests 11 and 12, to make it
1178        possible to check the Perl 5.10 features against Perl 5.10.
1180    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1181        pattern matches a fixed length string. PCRE did not allow this; now it
1182        does. Neither allows recursion.
1184    25. I finally figured out how to implement a request to provide the minimum
1185        length of subject string that was needed in order to match a given pattern.
1186        (It was back references and recursion that I had previously got hung up
1187        on.) This code has now been added to pcre_study(); it finds a lower bound
1188        to the length of subject needed. It is not necessarily the greatest lower
1189        bound, but using it to avoid searching strings that are too short does give
1190        some useful speed-ups. The value is available to calling programs via
1191        pcre_fullinfo().
1193    26. While implementing 25, I discovered to my embarrassment that pcretest had
1194        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1195        study optimizations had never been tested with that matching function.
1196        Oops. What is worse, even when it was passed study data, there was a bug in
1197        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1198        were also very few tests of studied patterns with pcre_dfa_exec().
1200    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1201        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1202        on the other side of the coin, they are no longer allowed to have different
1203        names, because these cannot be distinguished in PCRE, and this has caused
1204        confusion. (This is a difference from Perl.)
1206    28. When duplicate subpattern names are present (necessarily with different
1207        numbers, as required by 27 above), and a test is made by name in a
1208        conditional pattern, either for a subpattern having been matched, or for
1209        recursion in such a pattern, all the associated numbered subpatterns are
1210        tested, and the overall condition is true if the condition is true for any
1211        one of them. This is the way Perl works, and is also more like the way
1212        testing by number works.
1215    Version 7.9 11-Apr-09
1216    ---------------------
1218    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1219        (pcretest), all targets were linked against these libraries. This included
1220        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1221        libraries. This caused unwanted dependencies to be created. This problem
1222        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1223        pcretest is linked with readline.
1225    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1226        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1227        moved outside it again, because FALSE and TRUE are already defined in AIX,
1228        but BOOL is not.
1230    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1231        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1233    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1234        hyphens) following file names and line numbers when outputting matching
1235        lines. This is not true; no spaces are inserted. I have also clarified the
1236        wording for the --colour (or --color) option.
1238    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1239        was not coloured; this is different to GNU grep, so I have changed it to be
1240        the same.
1242    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1243        each matching line was coloured. Now it goes on to look for further matches
1244        of any of the test patterns, which is the same behaviour as GNU grep.
1246    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1247        doesn't make sense to accept an empty string match in pcregrep, so I have
1248        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1249        seems to be how GNU grep behaves.
1251    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1252        start or after a newline", because the conditional assertion was not being
1253        correctly handled. The rule now is that both the assertion and what follows
1254        in the first alternative must satisfy the test.
1256    9.  If auto-callout was enabled in a pattern with a conditional group whose
1257        condition was an assertion, PCRE could crash during matching, both with
1258        pcre_exec() and pcre_dfa_exec().
1260    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1261        used for matching.
1263    11. Unicode property support in character classes was not working for
1264        characters (bytes) greater than 127 when not in UTF-8 mode.
1266    12. Added the -M command line option to pcretest.
1268    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1270    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1272    16. Added comments and documentation about mis-use of no_arg in the C++
1273        wrapper.
1275    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1276        from Martin Jerabek that uses macro names for all relevant character and
1277        string constants.
1279    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1280        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1281        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1282        these, but not everybody uses configure.
1284    19. A conditional group that had only one branch was not being correctly
1285        recognized as an item that could match an empty string. This meant that an
1286        enclosing group might also not be so recognized, causing infinite looping
1287        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1288        with the subject "ab", where knowledge that the repeated group can match
1289        nothing is needed in order to break the loop.
1291    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1292        exec(), but without supplying a callout function, matching went wrong.
1294    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1295        leak if the size of the offset vector was greater than 30. When the vector
1296        is smaller, the saved offsets during recursion go onto a local stack
1297        vector, but for larger vectors malloc() is used. It was failing to free
1298        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1299        error, in fact).
1301    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1302        heapframe that is used only when UTF-8 support is enabled. This caused no
1303        problem, but was untidy.
1305    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1306        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1307        included within another project.
1309    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1310        slightly modified by me:
1312          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1313              not building pcregrep.
1315          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1316              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1318    25. Forward references, both numeric and by name, in patterns that made use of
1319        duplicate group numbers, could behave incorrectly or give incorrect errors,
1320        because when scanning forward to find the reference group, PCRE was not
1321        taking into account the duplicate group numbers. A pattern such as
1322        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1324    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1325        the feature of a custom pointer more persuasive (as requested by a user).
1327    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1328        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1329        pre-defined.
1331    28. Added support for (*UTF8) at the start of a pattern.
1333    29. Arrange for flags added by the "release type" setting in CMake to be shown
1334        in the configuration summary.
1337    Version 7.8 05-Sep-08
1338    ---------------------
1340    1.  Replaced UCP searching code with optimized version as implemented for Ad
1341        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1342        stage table and inline lookup instead of a function, giving speed ups of 2
1343        to 5 times on some simple patterns that I tested. Permission was given to
1344        distribute the MultiStage2.py script that generates the tables (it's not in
1345        the tarball, but is in the Subversion repository).
1347    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1348        scripts.
1350    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1351        a group with a zero qualifier. The result of the study could be incorrect,
1352        or the function might crash, depending on the pattern.
1354    4.  Caseless matching was not working for non-ASCII characters in back
1355        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1356        It now works when Unicode Property Support is available.
1358    5.  In pcretest, an escape such as \x{de} in the data was always generating
1359        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1360        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1361        truncation.
1363    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1365    7.  Added two (int) casts to pcregrep when printing the difference of two
1366        pointers, in case they are 64-bit values.
1368    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1369        test 2 if it fails.
1371    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1372        and a #define of that name to empty if it is not externally set. This is to
1373        allow users of MSVC to set it if necessary.
1375    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1376        the convenience functions in the pcre_get.c source file.
1378    11. An option change at the start of a pattern that had top-level alternatives
1379        could cause overwriting and/or a crash. This command provoked a crash in
1380        some environments:
1382          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1384        This potential security problem was recorded as CVE-2008-2371.
1386    12. For a pattern where the match had to start at the beginning or immediately
1387        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1388        pcre_dfa_exec() could read past the end of the passed subject if there was
1389        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1390        pcretest so that it places the subject at the end of its malloc-ed buffer.
1392    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1393        exec() might read past the end of the data buffer in UTF-8 mode.
1395    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1396        the data contained the byte 0x85 as part of a UTF-8 character within its
1397        first line. This applied both to normal and DFA matching.
1399    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1400        /^[^d]*?$/8 failed to match "abc".
1402    16. Added a missing copyright notice to pcrecpp_internal.h.
1404    17. Make it more clear in the documentation that values returned from
1405        pcre_exec() in ovector are byte offsets, not character counts.
1407    18. Tidied a few places to stop certain compilers from issuing warnings.
1409    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1410        supplied by Stefan Weber. I made a further small update for 7.8 because
1411        there is a change of source arrangements: the pcre_searchfuncs.c module is
1412        replaced by pcre_ucd.c.
1415    Version 7.7 07-May-08
1416    ---------------------
1418    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1419        a string to a long long, pretend we don't even have a long long." This is
1420        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1422    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1423        pre-7.6 versions, which defined a global no_arg variable instead of putting
1424        it in the RE class. (See also #8 below.)
1426    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1427        Lopes.
1429    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1431        (1) The include/exclude patterns were being applied to the whole pathnames
1432            of files, instead of just to the final components.
1434        (2) If there was more than one level of directory, the subdirectories were
1435            skipped unless they satisfied the include/exclude conditions. This is
1436            inconsistent with GNU grep (and could even be seen as contrary to the
1437            pcregrep specification - which I improved to make it absolutely clear).
1438            The action now is always to scan all levels of directory, and just
1439            apply the include/exclude patterns to regular files.
1441    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1442        --exclude_dir in the tests to avoid scanning .svn directories.
1444    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1445        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1446        doesn't support NULs in patterns.
1448    7.  Added some missing "const"s to declarations of static tables in
1449        pcre_compile.c and pcre_dfa_exec.c.
1451    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1452        caused by fix #2  above. (Subsequently also a second patch to fix the
1453        first patch. And a third patch - this was a messy problem.)
1455    9.  Applied Craig's patch to remove the use of push_back().
1457    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1458        matching function regexec().
1460    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1461        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1462        references. PCRE supports relative numbers with this syntax (I don't think
1463        Oniguruma does).
1465    12. Previously, a group with a zero repeat such as (...){0} was completely
1466        omitted from the compiled regex. However, this means that if the group
1467        was called as a subroutine from elsewhere in the pattern, things went wrong
1468        (an internal error was given). Such groups are now left in the compiled
1469        pattern, with a new opcode that causes them to be skipped at execution
1470        time.
1472    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1473        to the way PCRE behaves:
1475        (a) A lone ] character is dis-allowed (Perl treats it as data).
1477        (b) A back reference to an unmatched subpattern matches an empty string
1478            (Perl fails the current match path).
1480        (c) A data ] in a character class must be notated as \] because if the
1481            first data character in a class is ], it defines an empty class. (In
1482            Perl it is not possible to have an empty class.) The empty class []
1483            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1484            The negative empty class [^] matches any one character, independently
1485            of the DOTALL setting.
1487    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1488        non-existent subpattern following a character class starting with ']' and
1489        containing () gave an internal compiling error instead of "reference to
1490        non-existent subpattern". Fortunately, when the pattern did exist, the
1491        compiled code was correct. (When scanning forwards to check for the
1492        existencd of the subpattern, it was treating the data ']' as terminating
1493        the class, so got the count wrong. When actually compiling, the reference
1494        was subsequently set up correctly.)
1496    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1497        it was being rejected as not supported by pcre_dfa_exec(), even though
1498        other assertions are supported. I have made pcre_dfa_exec() support
1499        (*FAIL).
1501    16. The implementation of 13c above involved the invention of a new opcode,
1502        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1503        cannot be changed at match time, I realized I could make a small
1504        improvement to matching performance by compiling OP_ALLANY instead of
1505        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1506        on the OP_ANY path.
1508    17. Compiling pcretest on Windows with readline support failed without the
1509        following two fixes: (1) Make the unistd.h include conditional on
1510        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1512    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1513        ncurses library to be included for pcretest when ReadLine support is
1514        requested, but also to allow for it to be overridden. This patch came from
1515        Daniel Bergström.
1517    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1518        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1519        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1520        spotting this.
1523    Version 7.6 28-Jan-08
1524    ---------------------
1526    1.  A character class containing a very large number of characters with
1527        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1528        overflow.
1530    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1531        HAVE_LONG_LONG is not defined.
1533    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1534        bring them up to date and include new features. This patch includes:
1536        - Fixed PH's badly added libz and libbz2 support.
1537        - Fixed a problem with static linking.
1538        - Added pcredemo. [But later removed - see 7 below.]
1539        - Fixed dftables problem and added an option.
1540        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1541            HAVE_LONG_LONG.
1542        - Added readline support for pcretest.
1543        - Added an listing of the option settings after cmake has run.
1545    4.  A user submitted a patch to Makefile that makes it easy to create
1546        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1547        Makefile.am that cause it to include this special target, without
1548        affecting anything else. Note that the same mingw target plus all
1549        the other distribution libraries and programs are now supported
1550        when configuring with CMake (see 6 below) instead of with
1551        Configure/Make.
1553    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1554        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1555        exported in the Windows port". It has not yet been confirmed that the patch
1556        solves the problem, but it does no harm.
1558    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1559        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1560        with CMake, and also correct the comment about stack recursion.
1562    7.  Remove the automatic building of pcredemo from the ./configure system and
1563        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1564        of a program that users should build themselves after PCRE is installed, so
1565        building it automatically is not really right. What is more, it gave
1566        trouble in some build environments.
1568    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1571    Version 7.5 10-Jan-08
1572    ---------------------
1574    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1575        values in parens when parsing an RE using the C++ wrapper."
1577    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1578        Characters greater than 255 were excluded from the class instead of being
1579        included.
1581    3.  The same bug as (2) above applied to negated POSIX classes such as
1582        [:^space:].
1584    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1585        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1586        I have changed it.
1588    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1589        first named subpattern) and a construct such as (?&a) would reference the
1590        first named subpattern whose name started with "a" (in other words, the
1591        length check was missing). Both these problems are fixed. "Subpattern name
1592        expected" is now given for (?&) (a zero-length name), and this patch also
1593        makes it give the same error for \k'' (previously it complained that that
1594        was a reference to a non-existent subpattern).
1596    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1597        this is right because (?- can be followed by option settings as well as by
1598        digits. I have, however, made the messages clearer.
1600    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1601        than the number used in the conditional) now cause a compile-time error.
1602        This is actually not compatible with Perl, which accepts such patterns, but
1603        treats the conditional as always being FALSE (as PCRE used to), but it
1604        seems to me that giving a diagnostic is better.
1606    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1607        and messages.
1609    9.  Fix two occurrences of "backslash" in comments that should have been
1610        "backspace".
1612    10. Remove two redundant lines of code that can never be obeyed (their function
1613        was moved elsewhere).
1615    11. The program that makes PCRE's Unicode character property table had a bug
1616        which caused it to generate incorrect table entries for sequences of
1617        characters that have the same character type, but are in different scripts.
1618        It amalgamated them into a single range, with the script of the first of
1619        them. In other words, some characters were in the wrong script. There were
1620        thirteen such cases, affecting characters in the following ranges:
1622          U+002b0 - U+002c1
1623          U+0060c - U+0060d
1624          U+0061e - U+00612
1625          U+0064b - U+0065e
1626          U+0074d - U+0076d
1627          U+01800 - U+01805
1628          U+01d00 - U+01d77
1629          U+01d9b - U+01dbf
1630          U+0200b - U+0200f
1631          U+030fc - U+030fe
1632          U+03260 - U+0327f
1633          U+0fb46 - U+0fbb1
1634          U+10450 - U+1049d
1636    12. The -o option (show only the matching part of a line) for pcregrep was not
1637        compatible with GNU grep in that, if there was more than one match in a
1638        line, it showed only the first of them. It now behaves in the same way as
1639        GNU grep.
1641    13. If the -o and -v options were combined for pcregrep, it printed a blank
1642        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1643        does the same. The return code can be used to tell if there were any
1644        non-matching lines.
1646    14. Added --file-offsets and --line-offsets to pcregrep.
1648    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1649        infinitely looping recursion. The bug was that positive lookaheads were not
1650        being skipped when checking for a possible empty match (negative lookaheads
1651        and both kinds of lookbehind were skipped).
1653    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1654        inclusion of <windows.h> to before rather than after the definition of
1655        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1657    17. Specifying a possessive quantifier with a specific limit for a Unicode
1658        character property caused pcre_compile() to compile bad code, which led at
1659        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1660        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1661        caused the error; without that there was no problem.
1663    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1665    19. Added --enable-pcretest-libreadline.
1667    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1668        RE::GlobalReplace(). As a result, the number of replacements returned was
1669        double what it should be. I removed one of the increments, but Craig sent a
1670        later patch that removed the other one (the right fix) and added unit tests
1671        that check the return values (which was not done before).
1673    21. Several CMake things:
1675        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1676            the names libpcre and libpcreposix, not just pcre and pcreposix.
1678        (2) The above change means that pcretest and pcregrep are now correctly
1679            linked with the newly-built libraries, not previously installed ones.
1683    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1684        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1685        UTF-8 newline character). The key issue is that the pattern starts .*;
1686        this means that the match must be either at the beginning, or after a
1687        newline. The bug was in the code for advancing after a failed match and
1688        checking that the new position followed a newline. It was not taking
1689        account of UTF-8 characters correctly.
1691    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1692        character classes. PCRE was not treating the sequence [:...:] as a
1693        character class unless the ... were all letters. Perl, however, seems to
1694        allow any characters between [: and :], though of course it rejects as
1695        unknown any "names" that contain non-letters, because all the known class
1696        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1697        for example, whereas PCRE did not - it did not recognize a POSIX character
1698        class. This seemed a bit dangerous, so the code has been changed to be
1699        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1700        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1701        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1702        Perl does, and where it didn't before.
1704    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1705        Windows environments %n is disabled by default.
1708    Version 7.4 21-Sep-07
1709    ---------------------
1711    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1712        means that a class such as [\s] counted as "explicit reference to CR or
1713        LF". That isn't really right - the whole point of the change was to try to
1714        help when there was an actual mention of one of the two characters. So now
1715        the change happens only if \r or \n (or a literal CR or LF) character is
1716        encountered.
1718    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1719        of both had grown to the point where there were only 3 bits left.
1720        Fortunately, there was spare space in the data structure, and so I have
1721        moved the internal flags into a new 16-bit field to free up more option
1722        bits.
1724    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1725        but did not set the internal JCHANGED flag - either of these is enough to
1726        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1727        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1728        start sets both bits.
1730    4.  Added options (at build time, compile time, exec time) to change \R from
1731        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1733    5.  doc/pcresyntax.html was missing from the distribution.
1735    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1736        compatibility, even though it is no longer used.
1738    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1739        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1740        windows.h file is present (where different names are used). [This was
1741        reversed later after testing - see 16 below.]
1743    8.  Changed all #include <config.h> to #include "config.h". There were also
1744        some further <pcre.h> cases that I changed to "pcre.h".
1746    9.  When pcregrep was used with the --colour option, it missed the line ending
1747        sequence off the lines that it output.
1749    10. It was pointed out to me that arrays of string pointers cause lots of
1750        relocations when a shared library is dynamically loaded. A technique of
1751        using a single long string with a table of offsets can drastically reduce
1752        these. I have refactored PCRE in four places to do this. The result is
1753        dramatic:
1755          Originally:                          290
1756          After changing UCP table:            187
1757          After changing error message table:   43
1758          After changing table of "verbs"       36
1759          After changing table of Posix names   22
1761        Thanks to the folks working on Gregex for glib for this insight.
1763    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1764        unicode-properties was also set.
1766    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1768    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1769        checked only for CRLF.
1771    14. Added casts to pcretest.c to avoid compiler warnings.
1773    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1775    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1776        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1777        entirely. This removes changes made in 7 above.
1779    17. The CMake files have been updated, and there is now more information about
1780        building with CMake in the NON-UNIX-USE document.
1783    Version 7.3 28-Aug-07
1784    ---------------------
1786     1. In the rejigging of the build system that eventually resulted in 7.1, the
1787        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1788        brackets there is not right, since it causes compilers to look for an
1789        installed pcre.h, not the version that is in the source that is being
1790        compiled (which of course may be different). I have changed it back to:
1792          #include "pcre.h"
1794        I have a vague recollection that the change was concerned with compiling in
1795        different directories, but in the new build system, that is taken care of
1796        by the VPATH setting the Makefile.
1798     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1799        when the subject happened to end in the byte 0x85 (e.g. if the last
1800        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1801        characters but of course it shouldn't be taken as a newline when it is part
1802        of another character. The bug was that, for an unlimited repeat of . in
1803        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1804        characters when looking for a newline.
1806     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1808     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1809        in debug output.
1811     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1812        long printing in the pcrecpp unittest when running under MinGW.
1814     6. ESC_K was left out of the EBCDIC table.
1816     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1817        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1818        limit also applies to "virtual nesting" when a pattern is recursive, and in
1819        this case 1000 isn't so big. I have been able to remove this limit at the
1820        expense of backing off one optimization in certain circumstances. Normally,
1821        when pcre_exec() would call its internal match() function recursively and
1822        immediately return the result unconditionally, it uses a "tail recursion"
1823        feature to save stack. However, when a subpattern that can match an empty
1824        string has an unlimited repetition quantifier, it no longer makes this
1825        optimization. That gives it a stack frame in which to save the data for
1826        checking that an empty string has been matched. Previously this was taken
1827        from the 1000-entry workspace that had been reserved. So now there is no
1828        explicit limit, but more stack is used.
1830     8. Applied Daniel's patches to solve problems with the import/export magic
1831        syntax that is required for Windows, and which was going wrong for the
1832        pcreposix and pcrecpp parts of the library. These were overlooked when this
1833        problem was solved for the main library.
1835     9. There were some crude static tests to avoid integer overflow when computing
1836        the size of patterns that contain repeated groups with explicit upper
1837        limits. As the maximum quantifier is 65535, the maximum group length was
1838        set at 30,000 so that the product of these two numbers did not overflow a
1839        32-bit integer. However, it turns out that people want to use groups that
1840        are longer than 30,000 bytes (though not repeat them that many times).
1841        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1842        made it possible to implement the integer overflow checks in a much more
1843        dynamic way, which I have now done. The artificial limitation on group
1844        length has been removed - we now have only the limit on the total length of
1845        the compiled pattern, which depends on the LINK_SIZE setting.
1847    10. Fixed a bug in the documentation for get/copy named substring when
1848        duplicate names are permitted. If none of the named substrings are set, the
1849        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1850        empty string.
1852    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1853        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1854        because the ] is interpreted as the first data character and the
1855        terminating ] is not found. PCRE has been made compatible with Perl in this
1856        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1857        cause memory overwriting.
1859    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1860        string has been matched (to stop an infinite loop). It was not recognizing
1861        a conditional subpattern that could match an empty string if that
1862        subpattern was within another subpattern. For example, it looped when
1863        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1864        condition was not nested. This bug has been fixed.
1866    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1867        past the start of the subject in the presence of bytes with the top bit
1868        set, for example "\x8aBCD".
1870    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1871        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1873    14. Optimized (?!) to (*FAIL).
1875    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1876        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1877        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1878        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1879        does: it's just the validity check that is more restrictive.
1881    16. Inserted checks for integer overflows during escape sequence (backslash)
1882        processing, and also fixed erroneous offset values for syntax errors during
1883        backslash processing.
1885    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1886        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1888    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1889        caused an overrun.
1891    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1892        something other than just ASCII characters) inside a group that had an
1893        unlimited repeat caused a loop at compile time (while checking to see
1894        whether the group could match an empty string).
1896    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1897        [\P{Any}] did so. (Error in the code for printing property names.)
1899    21. An orphan \E inside a character class could cause a crash.
1901    22. A repeated capturing bracket such as (A)? could cause a wild memory
1902        reference during compilation.
1904    23. There are several functions in pcre_compile() that scan along a compiled
1905        expression for various reasons (e.g. to see if it's fixed length for look
1906        behind). There were bugs in these functions when a repeated \p or \P was
1907        present in the pattern. These operators have additional parameters compared
1908        with \d, etc, and these were not being taken into account when moving along
1909        the compiled data. Specifically:
1911        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1912            length.
1914        (b) An item such as \pL+ within a repeated group could cause crashes or
1915            loops.
1917        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1918            "reference to non-existent subpattern" error.
1920        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1922    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1923        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1925    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1927    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1928        character were causing crashes (broken optimization).
1930    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1931        \p or \P) caused a compile-time loop.
1933    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1934        break. For example, the unstudied pattern [\r\n]A does not match the string
1935        "\r\nA" because change 7.0/46 below moves the current point on by two
1936        characters after failing to match at the start. However, the pattern \nA
1937        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1938        the same is true. There doesn't seem any very clean way out of this, but
1939        what I have chosen to do makes the common cases work: PCRE now takes note
1940        of whether there can be an explicit match for \r or \n anywhere in the
1941        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1942        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1943        pattern has explicit CR or LF references.
1945    29. Added (*CR) etc for changing newline setting at start of pattern.
1948    Version 7.2 19-Jun-07
1949    ---------------------
1951     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1952        which is apparently normally available under Windows.
1954     2. Re-jig the pcregrep tests with different newline settings in an attempt
1955        to make them independent of the local environment's newline setting.
1957     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1959     4. Some of the "internals" tests were previously cut out when the link size
1960        was not 2, because the output contained actual offsets. The recent new
1961        "Z" feature of pcretest means that these can be cut out, making the tests
1962        usable with all link sizes.
1964     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1965        stack recursion. This gives a massive performance boost under BSD, but just
1966        a small improvement under Linux. However, it saves one field in the frame
1967        in all cases.
1969     6. Added more features from the forthcoming Perl 5.10:
1971        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1972            recursion call. It refers to the nth most recently opened parentheses.
1974        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1975            to be opened parentheses.
1977        (c) Conditions that refer to capturing parentheses can be specified
1978            relatively, for example, (?(-2)... or (?(+3)...
1980        (d) \K resets the start of the current match so that everything before
1981            is not part of it.
1983        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1985        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1986            reference syntax.
1988        (g) (?| introduces a group in which the numbering of parentheses in each
1989            alternative starts with the same number.
1991        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1993     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1996     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1997        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1998        for detecting groups that can match an empty string.
2000     9. A pattern with a very large number of alternatives (more than several
2001        hundred) was running out of internal workspace during the pre-compile
2002        phase, where pcre_compile() figures out how much memory will be needed. A
2003        bit of new cunning has reduced the workspace needed for groups with
2004        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2005        workspace instead of running out of the 4096 that are available.
2007    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2009    11. Applied patch from Google to remove an optimization that didn't quite work.
2010        The report of the bug said:
2012          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2013          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2014          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2016    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2017        it matched the wrong number of bytes.
2020    Version 7.1 24-Apr-07
2021    ---------------------
2023     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2024        that is more "standard", making use of automake and other Autotools. There
2025        is some re-arrangement of the files and adjustment of comments consequent
2026        on this.
2028     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2029        for recursive directory scanning broke on some systems because the files
2030        are not scanned in any specific order and on different systems the order
2031        was different. A call to "sort" has been inserted into RunGrepTest for the
2032        approprate test as a short-term fix. In the longer term there may be an
2033        alternative.
2035     3. I had an email from Eric Raymond about problems translating some of PCRE's
2036        man pages to HTML (despite the fact that I distribute HTML pages, some
2037        people do their own conversions for various reasons). The problems
2038        concerned the use of low-level troff macros .br and .in. I have therefore
2039        removed all such uses from the man pages (some were redundant, some could
2040        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2041        HTML has been updated to handle .nf/.fi and to complain if it encounters
2042        .br or .in.
2044     4. Updated comments in configure.ac that get placed in config.h.in and also
2045        arranged for config.h to be included in the distribution, with the name
2046        config.h.generic, for the benefit of those who have to compile without
2047        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2049     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2050        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2051        makevp.bat for the current PCRE, using the additional files
2052        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2054     6. A Windows user reported a minor discrepancy with test 2, which turned out
2055        to be caused by a trailing space on an input line that had got lost in his
2056        copy. The trailing space was an accident, so I've just removed it.
2058     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2059        that is needed.
2061     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2062        as "const" (a) because they are and (b) because it helps the PHP
2063        maintainers who have recently made a script to detect big data structures
2064        in the php code that should be moved to the .rodata section. I remembered
2065        to update Builducptable as well, so it won't revert if ucptable.h is ever
2066        re-created.
2068     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2069        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2070        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2071        support is not required. This saves 1.5-2K of code, which is important in
2072        some applications.
2074        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2075        so as not to refer to the tables, even though these functions will never be
2076        called when UTF-8 support is disabled. Otherwise there are problems with a
2077        shared library.
2079    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2081        (a) It was defining its arguments as char * instead of void *.
2083        (b) It was assuming that all moves were upwards in memory; this was true
2084            a long time ago when I wrote it, but is no longer the case.
2086        The emulated memove() is provided for those environments that have neither
2087        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2088        is clearly not the case, as these two bugs were recently reported.
2090    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2091        and Detrail to create the HTML documentation, the .txt form of the man
2092        pages, and it removes trailing spaces from listed files. It also creates
2093        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2094        case, it wraps all the #defines with #ifndefs. This script should be run
2095        before "make dist".
2097    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2098        with Unicode property support.
2100        (a) For a maximizing quantifier, if the two different cases of the
2101            character were of different lengths in their UTF-8 codings (there are
2102            some cases like this - I found 11), and the matching function had to
2103            back up over a mixture of the two cases, it incorrectly assumed they
2104            were both the same length.
2106        (b) When PCRE was configured to use the heap rather than the stack for
2107            recursion during matching, it was not correctly preserving the data for
2108            the other case of a UTF-8 character when checking ahead for a match
2109            while processing a minimizing repeat. If the check also involved
2110            matching a wide character, but failed, corruption could cause an
2111            erroneous result when trying to check for a repeat of the original
2112            character.
2114    13. Some tidying changes to the testing mechanism:
2116        (a) The RunTest script now detects the internal link size and whether there
2117            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2118            values substituted by "configure". (The RunGrepTest script already did
2119            this for UTF-8.) The configure.ac script no longer substitutes the
2120            relevant variables.
2122        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2123            with length and offset values. This means that the output is different
2124            for different internal link sizes. Test 2 is skipped for link sizes
2125            other than 2 because of this, bypassing the problem. Unfortunately,
2126            there was also a test in test 3 (the locale tests) that used /B and
2127            failed for link sizes other than 2. Rather than cut the whole test out,
2128            I have added a new /Z option to pcretest that replaces the length and
2129            offset values with spaces. This is now used to make test 3 independent
2130            of link size. (Test 2 will be tidied up later.)
2132    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2133        segmentation fault instead of returning the appropriate error message.
2135    15. In multiline mode when the newline sequence was set to "any", the pattern
2136        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2137        This doesn't seem right; it now treats the CRLF combination as the line
2138        ending, and so does not match in that case. It's only a pattern such as ^$
2139        that would hit this one: something like ^ABC$ would have failed after \r
2140        and then tried again after \r\n.
2142    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2143        in an attempt to make files that differ only in their line terminators
2144        compare equal. This works on Linux.
2146    17. Under certain error circumstances pcregrep might try to free random memory
2147        as it exited. This is now fixed, thanks to valgrind.
2149    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2150        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2151        was because its rules for how to advance for /g after matching an empty
2152        string at the end of a line did not allow for this case. They now check for
2153        it specially.
2155    20. pcretest is supposed to handle patterns and data of any length, by
2156        extending its buffers when necessary. It was getting this wrong when the
2157        buffer for a data line had to be extended.
2159    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2160        CRLF as a newline sequence.
2162    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2163        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2164        I have nevertheless tidied it up.
2166    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2168    24. Added a man page for pcre-config.
2171    Version 7.0 19-Dec-06
2172    ---------------------
2174     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2175        moving to gcc 4.1.1.
2177     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2178        sys/time.h, which is documented as needed for this function. It doesn't
2179        seem to matter on Linux, but it showed up on some releases of OS X.
2181     3. It seems that there are systems where bytes whose values are greater than
2182        127 match isprint() in the "C" locale. The "C" locale should be the
2183        default when a C program starts up. In most systems, only ASCII printing
2184        characters match isprint(). This difference caused the output from pcretest
2185        to vary, making some of the tests fail. I have changed pcretest so that:
2187        (a) When it is outputting text in the compiled version of a pattern, bytes
2188            other than 32-126 are always shown as hex escapes.
2190        (b) When it is outputting text that is a matched part of a subject string,
2191            it does the same, unless a different locale has been set for the match
2192            (using the /L modifier). In this case, it uses isprint() to decide.
2194     4. Fixed a major bug that caused incorrect computation of the amount of memory
2195        required for a compiled pattern when options that changed within the
2196        pattern affected the logic of the preliminary scan that determines the
2197        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2198        that the computed length was too small. The symptoms of this bug were
2199        either the PCRE error "internal error: code overflow" from pcre_compile(),
2200        or a glibc crash with a message such as "pcretest: free(): invalid next
2201        size (fast)". Examples of patterns that provoked this bug (shown in
2202        pcretest format) are:
2204          /(?-x: )/x
2205          /(?x)(?-x: \s*#\s*)/
2206          /((?i)[\x{c0}])/8
2207          /(?i:[\x{c0}])/8
2209        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2210        is now done differently.
2212     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2213        wrapper classes; (b) implement a new function in the C++ scanner that is
2214        more efficient than the old way of doing things because it avoids levels of
2215        recursion in the regex matching; (c) add a paragraph to the documentation
2216        for the FullMatch() function.
2218     6. The escape sequence \n was being treated as whatever was defined as
2219        "newline". Not only was this contrary to the documentation, which states
2220        that \n is character 10 (hex 0A), but it also went horribly wrong when
2221        "newline" was defined as CRLF. This has been fixed.
2223     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2224        was being set to -1 for the "end of line" case (supposedly a value that no
2225        character can have). Though this value is never used (the check for end of
2226        line is "zero bytes in current character"), it caused compiler complaints.
2227        I've changed it to 0xffffffff.
2229     8. In pcre_version.c, the version string was being built by a sequence of
2230        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2231        string (as it is for production releases) called a macro with an empty
2232        argument. The C standard says the result of this is undefined. The gcc
2233        compiler treats it as an empty string (which was what was wanted) but it is
2234        reported that Visual C gives an error. The source has been hacked around to
2235        avoid this problem.
2237     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2238        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2239        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2240        of them did).
2242    10. Originally, pcretest opened its input and output without "b"; then I was
2243        told that "b" was needed in some environments, so it was added for release
2244        5.0 to both the input and output. (It makes no difference on Unix-like
2245        systems.) Later I was told that it is wrong for the input on Windows. I've
2246        now abstracted the modes into two macros, to make it easier to fiddle with
2247        them, and removed "b" from the input mode under Windows.
2249    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2251    12. Added -help and --help to pcretest as an official way of being reminded
2252        of the options.
2254    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2255        and pcrecpp.cc because they annoy compilers at high warning levels.
2257    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2259    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2260        scripts are not C programs :-) and which was not noticed because it works
2261        on Linux.
2263    16. pcretest is supposed to handle any length of pattern and data line (as one
2264        line or as a continued sequence of lines) by extending its input buffer if
2265        necessary. This feature was broken for very long pattern lines, leading to
2266        a string of junk being passed to pcre_compile() if the pattern was longer
2267        than about 50K.
2269    17. I have done a major re-factoring of the way pcre_compile() computes the
2270        amount of memory needed for a compiled pattern. Previously, there was code
2271        that made a preliminary scan of the pattern in order to do this. That was
2272        OK when PCRE was new, but as the facilities have expanded, it has become
2273        harder and harder to keep it in step with the real compile phase, and there
2274        have been a number of bugs (see for example, 4 above). I have now found a
2275        cunning way of running the real compile function in a "fake" mode that
2276        enables it to compute how much memory it would need, while actually only
2277        ever using a few hundred bytes of working memory and without too many
2278        tests of the mode. This should make future maintenance and development
2279        easier. A side effect of this work is that the limit of 200 on the nesting
2280        depth of parentheses has been removed (though this was never a serious
2281        limitation, I suspect). However, there is a downside: pcre_compile() now
2282        runs more slowly than before (30% or more, depending on the pattern). I
2283        hope this isn't a big issue. There is no effect on runtime performance.
2285    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2286        newline (only possible for the last line of a file) and it was a
2287        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2289    19. Added additional timing features to pcretest. (1) The -tm option now times
2290        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2291        separate command line item, by a number that specifies the number of
2292        repeats to use when timing. The default is 50000; this gives better
2293        precision, but takes uncomfortably long for very large patterns.
2295    20. Extended pcre_study() to be more clever in cases where a branch of a
2296        subpattern has no definite first character. For example, (a*|b*)[cd] would
2297        previously give no result from pcre_study(). Now it recognizes that the
2298        first character must be a, b, c, or d.
2300    21. There was an incorrect error "recursive call could loop indefinitely" if
2301        a subpattern (or the entire pattern) that was being tested for matching an
2302        empty string contained only one non-empty item after a nested subpattern.
2303        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2304        incorrectly, because the \d was being skipped in the check.
2306    22. The pcretest program now has a new pattern option /B and a command line
2307        option -b, which is equivalent to adding /B to every pattern. This causes
2308        it to show the compiled bytecode, without the additional information that
2309        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2310        is the same as /B/I).
2312    23. A new optimization is now able automatically to treat some sequences such
2313        as a*b as a*+b. More specifically, if something simple (such as a character
2314        or a simple class like \d) has an unlimited quantifier, and is followed by
2315        something that cannot possibly match the quantified thing, the quantifier
2316        is automatically "possessified".
2318    24. A recursive reference to a subpattern whose number was greater than 39
2319        went wrong under certain circumstances in UTF-8 mode. This bug could also
2320        have affected the operation of pcre_study().
2322    25. Realized that a little bit of performance could be had by replacing
2323        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2325    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2327    27. Possessive quantifiers such as a++ were previously implemented by turning
2328        them into atomic groups such as ($>a+). Now they have their own opcodes,
2329        which improves performance. This includes the automatically created ones
2330        from 23 above.
2332    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2333        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2334        the first matched character to be a colon. This applied both to named and
2335        numbered groups.
2337    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2339    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2340        building PCRE on Windows easier, so I have included it in the distribution.
2342    31. There is now a check in pcretest against a ridiculously large number being
2343        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2344        loop, the loop is abandoned.
2346    32. Forward references to subpatterns in conditions such as (?(2)...) where
2347        subpattern 2 is defined later cause pcre_compile() to search forwards in
2348        the pattern for the relevant set of parentheses. This search went wrong
2349        when there were unescaped parentheses in a character class, parentheses
2350        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2352    33. "Subroutine" calls and backreferences were previously restricted to
2353        referencing subpatterns earlier in the regex. This restriction has now
2354        been removed.
2356    34. Added a number of extra features that are going to be in Perl 5.10. On the
2357        whole, these are just syntactic alternatives for features that PCRE had
2358        previously implemented using the Python syntax or my own invention. The
2359        other formats are all retained for compatibility.
2361        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2362            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2363            also .NET compatible.
2365        (b) A recursion or subroutine call to a named group can now be defined as
2366            (?&name) as well as (?P>name).
2368        (c) A backreference to a named group can now be defined as \k<name> or
2369            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2370            5.10, are also .NET compatible.
2372        (d) A conditional reference to a named group can now use the syntax
2373            (?(<name>) or (?('name') as well as (?(name).
2375        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2376            groups (named and numbered) that are never evaluated inline, but can be
2377            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2378            is always false. There may be only one alternative in such a group.
2380        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2381            as the simple (?(R). The condition is true only if the most recent
2382            recursion is that of the given number or name. It does not search out
2383            through the entire recursion stack.
2385        (g) The escape \gN or \g{N} has been added, where N is a positive or
2386            negative number, specifying an absolute or relative reference.
2388    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2389        some "unreachable code" warnings.
2391    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2392        things, this adds five new scripts.
2394    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2395        There were also incompatibilities regarding the handling of \Q..\E inside
2396        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2397        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2399    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2400        matches an empty string, and forcibly breaks the loop. There were bugs in
2401        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2402        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2403        separate and independent bugs (that affected different cases) have been
2404        fixed.
2406    39. Refactored the code to abolish the use of different opcodes for small
2407        capturing bracket numbers. This is a tidy that I avoided doing when I
2408        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2409        The new approach is not only tidier, it makes it possible to reduce the
2410        memory needed to fix the previous bug (38).
2412    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2413        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2414        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2415        mode.
2417    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2418        report.
2420    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2421        copy construction and assignment in the C++ wrapper.
2423    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2424        couple of bugs that could have given wrong results in the "--newline=crlf"
2425        case.
2427    44. Added a number of casts and did some reorganization of signed/unsigned int
2428        variables following suggestions from Dair Grant. Also renamed the variable
2429        "this" as "item" because it is a C++ keyword.
2431    45. Arranged for dftables to add
2433          #include "pcre_internal.h"
2435        to pcre_chartables.c because without it, gcc 4.x may remove the array
2436        definition from the final binary if PCRE is built into a static library and
2437        dead code stripping is activated.
2439    46. For an unanchored pattern, if a match attempt fails at the start of a
2440        newline sequence, and the newline setting is CRLF or ANY, and the next two
2441        characters are CRLF, advance by two characters instead of one.
2444    Version 6.7 04-Jul-06
2445    ---------------------
2447     1. In order to handle tests when input lines are enormously long, pcretest has
2448        been re-factored so that it automatically extends its buffers when
2449        necessary. The code is crude, but this _is_ just a test program. The
2450        default size has been increased from 32K to 50K.
2452     2. The code in pcre_study() was using the value of the re argument before
2453        testing it for NULL. (Of course, in any sensible call of the function, it
2454        won't be NULL.)
2456     3. The memmove() emulation function in pcre_internal.h, which is used on
2457        systems that lack both memmove() and bcopy() - that is, hardly ever -
2458        was missing a "static" storage class specifier.
2460     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2461        containing an extended class (one that cannot be represented by a bitmap
2462        because it contains high-valued characters or Unicode property items, e.g.
2463        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2464        pattern, but PCRE should not loop if you do not (it no longer does).
2465        [Detail: two cases were found: (a) a repeated subpattern containing an
2466        extended class; (b) a recursive reference to a subpattern that followed a
2467        previous extended class. It wasn't skipping over the extended class
2468        correctly when UTF-8 mode was not set.]
2470     5. A negated single-character class was not being recognized as fixed-length
2471        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2472        compile error "lookbehind assertion is not fixed length".
2474     6. The RunPerlTest auxiliary script was showing an unexpected difference
2475        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2476        write a Perl script that can interpret lines of an input file either as
2477        byte characters or as UTF-8, which is what "perltest" was being required to
2478        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2479        can't do is switch easily at run time between having the "use utf8;" pragma
2480        or not. In the end, I fudged it by using the RunPerlTest script to insert
2481        "use utf8;" explicitly for the UTF-8 tests.
2483     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2484        the end of the subject string, contrary to the documentation and to what
2485        Perl does. This was true of both matching functions. Now it matches only at
2486        the start of the subject and immediately after *internal* newlines.
2488     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2489        a pointer to an int instead of a pointer to an unsigned long int. This
2490        caused problems on 64-bit systems.
2492     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2493        instance of the 'standard' template library not being so standard".
2495    10. There was no check on the number of named subpatterns nor the maximum
2496        length of a subpattern name. The product of these values is used to compute
2497        the size of the memory block for a compiled pattern. By supplying a very
2498        long subpattern name and a large number of named subpatterns, the size
2499        computation could be caused to overflow. This is now prevented by limiting
2500        the length of names to 32 characters, and the number of named subpatterns
2501        to 10,000.
2503    11. Subpatterns that are repeated with specific counts have to be replicated in
2504        the compiled pattern. The size of memory for this was computed from the
2505        length of the subpattern and the repeat count. The latter is limited to
2506        65535, but there was no limit on the former, meaning that integer overflow
2507        could in principle occur. The compiled length of a repeated subpattern is
2508        now limited to 30,000 bytes in order to prevent this.
2510    12. Added the optional facility to have named substrings with the same name.
2512    13. Added the ability to use a named substring as a condition, using the
2513        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2514        are numbers (not recommended). Forward references are permitted.
2516    14. Added forward references in named backreferences (if you see what I mean).
2518    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2519        pattern could run off the end of the subject. For example, the pattern
2520        "(?s)(.{1,5})"8 did this with the subject "ab".
2522    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2523        PCRE_CASELESS was set when matching characters that were quantified with ?
2524        or *.
2526    17. A character class other than a single negated character that had a minimum
2527        but no maximum quantifier - for example [ab]{6,} - was not handled
2528        correctly by pce_dfa_exec(). It would match only one character.
2530    18. A valid (though odd) pattern that looked like a POSIX character
2531        class but used an invalid character after [ (for example [[,abc,]]) caused
2532        pcre_compile() to give the error "Failed: internal error: code overflow" or
2533        in some cases to crash with a glibc free() error. This could even happen if
2534        the pattern terminated after [[ but there just happened to be a sequence of
2535        letters, a binary zero, and a closing ] in the memory that followed.
2537    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2538        over the years. Originally (before any Unicode support), just the bottom 8
2539        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2540        output from "man perlunicode" includes this:
2542          The regular expression compiler produces polymorphic opcodes.  That
2543          is, the pattern adapts to the data and automatically switches to
2544          the Unicode character scheme when presented with Unicode data--or
2545          instead uses a traditional byte scheme when presented with byte
2546          data.
2548        Sadly, a wide octal escape does not cause a switch, and in a string with
2549        no other multibyte characters, these octal escapes are treated as before.
2550        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2551        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2552        Unicode string.
2554        I have not perpetrated such confusion in PCRE. Up till now, it took just
2555        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2556        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2557        translate to the appropriate multibyte character.
2559    29. Applied some refactoring to reduce the number of warnings from Microsoft
2560        and Borland compilers. This has included removing the fudge introduced
2561        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2562        a warning about an unused variable.
2564    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2565        characters since release 4.0, because Perl (from release 5.004) does not.
2566        [Or at least, is documented not to: some releases seem to be in conflict
2567        with the documentation.] However, when a pattern was studied with
2568        pcre_study() and all its branches started with \s, PCRE still included VT
2569        as a possible starting character. Of course, this did no harm; it just
2570        caused an unnecessary match attempt.
2572    22. Removed a now-redundant internal flag bit that recorded the fact that case
2573        dependency changed within the pattern. This was once needed for "required
2574        byte" processing, but is no longer used. This recovers a now-scarce options
2575        bit. Also moved the least significant internal flag bit to the most-
2576        significant bit of the word, which was not previously used (hangover from
2577        the days when it was an int rather than a uint) to free up another bit for
2578        the future.
2580    23. Added support for CRLF line endings as well as CR and LF. As well as the
2581        default being selectable at build time, it can now be changed at runtime
2582        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2583        specify that it is scanning data with non-default line endings.
2585    24. Changed the definition of CXXLINK to make it agree with the definition of
2586        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2588    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2589        recursions. This makes a big different to stack usage for some patterns.
2591    26. If a subpattern containing a named recursion or subroutine reference such
2592        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2593        the space required for the compiled pattern went wrong and gave too small a
2594        value. Depending on the environment, this could lead to "Failed: internal
2595        error: code overflow at offset 49" or "glibc detected double free or
2596        corruption" errors.
2598    27. Applied patches from Google (a) to support the new newline modes and (b) to
2599        advance over multibyte UTF-8 characters in GlobalReplace.
2601    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2602        difference for some implementation of PCRE in some Windows version.
2604    29. Added some extra testing facilities to pcretest:
2606        \q<number>   in a data line sets the "match limit" value
2607        \Q<number>   in a data line sets the "match recursion limt" value
2608        -S <number>  sets the stack size, where <number> is in megabytes
2610        The -S option isn't available for Windows.
2613    Version 6.6 06-Feb-06
2614    ---------------------
2616     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2617        in pcreposix.h. I have copied the definition from pcre.h.
2619     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2620        because pcre.h is no longer a built file.
2622     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2623        not normally included in the compiled code.
2626    Version 6.5 01-Feb-06
2627    ---------------------
2629     1. When using the partial match feature with pcre_dfa_exec(), it was not
2630        anchoring the second and subsequent partial matches at the new starting
2631        point. This could lead to incorrect results. For example, with the pattern
2632        /1234/, partially matching against "123" and then "a4" gave a match.
2634     2. Changes to pcregrep:
2636        (a) All non-match returns from pcre_exec() were being treated as failures
2637            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2638            error message is output. Some extra information is given for the
2639            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2640            probably the only errors that are likely to be caused by users (by
2641            specifying a regex that has nested indefinite repeats, for instance).
2642            If there are more than 20 of these errors, pcregrep is abandoned.
2644        (b) A binary zero was treated as data while matching, but terminated the
2645            output line if it was written out. This has been fixed: binary zeroes
2646            are now no different to any other data bytes.
2648        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2649            used to set a locale for matching. The --locale=xxxx long option has
2650            been added (no short equivalent) to specify a locale explicitly on the
2651            pcregrep command, overriding the environment variables.
2653        (d) When -B was used with -n, some line numbers in the output were one less
2654            than they should have been.
2656        (e) Added the -o (--only-matching) option.
2658        (f) If -A or -C was used with -c (count only), some lines of context were
2659            accidentally printed for the final match.
2661        (g) Added the -H (--with-filename) option.
2663        (h) The combination of options -rh failed to suppress file names for files
2664            that were found from directory arguments.
2666        (i) Added the -D (--devices) and -d (--directories) options.
2668        (j) Added the -F (--fixed-strings) option.
2670        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2672        (l) Added the --colo(u)r option.
2674        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2675            is not present by default.
2677     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2678        items such as (?R) or (?1), when the recursion could match a number of
2679        alternatives. If it matched one of the alternatives, but subsequently,
2680        outside the recursion, there was a failure, the code tried to back up into
2681        the recursion. However, because of the way PCRE is implemented, this is not
2682        possible, and the result was an incorrect result from the match.
2684        In order to prevent this happening, the specification of recursion has
2685        been changed so that all such subpatterns are automatically treated as
2686        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2688     4. I had overlooked the fact that, in some locales, there are characters for
2689        which isalpha() is true but neither isupper() nor islower() are true. In
2690        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2691        and ordfeminine) are like this. This affected the treatment of \w and \W
2692        when they appeared in character classes, but not when they appeared outside
2693        a character class. The bit map for "word" characters is now created
2694        separately from the results of isalnum() instead of just taking it from the
2695        upper, lower, and digit maps. (Plus the underscore character, of course.)
2697     5. The above bug also affected the handling of POSIX character classes such as
2698        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2699        permanent tables. Instead, the bit maps for such a class were previously
2700        created as the appropriate unions of the upper, lower, and digit bitmaps.
2701        Now they are created by subtraction from the [[:word:]] class, which has
2702        its own bitmap.
2704     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2705        It is created by subtracting the vertical space characters (\x09, \x0a,
2706        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2707        subtraction was done in the overall bitmap for a character class, meaning
2708        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2709        be recognized. This bug has been fixed.
2711     7. Patches from the folks at Google:
2713          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2714          real life, but is still worth protecting against".
2716          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2717          regular expressions".
2719          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2720          have it.
2722          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2723          "configure" and the latter not, in order to fix a problem somebody had
2724          with compiling the Arg class on HP-UX.
2726          (e) Improve the error-handling of the C++ wrapper a little bit.
2728          (f) New tests for checking recursion limiting.
2730     8. The pcre_memmove() function, which is used only if the environment does not
2731        have a standard memmove() function (and is therefore rarely compiled),
2732        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2733        returning a result (though PCRE never actually uses the result).
2735     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2736        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2737        returned instead of calling malloc() with an overflowing number that would
2738        most likely cause subsequent chaos.
2740    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2742    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2743        with this option is matched, the nmatch and pmatch options of regexec() are
2744        ignored.
2746    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2747        provided in case anyone wants to the the POSIX interface with UTF-8
2748        strings.
2750    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2751        C++ linking (needed for some HP-UX environments).
2753    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2754        (unused parameter) and in the pcre_printint() function (omitted "default"
2755        switch label when the default is to do nothing).
2757    15. Added some code to make it possible, when PCRE is compiled as a C++
2758        library, to replace subject pointers for pcre_exec() with a smart pointer
2759        class, thus making it possible to process discontinuous strings.
2761    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2762        much the same function. They were added by different people who were trying
2763        to make PCRE easy to compile on non-Unix systems. It has been suggested
2764        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2765        for compiling on Windows systems. I have therefore replaced it with
2766        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2767        defaults to "extern" for C or "extern C" for C++, which works fine on
2768        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2769        SCOPE with something explicit in config.h. In addition:
2771        (a) pcreposix.h still had just "extern" instead of either of these macros;
2772            I have replaced it with PCRE_DATA_SCOPE.
2774        (b) Functions such as _pcre_xclass(), which are internal to the library,
2775            but external in the C sense, all had PCRE_EXPORT in their definitions.
2776            This is apparently wrong for the Windows case, so I have removed it.
2777            (It makes no difference on Unix-like systems.)
2779    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2780        of recursive calls to match(). This is different to MATCH_LIMIT because
2781        that limits the total number of calls to match(), not all of which increase
2782        the depth of recursion. Limiting the recursion depth limits the amount of
2783        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2784        when PCRE is compiled, and changed at run time. A patch from Google adds
2785        this functionality to the C++ interface.
2787    18. Changes to the handling of Unicode character properties:
2789        (a) Updated the table to Unicode 4.1.0.
2791        (b) Recognize characters that are not in the table as "Cn" (undefined).
2793        (c) I revised the way the table is implemented to a much improved format
2794            which includes recognition of ranges. It now supports the ranges that
2795            are defined in UnicodeData.txt, and it also amalgamates other
2796            characters into ranges. This has reduced the number of entries in the
2797            table from around 16,000 to around 3,000, thus reducing its size
2798            considerably. I realized I did not need to use a tree structure after
2799            all - a binary chop search is just as efficient. Having reduced the
2800            number of entries, I extended their size from 6 bytes to 8 bytes to
2801            allow for more data.
2803        (d) Added support for Unicode script names via properties such as \p{Han}.
2805    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2806        matching that character.
2808    20. When matching a repeated Unicode property with a minimum greater than zero,
2809        (for example \pL{2,}), PCRE could look past the end of the subject if it
2810        reached it while seeking the minimum number of characters. This could
2811        happen only if some of the characters were more than one byte long, because
2812        there is a check for at least the minimum number of bytes.
2814    21. Refactored the implementation of \p and \P so as to be more general, to
2815        allow for more different types of property in future. This has changed the
2816        compiled form incompatibly. Anybody with saved compiled patterns that use
2817        \p or \P will have to recompile them.
2819    22. Added "Any" and "L&" to the supported property types.
2821    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2822        but give a compile time error if the value is greater than 0xff.
2824    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2825        accidentally not being installed or uninstalled.
2827    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2828        made were to insert the current release number. This seemed silly, because
2829        it made things harder for people building PCRE on systems that don't run
2830        "configure". I have turned pcre.h into a distributed file, no longer built
2831        by "configure", with the version identification directly included. There is
2832        no longer a pcre.h.in file.
2834        However, this change necessitated a change to the pcre-config script as
2835        well. It is built from pcre-config.in, and one of the substitutions was the
2836        release number. I have updated configure.ac so that ./configure now finds
2837        the release number by grepping pcre.h.
2839    26. Added the ability to run the tests under valgrind.
2842    Version 6.4 05-Sep-05
2843    ---------------------
2845     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2846        "--" to be printed when multiple files were scanned, even when none of the
2847        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2848        consider it to be a bug, and have restored the previous behaviour.
2850     2. A couple of code tidies to get rid of compiler warnings.
2852     3. The pcretest program used to cheat by referring to symbols in the library
2853        whose names begin with _pcre_. These are internal symbols that are not
2854        really supposed to be visible externally, and in some environments it is
2855        possible to suppress them. The cheating is now confined to including
2856        certain files from the library's source, which is a bit cleaner.
2858     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2859        file's purpose clearer.
2861     5. Reorganized pcre_ucp_findchar().
2864    Version 6.3 15-Aug-05
2865    ---------------------
2867     1. The file libpcre.pc.in did not have general read permission in the tarball.
2869     2. There were some problems when building without C++ support:
2871        (a) If C++ support was not built, "make install" and "make test" still
2872            tried to test it.
2874        (b) There were problems when the value of CXX was explicitly set. Some
2875            changes have been made to try to fix these, and ...
2877        (c) --disable-cpp can now be used to explicitly disable C++ support.
2879        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2880            backslash in a target when C++ was disabled. This confuses some
2881            versions of "make", apparently. Using an intermediate variable solves
2882            this. (Same for CPP_LOBJ.)
2884     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2885        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2886        necessary on certain architectures.
2888     4. Added a setting of -export-symbols-regex to the link command to remove
2889        those symbols that are exported in the C sense, but actually are local
2890        within the library, and not documented. Their names all begin with
2891        "_pcre_". This is not a perfect job, because (a) we have to except some
2892        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2893        available (and never for static libraries). I have made a note to try to
2894        find a way round (a) in the future.
2897    Version 6.2 01-Aug-05
2898    ---------------------
2900     1. There was no test for integer overflow of quantifier values. A construction
2901        such as {1111111111111111} would give undefined results. What is worse, if
2902        a minimum quantifier for a parenthesized subpattern overflowed and became
2903        negative, the calculation of the memory size went wrong. This could have
2904        led to memory overwriting.
2906     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2908     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2909        operating environments where this matters.
2911     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2912        PCRE options from within the C++ wrapper.
2914     5. Named capturing subpatterns were not being correctly counted when a pattern
2915        was compiled. This caused two problems: (a) If there were more than 100
2916        such subpatterns, the calculation of the memory needed for the whole
2917        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2918        back references of the form \12, where the number was greater than 9, were
2919        not recognized as back references, even though there were sufficient
2920        previous subpatterns.
2922     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2923        versions of gcc, e.g. 2.95.4.
2926    Version 6.1 21-Jun-05
2927    ---------------------
2929     1. There was one reference to the variable "posix" in pcretest.c that was not
2930        surrounded by "#if !defined NOPOSIX".
2932     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2933        the cross-check on the old pcre_info() function, for the benefit of the
2934        cut-down version of PCRE that is currently imported into Exim.
2936     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2937        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2938        patterns that start (?i) but I don't think that matters. The use of (?i) is
2939        just an example; this all applies to the other options as well.
2941     4. Since libtool seems to echo the compile commands it is issuing, the output
2942        from "make" can be reduced a bit by putting "@" in front of each libtool
2943        compile command.
2945     5. Patch from the folks at Google for configure.in to be a bit more thorough
2946        in checking for a suitable C++ installation before trying to compile the
2947        C++ stuff. This should fix a reported problem when a compiler was present,
2948        but no suitable headers.
2950     6. The man pages all had just "PCRE" as their title. I have changed them to
2951        be the relevant file name. I have also arranged that these names are
2952        retained in the file doc/pcre.txt, which is a concatenation in text format
2953        of all the man pages except the little individual ones for each function.
2955     7. The NON-UNIX-USE file had not been updated for the different set of source
2956        files that come with release 6. I also added a few comments about the C++
2957        wrapper.
2960    Version 6.0 07-Jun-05
2961    ---------------------
2963     1. Some minor internal re-organization to help with my DFA experiments.
2965     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2966        didn't matter for the library itself when fully configured, but did matter
2967        when compiling without UCP support, or within Exim, where the ucp files are
2968        not imported.
2970     3. Refactoring of the library code to split up the various functions into
2971        different source modules. The addition of the new DFA matching code (see
2972        below) to a single monolithic source would have made it really too
2973        unwieldy, quite apart from causing all the code to be include in a
2974        statically linked application, when only some functions are used. This is
2975        relevant even without the DFA addition now that patterns can be compiled in
2976        one application and matched in another.
2978        The downside of splitting up is that there have to be some external
2979        functions and data tables that are used internally in different modules of
2980        the library but which are not part of the API. These have all had their
2981        names changed to start with "_pcre_" so that they are unlikely to clash
2982        with other external names.
2984     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2985        a different (DFA) algorithm. Although it is slower than the original
2986        function, it does have some advantages for certain types of matching
2987        problem.
2989     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2990        including restarting after a partial match.
2992     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2993        defined when compiling for Windows was sent to me. I have put it into the
2994        code, though I have no means of testing or verifying it.
2996     7. Added the pcre_refcount() auxiliary function.
2998     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2999        match before or at the first newline in the subject string. In pcretest,
3000        the /f option on a pattern can be used to set this.
3002     9. A repeated \w when used in UTF-8 mode with characters greater than 256
3003        would behave wrongly. This has been present in PCRE since release 4.0.
3005    10. A number of changes to the pcregrep command:
3007        (a) Refactored how -x works; insert ^(...)$ instead of setting
3008            PCRE_ANCHORED and checking the length, in preparation for adding
3009            something similar for -w.
3011        (b) Added the -w (match as a word) option.
3013        (c) Refactored the way lines are read and buffered so as to have more
3014            than one at a time available.
3016        (d) Implemented a pcregrep test script.
3018        (e) Added the -M (multiline match) option. This allows patterns to match
3019            over several lines of the subject. The buffering ensures that at least
3020            8K, or the rest of the document (whichever is the shorter) is available
3021            for matching (and similarly the previous 8K for lookbehind assertions).
3023        (f) Changed the --help output so that it now says
3025              -w, --word-regex(p)
3027            instead of two lines, one with "regex" and the other with "regexp"
3028            because that confused at least one person since the short forms are the
3029            same. (This required a bit of code, as the output is generated
3030            automatically from a table. It wasn't just a text change.)
3032        (g) -- can be used to terminate pcregrep options if the next thing isn't an
3033            option but starts with a hyphen. Could be a pattern or a path name
3034            starting with a hyphen, for instance.
3036        (h) "-" can be given as a file name to represent stdin.
3038        (i) When file names are being printed, "(standard input)" is used for
3039            the standard input, for compatibility with GNU grep. Previously
3040            "<stdin>" was used.
3042        (j) The option --label=xxx can be used to supply a name to be used for
3043            stdin when file names are being printed. There is no short form.
3045        (k) Re-factored the options decoding logic because we are going to add
3046            two more options that take data. Such options can now be given in four
3047            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3049        (l) Added the -A, -B, and -C options for requesting that lines of context
3050            around matches be printed.
3052        (m) Added the -L option to print the names of files that do not contain
3053            any matching lines, that is, the complement of -l.
3055        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3056            continue to scan other files.
3058        (o) The -s option was incorrectly implemented. For compatibility with other
3059            greps, it now suppresses the error message for a non-existent or non-
3060            accessible file (but not the return code). There is a new option called
3061            -q that suppresses the output of matching lines, which was what -s was
3062            previously doing.
3064        (p) Added --include and --exclude options to specify files for inclusion
3065            and exclusion when recursing.
3067    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3068        Hopefully, it now does.
3070    12. Missing cast in pcre_study().
3072    13. Added an "uninstall" target to the makefile.
3074    14. Replaced "extern" in the function prototypes in Makefile.in with
3075        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3076        world, but is set differently for Windows.
3078    15. Added a second compiling function called pcre_compile2(). The only
3079        difference is that it has an extra argument, which is a pointer to an
3080        integer error code. When there is a compile-time failure, this is set
3081        non-zero, in addition to the error test pointer being set to point to an
3082        error message. The new argument may be NULL if no error number is required
3083        (but then you may as well call pcre_compile(), which is now just a
3084        wrapper). This facility is provided because some applications need a
3085        numeric error indication, but it has also enabled me to tidy up the way
3086        compile-time errors are handled in the POSIX wrapper.
3088    16. Added VPATH=.libs to the makefile; this should help when building with one
3089        prefix path and installing with another. (Or so I'm told by someone who
3090        knows more about this stuff than I do.)
3092    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3093        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3094        match everything, including newlines. This is not POSIX-compatible, but
3095        somebody wanted the feature. From pcretest it can be activated by using
3096        both the P and the s flags.
3098    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3100    19. libpcre.pc was being incorrectly installed as executable.
3102    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3103        it now also looks for '\r' so that it will work unmodified on Windows.
3105    21. Added Google's contributed C++ wrapper to the distribution.
3107    22. Added some untidy missing memory free() calls in pcretest, to keep
3108        Electric Fence happy when testing.
3112    Version 5.0 13-Sep-04
3113    ---------------------
3115     1. Internal change: literal characters are no longer packed up into items
3116        containing multiple characters in a single byte-string. Each character
3117        is now matched using a separate opcode. However, there may be more than one
3118        byte in the character in UTF-8 mode.
3120     2. The pcre_callout_block structure has two new fields: pattern_position and
3121        next_item_length. These contain the offset in the pattern to the next match
3122        item, and its length, respectively.
3124     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3125        insertion of callouts before each pattern item. Added the /C option to
3126        pcretest to make use of this.
3128     4. On the advice of a Windows user, the lines
3130          #if defined(_WIN32) || defined(WIN32)
3131          _setmode( _fileno( stdout ), 0x8000 );
3132          #endif  /* defined(_WIN32) || defined(WIN32) */
3134        have been added to the source of pcretest. This apparently does useful
3135        magic in relation to line terminators.
3137     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3138        for the benefit of those environments where the "b" makes a difference.
3140     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3141        to know about it. I have put a hack into configure.in that adds in code
3142        to set GCC=yes if CC=icc. This seems to end up at a point in the
3143        generated configure script that is early enough to affect the setting of
3144        compiler options, which is what is needed, but I have no means of testing
3145        whether it really works. (The user who reported this had patched the
3146        generated configure script, which of course I cannot do.)
3148        LATER: After change 22 below (new libtool files), the configure script
3149        seems to know about icc (and also ecc). Therefore, I have commented out
3150        this hack in configure.in.
3152     7. Added support for pkg-config (2 patches were sent in).
3154     8. Negated POSIX character classes that used a combination of internal tables
3155        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3156        [[:^ascii]]. Typically, they would match almost any characters. The other
3157        POSIX classes were not broken in this way.
3159     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3160        to find the match, as PCRE was deluded into thinking that the match had to
3161        start at the start point or following a newline. The same bug applied to
3162        patterns with negative forward assertions or any backward assertions
3163        preceding ".*" at the start, unless the pattern required a fixed first
3164        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3166    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3167        starting at the last subject character, bytes beyond the end of the subject
3168        string were read.
3170    11. Renamed the variable "class" as "classbits" to make life easier for C++
3171        users. (Previously there was a macro definition, but it apparently wasn't
3172        enough.)
3174    12. Added the new field "tables" to the extra data so that tables can be passed
3175        in at exec time, or the internal tables can be re-selected. This allows
3176        a compiled regex to be saved and re-used at a later time by a different
3177        program that might have everything at different addresses.
3179    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3180        -R library as well as a -L library.
3182    14. The debugging options of pcretest (-d on the command line or D on a
3183        pattern) showed incorrect output for anything following an extended class
3184        that contained multibyte characters and which was followed by a quantifier.
3186    15. Added optional support for general category Unicode character properties
3187        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3188        support. It adds about 90K to the size of the library. The meanings of the
3189        inbuilt class escapes such as \d and \s have NOT been changed.
3191    16. Updated pcredemo.c to include calls to free() to release the memory for the
3192        compiled pattern.
3194    17. The generated file chartables.c was being created in the source directory
3195        instead of in the building directory. This caused the build to fail if the
3196        source directory was different from the building directory, and was
3197        read-only.
3199    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3200        file. No doubt somebody will tell me if they don't make sense... Also added
3201        Dan Mooney's comments about building on OpenVMS.
3203    19. Added support for partial matching via the PCRE_PARTIAL option for
3204        pcre_exec() and the \P data escape in pcretest.
3206    20. Extended pcretest with 3 new pattern features:
3208        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3209              write the compiled pattern to the file whose name is "rest-of-line".
3210              This is a straight binary dump of the data, with the saved pointer to
3211              the character tables forced to be NULL. The study data, if any, is
3212              written too. After writing, pcretest reads a new pattern.
3214        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3215              compiled pattern from the given file. There must not be any
3216              occurrences of "<" in the file name (pretty unlikely); if there are,
3217              pcretest will instead treat the initial "<" as a pattern delimiter.
3218              After reading in the pattern, pcretest goes on to read data lines as
3219              usual.
3221        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3222              and 16-bit fields in a compiled pattern, to simulate a pattern that
3223              was compiled on a host of opposite endianness.
3225    21. The pcre-exec() function can now cope with patterns that were compiled on
3226        hosts of opposite endianness, with this restriction:
3228          As for any compiled expression that is saved and used later, the tables
3229          pointer field cannot be preserved; the extra_data field in the arguments
3230          to pcre_exec() should be used to pass in a tables address if a value
3231          other than the default internal tables were used at compile time.
3233    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3234        now diagnosed as an error. Previously, most of the time, a negative number
3235        would have been treated as zero, but if in addition "ovector" was passed as
3236        NULL, a crash could occur.
3238    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3239        new versions from the libtool 1.5 distribution (the last one is a copy of
3240        a file called libtool.m4). This seems to have fixed the need to patch
3241        "configure" to support Darwin 1.3 (which I used to do). However, I still
3242        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3243        workstation).
3245    24. Changed the PCRE licence to be the more standard "BSD" licence.
3248    Version 4.5 01-Dec-03
3249    ---------------------
3251     1. There has been some re-arrangement of the code for the match() function so
3252        that it can be compiled in a version that does not call itself recursively.
3253        Instead, it keeps those local variables that need separate instances for
3254        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3255        needs to "recurse". Keeping track of where control must go is done by means
3256        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3257        hide most of the details from the main code, and operates only if
3258        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3259        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3260        operating.
3262        To make it easier for callers to provide specially tailored get/free
3263        functions for this usage, two new functions, pcre_stack_malloc, and
3264        pcre_stack_free, are used. They are always called in strict stacking order,
3265        and the size of block requested is always the same.
3267        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3268        PCRE has been compiled to use the stack or the heap for recursion. The
3269        -C option of pcretest uses this to show which version is compiled.
3271        A new data escape \S, is added to pcretest; it causes the amounts of store
3272        obtained and freed by both kinds of malloc/free at match time to be added
3273        to the output.
3275     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3276        what's available on my current Linux desktop machine.
3278     3. When matching a UTF-8 string, the test for a valid string at the start has
3279        been extended. If start_offset is not zero, PCRE now checks that it points
3280        to a byte that is the start of a UTF-8 character. If not, it returns
3281        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3282        this is necessary because there may be backward assertions in the pattern.
3283        When matching the same subject several times, it may save resources to use
3284        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3286     4. The code for checking the validity of UTF-8 strings has been tightened so
3287        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3288        containing "overlong sequences".
3290     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3291        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3292        should have been "&", but it just so happened that all the cases this let
3293        through by mistake were picked up later in the function.
3295     6. I had used a variable called "isblank" - this is a C99 function, causing
3296        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3298     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3299        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3300        so the version is shown; (c) stop "make test" echoing "./RunTest".
3302     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3304     9. The prototype for memmove() for systems that don't have it was using
3305        size_t, but the inclusion of the header that defines size_t was later. I've
3306        moved the #includes for the C headers earlier to avoid this.
3308    10. Added some adjustments to the code to make it easier to compiler on certain
3309        special systems:
3311          (a) Some "const" qualifiers were missing.
3312          (b) Added the macro EXPORT before all exported functions; by default this
3313              is defined to be empty.
3314          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3315              that it reads its output file name as an argument instead of writing
3316              to the standard output and assuming this can be redirected.
3318    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3319        class containing characters with values greater than 255, PCRE compilation
3320        went into a loop.
3322    12. A recursive reference to a subpattern that was within another subpattern
3323        that had a minimum quantifier of zero caused PCRE to crash. For example,
3324        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3325        recursion. If the recursively-called subpattern itself had a zero repeat,
3326        that was OK.
3328    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3329        buffer into which it was copied (for escape processing) was still set at
3330        1024, so long lines caused crashes.
3332    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3333        "internal error: code overflow...". This applied to any character class
3334        that was followed by a possessive quantifier.
3336    15. Modified the Makefile to add libpcre.la as a prerequisite for
3337        libpcreposix.la because I was told this is needed for a parallel build to
3338        work.
3340    16. If a pattern that contained .* following optional items at the start was
3341        studied, the wrong optimizing data was generated, leading to matching
3342        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3343        matching string must start with a or b or c. The correct conclusion for
3344        this pattern is that a match can start with any character.
3347    Version 4.4 13-Aug-03
3348    ---------------------
3350     1. In UTF-8 mode, a character class containing characters with values between
3351        127 and 255 was not handled correctly if the compiled pattern was studied.
3352        In fixing this, I have also improved the studying algorithm for such
3353        classes (slightly).
3355     2. Three internal functions had redundant arguments passed to them. Removal
3356        might give a very teeny performance improvement.
3358     3. Documentation bug: the value of the capture_top field in a callout is *one
3359        more than* the number of the hightest numbered captured substring.
3361     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3362        in incorrectly linking with a previously installed version. They now link
3363        explicitly with libpcre.la.
3365     5. configure.in no longer needs to recognize Cygwin specially.
3367     6. A problem in pcre.in for Windows platforms is fixed.
3369     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3370        pcretest, it used to include the size of the study block as part of its
3371        output. Unfortunately, the structure contains a field that has a different
3372        size on different hardware architectures. This meant that the tests that
3373        showed this size failed. As the block is currently always of a fixed size,
3374        this information isn't actually particularly useful in pcretest output, so
3375        I have just removed it.