ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 5 by nigel, Sat Feb 24 21:38:05 2007 UTC revision 1033 by ph10, Mon Sep 10 11:02:48 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.32
5    ------------
7    1.  Improved JIT compiler optimizations for first character search and single
8        character iterators.
10    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
11        Patch by Daniel Richard G.
13    3.  Single character iterator optimizations in the JIT compiler.
15    4.  Improved JIT compiler optimizations for character ranges.
17    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
18        Reported by Giuseppe D'Angelo.
20    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
21        of a line, was being set incorrectly in cases where .* appeared inside
22        atomic brackets at the start of a pattern, or where there was a subsequent
23        *PRUNE or *SKIP.
25    7.  Improved instruction cache flush for POWER/PowerPC.
26        Patch by Daniel Richard G.
28    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
29        grep:
31        (a) There is now no limit to the number of patterns to be matched.
33        (b) An error is given if a pattern is too long.
35        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
36            are now supported.
38        (d) --exclude-from and --include-from (multiple use) have been added.
40        (e) Exclusions and inclusions now apply to all files and directories, not
41            just to those obtained from scanning a directory recursively.
43        (f) Multiple uses of -f and --file-list are now supported.
45        (g) In a Windows environment, the default for -d has been changed from
46            "read" (the GNU grep default) to "skip", because otherwise the presence
47            of a directory in the file list provokes an error.
49        (h) The documentation has been revised and clarified in places.
51    9.  Improve the matching speed of capturing brackets.
53    10. Changed the meaning of \X so that it now matches a Unicode extended
54        grapheme cluster.
56    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
57        out POSIX threads when JIT support is configured.
59    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
61    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
62        structure could go wrong in environments where size_t is not the same size
63        as int.
65    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
66        set.
68    15. The EBCDIC support had decayed; later updates to the code had included
69        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
70        general tidy up of EBCDIC-related issues, and the documentation was also
71        not quite right. There is now a test that can be run on ASCII systems to
72        check some of the EBCDIC-related things (but is it not a full test).
76    Version 8.31 06-July-2012
77    -------------------------
79    1.  Fixing a wrong JIT test case and some compiler warnings.
81    2.  Removed a bashism from the RunTest script.
83    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
84        to unsigned type, result still unsigned" that was given by an MS compiler
85        on encountering the code "-sizeof(xxx)".
87    4.  Partial matching support is added to the JIT compiler.
89    5.  Fixed several bugs concerned with partial matching of items that consist
90        of more than one character:
92        (a) /^(..)\1/ did not partially match "aba" because checking references was
93            done on an "all or nothing" basis. This also applied to repeated
94            references.
96        (b) \R did not give a hard partial match if \r was found at the end of the
97            subject.
99        (c) \X did not give a hard partial match after matching one or more
100            characters at the end of the subject.
102        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
103            a partial match for the string "\r".
105        (e) When newline was set to CRLF, the metacharacter "." did not recognize
106            a partial match for a CR character at the end of the subject string.
108    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
109        running pcretest, the text "(JIT)" added to the output whenever JIT is
110        actually used to run the match.
112    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
113        or /S+[+] with a digit between 1 and 7.
115    8.  OP_NOT now supports any UTF character not just single-byte ones.
117    9.  (*MARK) control verb is now supported by the JIT compiler.
119    10. The command "./RunTest list" lists the available tests without actually
120        running any of them. (Because I keep forgetting what they all are.)
124    12. Applied a (slightly modified) user-supplied patch that improves performance
125        when the heap is used for recursion (compiled with --disable-stack-for-
126        recursion). Instead of malloc and free for each heap frame each time a
127        logical recursion happens, frames are retained on a chain and re-used where
128        possible. This sometimes gives as much as 30% improvement.
130    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
131        call.
133    14. As documented, (*COMMIT) is now confined to within a positive assertion.
135    15. It is now possible to link pcretest with libedit as an alternative to
136        libreadline.
138    16. (*COMMIT) control verb is now supported by the JIT compiler.
140    17. The Unicode data tables have been updated to Unicode 6.1.0.
142    18. Added --file-list option to pcregrep.
144    19. Added binary file support to pcregrep, including the -a, --binary-files,
145        -I, and --text options.
147    20. The madvise function is renamed for posix_madvise for QNX compatibility
148        reasons. Fixed by Giuseppe D'Angelo.
150    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
152    22. Changed the option for creating symbolic links for 16-bit man pages from
153        -s to -sf so that re-installing does not cause issues.
155    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
157    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
158        matches in certain environments (the workspace was not being correctly
159        retained). Also added to pcre_dfa_exec() a simple plausibility check on
160        some of the workspace data at the beginning of a restart.
162    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
163        was not doing so when it should - probably a typo introduced by SVN 528
164        (change 8.10/14).
166    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
167        \w+ when the character tables indicated that \x{c4} was a word character.
168        There were several related cases, all because the tests for doing a table
169        lookup were testing for characters less than 127 instead of 255.
171    27. If a pattern contains capturing parentheses that are not used in a match,
172        their slots in the ovector are set to -1. For those that are higher than
173        any matched groups, this happens at the end of processing. In the case when
174        there were back references that the ovector was too small to contain
175        (causing temporary malloc'd memory to be used during matching), and the
176        highest capturing number was not used, memory off the end of the ovector
177        was incorrectly being set to -1. (It was using the size of the temporary
178        memory instead of the true size.)
180    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
181        ovector size, it uses memory at the end of the block that it has got.
183    29. Check for an overlong MARK name and give an error at compile time. The
184        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
186    30. JIT compiler update.
188    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
189        Rill for the patch.
191    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
193    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
195    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
196        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
197        was enabled.
199    35. Improve JIT code generation for greedy plus quantifier.
201    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
202        1 to "aa" instead of to an empty string. The bug affected repeated groups
203        that could potentially match an empty string.
205    37. Optimizing single character iterators in JIT.
207    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
208        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
209        codepoints that are too big for the mode are faulted, and in a UTF mode,
210        disallowed codepoints are also faulted.
212    39. If PCRE was compiled with UTF support, in three places in the DFA
213        matcher there was code that should only have been obeyed in UTF mode, but
214        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
215        processing when bytes with values greater than 127 were present. In 16-bit
216        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
217        both cases the values are those that cannot be the first data item in a UTF
218        character. The three items that might have provoked this were recursions,
219        possessively repeated groups, and atomic groups.
221    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
222        and pcregrep, because some OS require shared objects to be explicitly
223        passed to ld, causing the link step to fail if they are not.
225    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
226        mode, patterns that started with \h* or \R* might be incorrectly matched.
229    Version 8.30 04-February-2012
230    -----------------------------
232    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
233        name is defined in ctype.h.
235    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
236        only in quite long subpatterns.
238    3.  Removed the function pcre_info(), which has been obsolete and deprecated
239        since it was replaced by pcre_fullinfo() in February 2000.
241    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
242        match a (*MARK), and the match failed at the start of the subject, a
243        reference to memory before the start of the subject could occur. This bug
244        was introduced by fix 17 of release 8.21.
246    5.  A reference to an unset group with zero minimum repetition was giving
247        totally wrong answers (in non-JavaScript-compatibility mode). For example,
248        /(another)?(\1?)test/ matched against "hello world test". This bug was
249        introduced in release 8.13.
251    6.  Add support for 16-bit character strings (a large amount of work involving
252        many changes and refactorings).
254    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
255        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
256        from a file.
258    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
259        rounding is not applied in this particular case).
261    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
262        if they appear, or are escaped, in patterns.
264    10. Get rid of a number of -Wunused-but-set-variable warnings.
266    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
267        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
268        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
269        also returns the mark "x". This bug applied to capturing parentheses,
270        non-capturing parentheses, and atomic parentheses. It also applied to some
271        assertions.
273    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
274        information out of configure.ac instead of relying on pcre.h.generic, which
275        is not stored in the repository.
277    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
278        -lreadline.
280    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
282    15. Applied Graycode's patch to put the top-level frame on the stack rather
283        than the heap when not using the stack for recursion. This gives a
284        performance improvement in many cases when recursion is not deep.
286    16. Experimental code added to "pcretest -C" to output the stack frame size.
289    Version 8.21 12-Dec-2011
290    ------------------------
292    1.  Updating the JIT compiler.
294    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
295        are added as well.
297    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
298        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
299        calling _pcre_jit_exec. Some extra comments are added.
301    4.  (*MARK) settings inside atomic groups that do not contain any capturing
302        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
303        was introduced by change 18 for 8.20.
305    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
306        ECMA-262 standard.
308    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
309        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
310        This bug was probably introduced by change 9 of 8.13.
312    7.  While fixing 6 above, I noticed that a number of other items were being
313        incorrectly rejected as "not fixed length". This arose partly because newer
314        opcodes had not been added to the fixed-length checking code. I have (a)
315        corrected the bug and added tests for these items, and (b) arranged for an
316        error to occur if an unknown opcode is encountered while checking for fixed
317        length instead of just assuming "not fixed length". The items that were
318        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
319        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
320        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
322    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
323        being incorrectly compiled and would have given unpredicatble results.
325    9.  A possessively repeated subpattern with minimum repeat count greater than
326        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
327        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
328        the first (A) could occur when it should not.
330    10. Add a cast and remove a redundant test from the code.
332    11. JIT should use pcre_malloc/pcre_free for allocation.
334    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
335        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
336        is anything other than /usr, -L is still shown).
338    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
340    14. Perl does not support \N without a following name in a [] class; PCRE now
341        also gives an error.
343    15. If a forward reference was repeated with an upper limit of around 2000,
344        it caused the error "internal error: overran compiling workspace". The
345        maximum number of forward references (including repeats) was limited by the
346        internal workspace, and dependent on the LINK_SIZE. The code has been
347        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
348        the default depends on LINK_SIZE. There is a new upper limit (for safety)
349        of around 200,000 forward references. While doing this, I also speeded up
350        the filling in of repeated forward references.
352    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
353        incorrectly expecting the subject to contain another "a" after the start.
355    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
356        in the match, the SKIP should be ignored. This was not happening; instead
357        the SKIP was being treated as NOMATCH. For patterns such as
358        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
359        tested.
361    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
362        now much more compatible with Perl, in particular in cases where the result
363        is a non-match for a non-anchored pattern. For example, if
364        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
365        "m", where previously it did not return a name. A side effect of this
366        change is that for partial matches, the last encountered mark name is
367        returned, as for non matches. A number of tests that were previously not
368        Perl-compatible have been moved into the Perl-compatible test files. The
369        refactoring has had the pleasing side effect of removing one argument from
370        the match() function, thus reducing its stack requirements.
372    19. If the /S+ option was used in pcretest to study a pattern using JIT,
373        subsequent uses of /S (without +) incorrectly behaved like /S+.
375    21. Retrieve executable code size support for the JIT compiler and fixing
376        some warnings.
378    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
379        not work when the shorter character appeared right at the end of the
380        subject string.
382    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
383        systems.
385    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
386        output it when the /M option is used in pcretest.
388    25. The CheckMan script was not being included in the distribution. Also, added
389        an explicit "perl" to run Perl scripts from the PrepareRelease script
390        because this is reportedly needed in Windows.
392    26. If study data was being save in a file and studying had not found a set of
393        "starts with" bytes for the pattern, the data written to the file (though
394        never used) was taken from uninitialized memory and so caused valgrind to
395        complain.
397    27. Updated RunTest.bat as provided by Sheri Pierce.
399    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
401    29. Computation of memory usage for the table of capturing group names was
402        giving an unnecessarily large value.
405    Version 8.20 21-Oct-2011
406    ------------------------
408    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
409        a POSIX class. After further experiments with Perl, which convinced me that
410        Perl has bugs and confusions, a closing square bracket is no longer allowed
411        in a POSIX name. This bug also affected patterns with classes that started
412        with full stops.
414    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
415        captured substring, but while checking the failing first alternative,
416        substring 1 is temporarily captured. If the output vector supplied to
417        pcre_exec() was not big enough for this capture, the yield of the function
418        was still zero ("insufficient space for captured substrings"). This cannot
419        be totally fixed without adding another stack variable, which seems a lot
420        of expense for a edge case. However, I have improved the situation in cases
421        such as /(a)(b)x|abc/ matched against "abc", where the return code
422        indicates that fewer than the maximum number of slots in the ovector have
423        been set.
425    3.  Related to (2) above: when there are more back references in a pattern than
426        slots in the output vector, pcre_exec() uses temporary memory during
427        matching, and copies in the captures as far as possible afterwards. It was
428        using the entire output vector, but this conflicts with the specification
429        that only 2/3 is used for passing back captured substrings. Now it uses
430        only the first 2/3, for compatibility. This is, of course, another edge
431        case.
433    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
434        main code base, and can be used by building with --enable-jit. When this is
435        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
436        runtime --no-jit option is given.
438    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
439        ovector, the return from the function was zero, implying that there were
440        other matches that did not fit. The correct "exactly full" value is now
441        returned.
443    6.  If a subpattern that was called recursively or as a subroutine contained
444        (*PRUNE) or any other control that caused it to give a non-standard return,
445        invalid errors such as "Error -26 (nested recursion at the same subject
446        position)" or even infinite loops could occur.
448    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
449        computing the minimum length on reaching *ACCEPT, and so ended up with the
450        wrong value of 1 rather than 0. Further investigation indicates that
451        computing a minimum subject length in the presence of *ACCEPT is difficult
452        (think back references, subroutine calls), and so I have changed the code
453        so that no minimum is registered for a pattern that contains *ACCEPT.
455    8.  If (*THEN) was present in the first (true) branch of a conditional group,
456        it was not handled as intended. [But see 16 below.]
458    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
459        Sheri Pierce.
461    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
462        the first byte in a match must be "a".
464    11. Change 17 for 8.13 increased the recursion depth for patterns like
465        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
466        pattern contains any instances of (*THEN). If it does not, the old
467        optimizations are restored. It would be nice to do this on a per-group
468        basis, but at the moment that is not feasible.
470    12. In some environments, the output of pcretest -C is CRLF terminated. This
471        broke RunTest's code that checks for the link size. A single white space
472        character after the value is now allowed for.
474    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
475        For "fr", it uses the Windows-specific input and output files.
477    14. If (*THEN) appeared in a group that was called recursively or as a
478        subroutine, it did not work as intended. [But see next item.]
480    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
481        pattern fragments (but not containing any | characters). If A and B are
482        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
483        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
484        D. In other words, Perl considers parentheses that do not contain any |
485        characters to be part of a surrounding alternative, whereas PCRE was
486        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
487        differently. PCRE now behaves in the same way as Perl, except in the case
488        of subroutine/recursion calls such as (?1) which have in any case always
489        been different (but PCRE had them first :-).
491    16. Related to 15 above: Perl does not treat the | in a conditional group as
492        creating alternatives. Such a group is treated in the same way as an
493        ordinary group without any | characters when processing (*THEN). PCRE has
494        been changed to match Perl's behaviour.
496    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
497        RunGrepTest script failed.
499    18. Change 22 for version 13 caused atomic groups to use more stack. This is
500        inevitable for groups that contain captures, but it can lead to a lot of
501        stack use in large patterns. The old behaviour has been restored for atomic
502        groups that do not contain any capturing parentheses.
504    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
505        suppress the check for a minimum subject length at run time. (If it was
506        given to pcre_exec() or pcre_dfa_exec() it did work.)
508    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
509        fail to work when decoding hex characters in data strings in EBCDIC
510        environments.
512    21. It appears that in at least one Mac OS environment, the isxdigit() function
513        is implemented as a macro that evaluates to its argument more than once,
514        contravening the C 90 Standard (I haven't checked a later standard). There
515        was an instance in pcretest which caused it to go wrong when processing
516        \x{...} escapes in subject strings. The has been rewritten to avoid using
517        things like p++ in the argument of isxdigit().
520    Version 8.13 16-Aug-2011
521    ------------------------
523    1.  The Unicode data tables have been updated to Unicode 6.0.0.
525    2.  Two minor typos in pcre_internal.h have been fixed.
527    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
528        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
529        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
531    4.  There were a number of related bugs in the code for matching backrefences
532        caselessly in UTF-8 mode when codes for the characters concerned were
533        different numbers of bytes. For example, U+023A and U+2C65 are an upper
534        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
535        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
536        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
537        2-byte code at the end of the subject (it thought there wasn't enough data
538        left).
540    5.  Comprehensive information about what went wrong is now returned by
541        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
542        as the output vector has at least 2 elements. The offset of the start of
543        the failing character and a reason code are placed in the vector.
545    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
546        now returned is for the first byte of the failing character, instead of the
547        last byte inspected. This is an incompatible change, but I hope it is small
548        enough not to be a problem. It makes the returned offset consistent with
549        pcre_exec() and pcre_dfa_exec().
551    7.  pcretest now gives a text phrase as well as the error number when
552        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
553        failure, the offset and reason code are output.
555    8.  When \R was used with a maximizing quantifier it failed to skip backwards
556        over a \r\n pair if the subsequent match failed. Instead, it just skipped
557        back over a single character (\n). This seems wrong (because it treated the
558        two characters as a single entity when going forwards), conflicts with the
559        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
560        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
561        has been changed.
563    9.  Some internal refactoring has changed the processing so that the handling
564        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
565        time (the PCRE_DOTALL option was changed this way some time ago: version
566        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
567        which was always a bit of a fudge. It also means that there is one less
568        argument for the match() function, which reduces its stack requirements
569        slightly. This change also fixes an incompatibility with Perl: the pattern
570        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
572    10. More internal refactoring has drastically reduced the number of recursive
573        calls to match() for possessively repeated groups such as (abc)++ when
574        using pcre_exec().
576    11. While implementing 10, a number of bugs in the handling of groups were
577        discovered and fixed:
579        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
580        (a|)*(?1) gave a compile-time internal error.
581        ((a|)+)+  did not notice that the outer group could match an empty string.
582        (^a|^)+   was not marked as anchored.
583        (.*a|.*)+ was not marked as matching at start or after a newline.
585    12. Yet more internal refactoring has removed another argument from the match()
586        function. Special calls to this function are now indicated by setting a
587        value in a variable in the "match data" data block.
589    13. Be more explicit in pcre_study() instead of relying on "default" for
590        opcodes that mean there is no starting character; this means that when new
591        ones are added and accidentally left out of pcre_study(), testing should
592        pick them up.
594    14. The -s option of pcretest has been documented for ages as being an old
595        synonym of -m (show memory usage). I have changed it to mean "force study
596        for every regex", that is, assume /S for every regex. This is similar to -i
597        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
598        using it. It makes it easier to run collections of tests with and without
599        study enabled, and thereby test pcre_study() more easily. All the standard
600        tests are now run with and without -s (but some patterns can be marked as
601        "never study" - see 20 below).
603    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
604        restoration of the capturing data to the outer values was not happening
605        correctly.
607    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
608        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
609        pattern had matched an empty string, and so incorrectly returned a no
610        match.
612    17. There was optimizing code for the last branch of non-capturing parentheses,
613        and also for the obeyed branch of a conditional subexpression, which used
614        tail recursion to cut down on stack usage. Unfortunately, now that there is
615        the possibility of (*THEN) occurring in these branches, tail recursion is
616        no longer possible because the return has to be checked for (*THEN). These
617        two optimizations have therefore been removed. [But see 8.20/11 above.]
619    18. If a pattern containing \R was studied, it was assumed that \R always
620        matched two bytes, thus causing the minimum subject length to be
621        incorrectly computed because \R can also match just one byte.
623    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
624        was incorrectly computed.
626    20. If /S is present twice on a test pattern in pcretest input, it now
627        *disables* studying, thereby overriding the use of -s on the command line
628        (see 14 above). This is necessary for one or two tests to keep the output
629        identical in both cases.
631    21. When (*ACCEPT) was used in an assertion that matched an empty string and
632        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
634    22. When an atomic group that contained a capturing parenthesis was
635        successfully matched, but the branch in which it appeared failed, the
636        capturing was not being forgotten if a higher numbered group was later
637        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
638        group 1 to "a", when in fact it should be unset. This applied to multi-
639        branched capturing and non-capturing groups, repeated or not, and also to
640        positive assertions (capturing in negative assertions does not happen
641        in PCRE) and also to nested atomic groups.
643    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
644        subject after a captured substring, to make it easier to tell which of a
645        number of identical substrings has been captured.
647    24. The way atomic groups are processed by pcre_exec() has been changed so that
648        if they are repeated, backtracking one repetition now resets captured
649        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
650        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
651        "aaa". Previously, it would have been "a". As part of this code
652        refactoring, the way recursive calls are handled has also been changed.
654    25. If an assertion condition captured any substrings, they were not passed
655        back unless some other capturing happened later. For example, if
656        (?(?=(a))a) was matched against "a", no capturing was returned.
658    26. When studying a pattern that contained subroutine calls or assertions,
659        the code for finding the minimum length of a possible match was handling
660        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
661        group 1 called group 2 while simultaneously a separate group 2 called group
662        1). A stack overflow occurred in this case. I have fixed this by limiting
663        the recursion depth to 10.
665    27. Updated RunTest.bat in the distribution to the version supplied by Tom
666        Fortmann. This supports explicit test numbers on the command line, and has
667        argument validation and error reporting.
669    28. An instance of \X with an unlimited repeat could fail if at any point the
670        first character it looked at was a mark character.
672    29. Some minor code refactoring concerning Unicode properties and scripts
673        should reduce the stack requirement of match() slightly.
675    30. Added the '=' option to pcretest to check the setting of unused capturing
676        slots at the end of the pattern, which are documented as being -1, but are
677        not included in the return count.
679    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
680        compiled something random. Now it gives a compile-time error (as does
681        Perl).
683    32. A *MARK encountered during the processing of a positive assertion is now
684        recorded and passed back (compatible with Perl).
686    33. If --only-matching or --colour was set on a pcregrep call whose pattern
687        had alternative anchored branches, the search for a second match in a line
688        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
689        matched the line "0102" twice. The same bug affected patterns that started
690        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
691        twice.
693    34. Previously, PCRE did not allow quantification of assertions. However, Perl
694        does, and because of capturing effects, quantifying parenthesized
695        assertions may at times be useful. Quantifiers are now allowed for
696        parenthesized assertions.
698    35. A minor code tidy in pcre_compile() when checking options for \R usage.
700    36. \g was being checked for fancy things in a character class, when it should
701        just be a literal "g".
703    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
704        appearance of a nested POSIX class supersedes an apparent external class.
705        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
706        unescaped square brackets may also appear as part of class names. For
707        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
708        more like Perl. (But see 8.20/1 above.)
710    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
711        was because it thought it was \N{name}, which is not supported).
713    39. Add minix to OS list not supporting the -S option in pcretest.
715    40. PCRE tries to detect cases of infinite recursion at compile time, but it
716        cannot analyze patterns in sufficient detail to catch mutual recursions
717        such as ((?1))((?2)). There is now a runtime test that gives an error if a
718        subgroup is called recursively as a subpattern for a second time at the
719        same position in the subject string. In previous releases this might have
720        been caught by the recursion limit, or it might have run out of stack.
722    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
723        happen only once. PCRE was, however incorrectly giving a compile time error
724        "recursive call could loop indefinitely" because it cannot analyze the
725        pattern in sufficient detail. The compile time test no longer happens when
726        PCRE is compiling a conditional subpattern, but actual runaway loops are
727        now caught at runtime (see 40 above).
729    42. It seems that Perl allows any characters other than a closing parenthesis
730        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
731        has been changed to be the same.
733    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
734        as not to get warnings when autogen.sh is called. Also changed
735        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
737    44. To help people who use pcregrep to scan files containing exceedingly long
738        lines, the following changes have been made:
740        (a) The default value of the buffer size parameter has been increased from
741            8K to 20K. (The actual buffer used is three times this size.)
743        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
744            PCRE is built.
746        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
747            to be set at run time.
749        (d) Numerical values in pcregrep options can be followed by K or M, for
750            example --buffer-size=50K.
752        (e) If a line being scanned overflows pcregrep's buffer, an error is now
753            given and the return code is set to 2.
755    45. Add a pointer to the latest mark to the callout data block.
757    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
758        partial match of an empty string instead of no match. This was specific to
759        the use of ".".
761    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
762        complete match instead of a partial match. This bug was dependent on both
763        the PCRE_UTF8 and PCRE_DOTALL options being set.
765    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
766        starting byte set, because \b was not being ignored.
769    Version 8.12 15-Jan-2011
770    ------------------------
772    1.  Fixed some typos in the markup of the man pages, and wrote a script that
773        checks for such things as part of the documentation building process.
775    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
776        --match-limit and --recursion-limit options (added for 8.11). In
777        particular, this made one of the standard tests fail. (The integer value
778        went into the wrong half of a long int.)
780    3.  If the --colour option was given to pcregrep with -v (invert match), it
781        did strange things, either producing crazy output, or crashing. It should,
782        of course, ignore a request for colour when reporting lines that do not
783        match.
785    4.  Another pcregrep bug caused similar problems if --colour was specified with
786        -M (multiline) and the pattern match finished with a line ending.
788    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
789        matched in multiline mode, the following line was shown as part of the
790        match. This seems wrong, so I have changed it.
792    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
793        the check for further matches in the same line (so they could be coloured)
794        to overrun the end of the current line. If another match was found, it was
795        incorrectly shown (and then shown again when found in the next line).
797    7.  If pcregrep was compiled under Windows, there was a reference to the
798        function pcregrep_exit() before it was defined. I am assuming this was
799        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
800        reported by a user. I've moved the definition above the reference.
803    Version 8.11 10-Dec-2010
804    ------------------------
806    1.  (*THEN) was not working properly if there were untried alternatives prior
807        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
808        backtracked to try for "b" instead of moving to the next alternative branch
809        at the same level (in this case, to look for "c"). The Perl documentation
810        is clear that when (*THEN) is backtracked onto, it goes to the "next
811        alternative in the innermost enclosing group".
813    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
814        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
815        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
816        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
817        (*THEN).
819    3.  If \s appeared in a character class, it removed the VT character from
820        the class, even if it had been included by some previous item, for example
821        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
822        of \s, but is part of the POSIX "space" class.)
824    4.  A partial match never returns an empty string (because you can always
825        match an empty string at the end of the subject); however the checking for
826        an empty string was starting at the "start of match" point. This has been
827        changed to the "earliest inspected character" point, because the returned
828        data for a partial match starts at this character. This means that, for
829        example, /(?<=abc)def/ gives a partial match for the subject "abc"
830        (previously it gave "no match").
832    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
833        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
834        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
835        has an implication that the given string is incomplete (because a partial
836        match is preferred over a full match). For this reason, these items now
837        give a partial match in this situation. [Aside: previously, the one case
838        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
839        match rather than a full match, which was wrong by the old rules, but is
840        now correct.]
842    6.  There was a bug in the handling of #-introduced comments, recognized when
843        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
844        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
845        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
846        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
847        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
848        places in pcre_compile().
850    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
851        comments when looking ahead for named forward references to subpatterns,
852        the only newline sequence it recognized was NL. It now handles newlines
853        according to the set newline convention.
855    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
856        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
857        cater for a lack of strerror(). These oversights have been fixed.
859    9.  Added --match-limit and --recursion-limit to pcregrep.
861    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
863    11. When the -o option was used, pcregrep was setting a return code of 1, even
864        when matches were found, and --line-buffered was not being honoured.
866    12. Added an optional parentheses number to the -o and --only-matching options
867        of pcregrep.
869    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
870        can match an empty string. The code to do it in pcretest and pcredemo
871        needed fixing:
873        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
874            only one byte after an empty string match just before CRLF (this case
875            just got forgotten; "any" and "anycrlf" were OK).
877        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
878            mode when an empty string match preceded an ASCII character followed by
879            a non-ASCII character. (The code for advancing by one character rather
880            than one byte was nonsense.)
882        (c) The pcredemo.c sample program did not have any code at all to handle
883            the cases when CRLF is a valid newline sequence.
885    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
886        as a starting offset was within the subject string. There is now a new
887        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
888        negative or greater than the length of the string. In order to test this,
889        pcretest is extended to allow the setting of negative starting offsets.
891    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
892        starting offset points to the beginning of a UTF-8 character was
893        unnecessarily clumsy. I tidied it up.
895    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
896        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
898    17. Nobody had reported that the --include_dir option, which was added in
899        release 7.7 should have been called --include-dir (hyphen, not underscore)
900        for compatibility with GNU grep. I have changed it to --include-dir, but
901        left --include_dir as an undocumented synonym, and the same for
902        --exclude-dir, though that is not available in GNU grep, at least as of
903        release 2.5.4.
905    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
906        characters from a string of bytes) have been redefined so as not to use
907        loops, in order to improve performance in some environments. At the same
908        time, I abstracted some of the common code into auxiliary macros to save
909        repetition (this should not affect the compiled code).
911    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
912        compile-time error is now given if \c is not followed by an ASCII
913        character, that is, a byte less than 128. (In EBCDIC mode, the code is
914        different, and any byte value is allowed.)
916    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
917        START_OPTIMIZE option, which is now allowed at compile time - but just
918        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
919        to pcregrep and other applications that have no direct access to PCRE
920        options. The new /Y option in pcretest sets this option when calling
921        pcre_compile().
923    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
924        back references. Groups containing recursive back references were forced to
925        be atomic by that change, but in the case of named groups, the amount of
926        memory required was incorrectly computed, leading to "Failed: internal
927        error: code overflow". This has been fixed.
929    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
930        pcretest.c, to avoid build problems in some Borland environments.
933    Version 8.10 25-Jun-2010
934    ------------------------
936    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
937        THEN.
939    2.  (*ACCEPT) was not working when inside an atomic group.
941    3.  Inside a character class, \B is treated as a literal by default, but
942        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
943        causes the error). The code is unchanged, but I tidied the documentation.
945    4.  Inside a character class, PCRE always treated \R and \X as literals,
946        whereas Perl faults them if its -w option is set. I have changed PCRE so
947        that it faults them when PCRE_EXTRA is set.
949    5.  Added support for \N, which always matches any character other than
950        newline. (It is the same as "." when PCRE_DOTALL is not set.)
952    6.  When compiling pcregrep with newer versions of gcc which may have
953        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
954        declared with attribute warn_unused_result" were given. Just casting the
955        result to (void) does not stop the warnings; a more elaborate fudge is
956        needed. I've used a macro to implement this.
958    7.  Minor change to pcretest.c to avoid a compiler warning.
960    8.  Added four artifical Unicode properties to help with an option to make
961        \s etc use properties (see next item). The new properties are: Xan
962        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
964    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
965        use Unicode properties. (*UCP) at the start of a pattern can be used to set
966        this option. Modified pcretest to add /W to test this facility. Added
967        REG_UCP to make it available via the POSIX interface.
969    10. Added --line-buffered to pcregrep.
971    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
972        studied, and the match started with a letter with a code point greater than
973        127 whose first byte was different to the first byte of the other case of
974        the letter, the other case of this starting letter was not recognized
975        (#976).
977    12. If a pattern that was studied started with a repeated Unicode property
978        test, for example, \p{Nd}+, there was the theoretical possibility of
979        setting up an incorrect bitmap of starting bytes, but fortunately it could
980        not have actually happened in practice until change 8 above was made (it
981        added property types that matched character-matching opcodes).
983    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
984        possible starting bytes for non-anchored patterns.
986    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
987        \R, and also a number of cases that involve Unicode properties, both
988        explicit and implicit when PCRE_UCP is set.
990    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
991        input, it could crash or give wrong results if characters with values
992        greater than 0xc0 were present in the subject string. (Detail: it assumed
993        UTF-8 input when processing these items.)
995    16. Added a lot of (int) casts to avoid compiler warnings in systems where
996        size_t is 64-bit (#991).
998    17. Added a check for running out of memory when PCRE is compiled with
999        --disable-stack-for-recursion (#990).
1001    18. If the last data line in a file for pcretest does not have a newline on
1002        the end, a newline was missing in the output.
1004    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1005        less than 128) in its various bitmaps. However, there is a facility for
1006        generating tables according to the current locale when PCRE is compiled. It
1007        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1008        characters, are recognized by isspace() and therefore were getting set in
1009        these tables, and indeed these tables seem to approximate to ISO 8859. This
1010        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1011        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1012        which of course cannot start UTF-8 characters. I have changed the code so
1013        that only real ASCII characters (less than 128) and the correct starting
1014        bytes for UTF-8 encodings are set for characters greater than 127 when in
1015        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1016        altogether.)
1018    20. Added the /T option to pcretest so as to be able to run tests with non-
1019        standard character tables, thus making it possible to include the tests
1020        used for 19 above in the standard set of tests.
1022    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1023        reference to a subpattern the other side of a comment that contains an
1024        opening parenthesis caused either an internal compiling error, or a
1025        reference to the wrong subpattern.
1028    Version 8.02 19-Mar-2010
1029    ------------------------
1031    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1033    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1034        configured.
1036    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1037        original author of that file, following a query about its status.
1039    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1040        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1042    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1043        quantifier applied to a forward-referencing subroutine call, could compile
1044        incorrect code or give the error "internal error: previously-checked
1045        referenced subpattern not found".
1047    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1048        variables to point to external functions. For these systems, therefore,
1049        pcre_malloc etc. are now initialized to local functions that call the
1050        relevant global functions.
1052    7.  There were two entries missing in the vectors called coptable and poptable
1053        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1054        I've fixed the data, and added a kludgy way of testing at compile time that
1055        the lengths are correct (equal to the number of opcodes).
1057    8.  Following on from 7, I added a similar kludge to check the length of the
1058        eint vector in pcreposix.c.
1060    9.  Error texts for pcre_compile() are held as one long string to avoid too
1061        much relocation at load time. To find a text, the string is searched,
1062        counting zeros. There was no check for running off the end of the string,
1063        which could happen if a new error number was added without updating the
1064        string.
1066    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1068    11. \K was not working if it appeared in an atomic group or in a group that
1069        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1070        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1071        the assertion is positive, but not if it is negative.
1073    12. Change 11 fortuitously reduced the size of the stack frame used in the
1074        "match()" function of pcre_exec.c by one pointer. Forthcoming
1075        implementation of support for (*MARK) will need an extra pointer on the
1076        stack; I have reserved it now, so that the stack frame size does not
1077        decrease.
1079    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1080        item in branch that calls a recursion is a subroutine call - as in the
1081        second branch in the above example - was incorrectly given the compile-
1082        time error "recursive call could loop indefinitely" because pcre_compile()
1083        was not correctly checking the subroutine for matching a non-empty string.
1085    14. The checks for overrunning compiling workspace could trigger after an
1086        overrun had occurred. This is a "should never occur" error, but it can be
1087        triggered by pathological patterns such as hundreds of nested parentheses.
1088        The checks now trigger 100 bytes before the end of the workspace.
1090    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1093    Version 8.01 19-Jan-2010
1094    ------------------------
1096    1.  If a pattern contained a conditional subpattern with only one branch (in
1097        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1098        computed the wrong minimum data length (which is of course zero for such
1099        subpatterns). This could cause incorrect "no match" results.
1101    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1102        the pattern is reset in the first branch, pcre_compile() failed with
1103        "internal error: code overflow at offset...". This happened only when
1104        the reset was to the original external option setting. (An optimization
1105        abstracts leading options settings into an external setting, which was the
1106        cause of this.)
1108    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1109        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1110        assertion pattern did not match (meaning that the assertion was true), it
1111        was incorrectly treated as false if the SKIP had been reached during the
1112        matching. This also applied to assertions used as conditions.
1114    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1115        assertion subpattern, including such a pattern used as a condition,
1116        unpredictable results occurred, instead of the error return
1119    5.  The C++ GlobalReplace function was not working like Perl for the special
1120        situation when an empty string is matched. It now does the fancy magic
1121        stuff that is necessary.
1123    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1124        removed. (These were left over from very, very early versions of PCRE.)
1126    7.  Some cosmetic changes to the code to make life easier when compiling it
1127        as part of something else:
1129        (a) Change DEBUG to PCRE_DEBUG.
1131        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1132            called "current" as "current_branch", to prevent a collision with the
1133            Linux macro when compiled as a kernel module.
1135        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1136            prevent a collision with the Linux macro when compiled as a kernel
1137            module.
1139    8.  In pcre_compile() there are some checks for integer overflows that used to
1140        cast potentially large values to (double). This has been changed to that
1141        when building, a check for int64_t is made, and if it is found, it is used
1142        instead, thus avoiding the use of floating point arithmetic. (There is no
1143        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1144        double.
1146    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1147        2005 (difference between two addresses compared to an unsigned value).
1149    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1150        custom one, because of the following reported problem in Windows:
1152          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1153              under Win32.
1154          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1155              therefore missing the function definition.
1156          - The compiler thus generates a "C" signature for the test function.
1157          - The linker fails to find the "C" function.
1158          - PCRE fails to configure if asked to do so against libbz2.
1160    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1161        messages were output:
1163          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1164          rerunning libtoolize, to keep the correct libtool macros in-tree.
1165          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1167        I have done both of these things.
1169    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1170        most of the time, it *can* run out if it is given a pattern that contains a
1171        runaway infinite recursion. I updated the discussion in the pcrestack man
1172        page.
1174    13. Now that we have gone to the x.xx style of version numbers, the minor
1175        version may start with zero. Using 08 or 09 is a bad idea because users
1176        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1177        interpreted as invalid octal numbers. I've updated the previous comment in
1178        configure.ac, and also added a check that gives an error if 08 or 09 are
1179        used.
1181    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1182        causing partial matching to fail when the end of the subject matched \W
1183        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1185    15. There were some discrepancies between the declarations in pcre_internal.h
1186        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1187        their definitions. The declarations used "const uschar *" and the
1188        definitions used USPTR. Even though USPTR is normally defined as "const
1189        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1190        reported that: "This difference in casting confuses some C++ compilers, for
1191        example, SunCC recognizes above declarations as different functions and
1192        generates broken code for hbpcre." I have changed the declarations to use
1193        USPTR.
1195    16. GNU libtool is named differently on some systems. The autogen.sh script now
1196        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1197        (FreeBSD).
1199    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1200        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1201        comment: "Figure out how to create a longlong from a string: strtoll and
1202        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1203        instance, but it only takes 2 args instead of 3!"
1205    18. A subtle bug concerned with back references has been fixed by a change of
1206        specification, with a corresponding code fix. A pattern such as
1207        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1208        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1209        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1210        same bug. Such groups have to be quantified to be useful, or contained
1211        inside another quantified group. (If there's no repetition, the reference
1212        can never match.) The problem arises because, having left the group and
1213        moved on to the rest of the pattern, a later failure that backtracks into
1214        the group uses the captured value from the final iteration of the group
1215        rather than the correct earlier one. I have fixed this in PCRE by forcing
1216        any group that contains a reference to itself to be an atomic group; that
1217        is, there cannot be any backtracking into it once it has completed. This is
1218        similar to recursive and subroutine calls.
1221    Version 8.00 19-Oct-09
1222    ----------------------
1224    1.  The table for translating pcre_compile() error codes into POSIX error codes
1225        was out-of-date, and there was no check on the pcre_compile() error code
1226        being within the table. This could lead to an OK return being given in
1227        error.
1229    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1230        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1231        in a Windows environment.
1233    3.  The pcregrep --count option prints the count for each file even when it is
1234        zero, as does GNU grep. However, pcregrep was also printing all files when
1235        --files-with-matches was added. Now, when both options are given, it prints
1236        counts only for those files that have at least one match. (GNU grep just
1237        prints the file name in this circumstance, but including the count seems
1238        more useful - otherwise, why use --count?) Also ensured that the
1239        combination -clh just lists non-zero counts, with no names.
1241    4.  The long form of the pcregrep -F option was incorrectly implemented as
1242        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1243        but it seems right to fix it, and I didn't think it was worth preserving
1244        the old behaviour.
1246    5.  The command line items --regex=pattern and --regexp=pattern were not
1247        recognized by pcregrep, which required --regex pattern or --regexp pattern
1248        (with a space rather than an '='). The man page documented the '=' forms,
1249        which are compatible with GNU grep; these now work.
1251    6.  No libpcreposix.pc file was created for pkg-config; there was just
1252        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1254    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1255        when UCP support is not needed, by modifying the Python script that
1256        generates it from Unicode data files. This should not matter if the module
1257        is correctly used as a library, but I received one complaint about 50K of
1258        unwanted data. My guess is that the person linked everything into his
1259        program rather than using a library. Anyway, it does no harm.
1261    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1262        was a minimum greater than 1 for a wide character in a possessive
1263        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1264        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1265        character. Chaos in the form of incorrect output or a compiling loop could
1266        result.
1268    9.  The restrictions on what a pattern can contain when partial matching is
1269        requested for pcre_exec() have been removed. All patterns can now be
1270        partially matched by this function. In addition, if there are at least two
1271        slots in the offset vector, the offset of the earliest inspected character
1272        for the match and the offset of the end of the subject are set in them when
1273        PCRE_ERROR_PARTIAL is returned.
1275    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1276        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1277        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1278        and may be more useful for multi-segment matching.
1280    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1281        used to be given if ever the end of the subject was reached; now it is
1282        given only if matching could not proceed because another character was
1283        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1284        string "Z", which now yields "no match" instead of "partial match". In the
1285        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1286        final character ended with (*FAIL).
1288    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1289        if the pattern had a "must contain" character that was already found in the
1290        earlier partial match, unless partial matching was again requested. For
1291        example, with the pattern /dog.(body)?/, the "must contain" character is
1292        "g". If the first part-match was for the string "dog", restarting with
1293        "sbody" failed. This bug has been fixed.
1295    13. The string returned by pcre_dfa_exec() after a partial match has been
1296        changed so that it starts at the first inspected character rather than the
1297        first character of the match. This makes a difference only if the pattern
1298        starts with a lookbehind assertion or \b or \B (\K is not supported by
1299        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1300        matching functions compatible, and I think it's the right thing to do.
1302    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1303        so that the demonstration program is easily available in environments where
1304        PCRE has not been installed from source.
1306    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1307        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1308        library.
1310    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1311        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1312        is not the first non-POSIX option to be added. Clearly some people find
1313        these options useful.
1315    17. If a caller to the POSIX matching function regexec() passes a non-zero
1316        value for nmatch with a NULL value for pmatch, the value of
1317        nmatch is forced to zero.
1319    18. RunGrepTest did not have a test for the availability of the -u option of
1320        the diff command, as RunTest does. It now checks in the same way as
1321        RunTest, and also checks for the -b option.
1323    19. If an odd number of negated classes containing just a single character
1324        interposed, within parentheses, between a forward reference to a named
1325        subpattern and the definition of the subpattern, compilation crashed with
1326        an internal error, complaining that it could not find the referenced
1327        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1328        [The bug was that it was starting one character too far in when skipping
1329        over the character class, thus treating the ] as data rather than
1330        terminating the class. This meant it could skip too much.]
1332    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1333        /g option in pcretest when the pattern contains \K, which makes it possible
1334        to have an empty string match not at the start, even when the pattern is
1335        anchored. Updated pcretest and pcredemo to use this option.
1337    21. If the maximum number of capturing subpatterns in a recursion was greater
1338        than the maximum at the outer level, the higher number was returned, but
1339        with unset values at the outer level. The correct (outer level) value is
1340        now given.
1342    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1343        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1344        make it do so. The string so far is captured, making this feature
1345        compatible with Perl.
1347    23. The tests have been re-organized, adding tests 11 and 12, to make it
1348        possible to check the Perl 5.10 features against Perl 5.10.
1350    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1351        pattern matches a fixed length string. PCRE did not allow this; now it
1352        does. Neither allows recursion.
1354    25. I finally figured out how to implement a request to provide the minimum
1355        length of subject string that was needed in order to match a given pattern.
1356        (It was back references and recursion that I had previously got hung up
1357        on.) This code has now been added to pcre_study(); it finds a lower bound
1358        to the length of subject needed. It is not necessarily the greatest lower
1359        bound, but using it to avoid searching strings that are too short does give
1360        some useful speed-ups. The value is available to calling programs via
1361        pcre_fullinfo().
1363    26. While implementing 25, I discovered to my embarrassment that pcretest had
1364        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1365        study optimizations had never been tested with that matching function.
1366        Oops. What is worse, even when it was passed study data, there was a bug in
1367        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1368        were also very few tests of studied patterns with pcre_dfa_exec().
1370    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1371        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1372        on the other side of the coin, they are no longer allowed to have different
1373        names, because these cannot be distinguished in PCRE, and this has caused
1374        confusion. (This is a difference from Perl.)
1376    28. When duplicate subpattern names are present (necessarily with different
1377        numbers, as required by 27 above), and a test is made by name in a
1378        conditional pattern, either for a subpattern having been matched, or for
1379        recursion in such a pattern, all the associated numbered subpatterns are
1380        tested, and the overall condition is true if the condition is true for any
1381        one of them. This is the way Perl works, and is also more like the way
1382        testing by number works.
1385    Version 7.9 11-Apr-09
1386    ---------------------
1388    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1389        (pcretest), all targets were linked against these libraries. This included
1390        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1391        libraries. This caused unwanted dependencies to be created. This problem
1392        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1393        pcretest is linked with readline.
1395    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1396        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1397        moved outside it again, because FALSE and TRUE are already defined in AIX,
1398        but BOOL is not.
1400    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1401        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1403    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1404        hyphens) following file names and line numbers when outputting matching
1405        lines. This is not true; no spaces are inserted. I have also clarified the
1406        wording for the --colour (or --color) option.
1408    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1409        was not coloured; this is different to GNU grep, so I have changed it to be
1410        the same.
1412    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1413        each matching line was coloured. Now it goes on to look for further matches
1414        of any of the test patterns, which is the same behaviour as GNU grep.
1416    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1417        doesn't make sense to accept an empty string match in pcregrep, so I have
1418        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1419        seems to be how GNU grep behaves.
1421    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1422        start or after a newline", because the conditional assertion was not being
1423        correctly handled. The rule now is that both the assertion and what follows
1424        in the first alternative must satisfy the test.
1426    9.  If auto-callout was enabled in a pattern with a conditional group whose
1427        condition was an assertion, PCRE could crash during matching, both with
1428        pcre_exec() and pcre_dfa_exec().
1430    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1431        used for matching.
1433    11. Unicode property support in character classes was not working for
1434        characters (bytes) greater than 127 when not in UTF-8 mode.
1436    12. Added the -M command line option to pcretest.
1438    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1440    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1442    16. Added comments and documentation about mis-use of no_arg in the C++
1443        wrapper.
1445    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1446        from Martin Jerabek that uses macro names for all relevant character and
1447        string constants.
1449    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1450        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1451        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1452        these, but not everybody uses configure.
1454    19. A conditional group that had only one branch was not being correctly
1455        recognized as an item that could match an empty string. This meant that an
1456        enclosing group might also not be so recognized, causing infinite looping
1457        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1458        with the subject "ab", where knowledge that the repeated group can match
1459        nothing is needed in order to break the loop.
1461    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1462        exec(), but without supplying a callout function, matching went wrong.
1464    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1465        leak if the size of the offset vector was greater than 30. When the vector
1466        is smaller, the saved offsets during recursion go onto a local stack
1467        vector, but for larger vectors malloc() is used. It was failing to free
1468        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1469        error, in fact).
1471    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1472        heapframe that is used only when UTF-8 support is enabled. This caused no
1473        problem, but was untidy.
1475    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1476        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1477        included within another project.
1479    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1480        slightly modified by me:
1482          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1483              not building pcregrep.
1485          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1486              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1488    25. Forward references, both numeric and by name, in patterns that made use of
1489        duplicate group numbers, could behave incorrectly or give incorrect errors,
1490        because when scanning forward to find the reference group, PCRE was not
1491        taking into account the duplicate group numbers. A pattern such as
1492        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1494    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1495        the feature of a custom pointer more persuasive (as requested by a user).
1497    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1498        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1499        pre-defined.
1501    28. Added support for (*UTF8) at the start of a pattern.
1503    29. Arrange for flags added by the "release type" setting in CMake to be shown
1504        in the configuration summary.
1507    Version 7.8 05-Sep-08
1508    ---------------------
1510    1.  Replaced UCP searching code with optimized version as implemented for Ad
1511        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1512        stage table and inline lookup instead of a function, giving speed ups of 2
1513        to 5 times on some simple patterns that I tested. Permission was given to
1514        distribute the MultiStage2.py script that generates the tables (it's not in
1515        the tarball, but is in the Subversion repository).
1517    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1518        scripts.
1520    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1521        a group with a zero qualifier. The result of the study could be incorrect,
1522        or the function might crash, depending on the pattern.
1524    4.  Caseless matching was not working for non-ASCII characters in back
1525        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1526        It now works when Unicode Property Support is available.
1528    5.  In pcretest, an escape such as \x{de} in the data was always generating
1529        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1530        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1531        truncation.
1533    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1535    7.  Added two (int) casts to pcregrep when printing the difference of two
1536        pointers, in case they are 64-bit values.
1538    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1539        test 2 if it fails.
1541    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1542        and a #define of that name to empty if it is not externally set. This is to
1543        allow users of MSVC to set it if necessary.
1545    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1546        the convenience functions in the pcre_get.c source file.
1548    11. An option change at the start of a pattern that had top-level alternatives
1549        could cause overwriting and/or a crash. This command provoked a crash in
1550        some environments:
1552          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1554        This potential security problem was recorded as CVE-2008-2371.
1556    12. For a pattern where the match had to start at the beginning or immediately
1557        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1558        pcre_dfa_exec() could read past the end of the passed subject if there was
1559        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1560        pcretest so that it places the subject at the end of its malloc-ed buffer.
1562    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1563        exec() might read past the end of the data buffer in UTF-8 mode.
1565    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1566        the data contained the byte 0x85 as part of a UTF-8 character within its
1567        first line. This applied both to normal and DFA matching.
1569    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1570        /^[^d]*?$/8 failed to match "abc".
1572    16. Added a missing copyright notice to pcrecpp_internal.h.
1574    17. Make it more clear in the documentation that values returned from
1575        pcre_exec() in ovector are byte offsets, not character counts.
1577    18. Tidied a few places to stop certain compilers from issuing warnings.
1579    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1580        supplied by Stefan Weber. I made a further small update for 7.8 because
1581        there is a change of source arrangements: the pcre_searchfuncs.c module is
1582        replaced by pcre_ucd.c.
1585    Version 7.7 07-May-08
1586    ---------------------
1588    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1589        a string to a long long, pretend we don't even have a long long." This is
1590        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1592    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1593        pre-7.6 versions, which defined a global no_arg variable instead of putting
1594        it in the RE class. (See also #8 below.)
1596    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1597        Lopes.
1599    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1601        (1) The include/exclude patterns were being applied to the whole pathnames
1602            of files, instead of just to the final components.
1604        (2) If there was more than one level of directory, the subdirectories were
1605            skipped unless they satisfied the include/exclude conditions. This is
1606            inconsistent with GNU grep (and could even be seen as contrary to the
1607            pcregrep specification - which I improved to make it absolutely clear).
1608            The action now is always to scan all levels of directory, and just
1609            apply the include/exclude patterns to regular files.
1611    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1612        --exclude_dir in the tests to avoid scanning .svn directories.
1614    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1615        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1616        doesn't support NULs in patterns.
1618    7.  Added some missing "const"s to declarations of static tables in
1619        pcre_compile.c and pcre_dfa_exec.c.
1621    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1622        caused by fix #2  above. (Subsequently also a second patch to fix the
1623        first patch. And a third patch - this was a messy problem.)
1625    9.  Applied Craig's patch to remove the use of push_back().
1627    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1628        matching function regexec().
1630    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1631        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1632        references. PCRE supports relative numbers with this syntax (I don't think
1633        Oniguruma does).
1635    12. Previously, a group with a zero repeat such as (...){0} was completely
1636        omitted from the compiled regex. However, this means that if the group
1637        was called as a subroutine from elsewhere in the pattern, things went wrong
1638        (an internal error was given). Such groups are now left in the compiled
1639        pattern, with a new opcode that causes them to be skipped at execution
1640        time.
1642    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1643        to the way PCRE behaves:
1645        (a) A lone ] character is dis-allowed (Perl treats it as data).
1647        (b) A back reference to an unmatched subpattern matches an empty string
1648            (Perl fails the current match path).
1650        (c) A data ] in a character class must be notated as \] because if the
1651            first data character in a class is ], it defines an empty class. (In
1652            Perl it is not possible to have an empty class.) The empty class []
1653            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1654            The negative empty class [^] matches any one character, independently
1655            of the DOTALL setting.
1657    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1658        non-existent subpattern following a character class starting with ']' and
1659        containing () gave an internal compiling error instead of "reference to
1660        non-existent subpattern". Fortunately, when the pattern did exist, the
1661        compiled code was correct. (When scanning forwards to check for the
1662        existencd of the subpattern, it was treating the data ']' as terminating
1663        the class, so got the count wrong. When actually compiling, the reference
1664        was subsequently set up correctly.)
1666    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1667        it was being rejected as not supported by pcre_dfa_exec(), even though
1668        other assertions are supported. I have made pcre_dfa_exec() support
1669        (*FAIL).
1671    16. The implementation of 13c above involved the invention of a new opcode,
1672        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1673        cannot be changed at match time, I realized I could make a small
1674        improvement to matching performance by compiling OP_ALLANY instead of
1675        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1676        on the OP_ANY path.
1678    17. Compiling pcretest on Windows with readline support failed without the
1679        following two fixes: (1) Make the unistd.h include conditional on
1680        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1682    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1683        ncurses library to be included for pcretest when ReadLine support is
1684        requested, but also to allow for it to be overridden. This patch came from
1685        Daniel Bergström.
1687    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1688        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1689        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1690        spotting this.
1693    Version 7.6 28-Jan-08
1694    ---------------------
1696    1.  A character class containing a very large number of characters with
1697        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1698        overflow.
1700    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1701        HAVE_LONG_LONG is not defined.
1703    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1704        bring them up to date and include new features. This patch includes:
1706        - Fixed PH's badly added libz and libbz2 support.
1707        - Fixed a problem with static linking.
1708        - Added pcredemo. [But later removed - see 7 below.]
1709        - Fixed dftables problem and added an option.
1710        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1711            HAVE_LONG_LONG.
1712        - Added readline support for pcretest.
1713        - Added an listing of the option settings after cmake has run.
1715    4.  A user submitted a patch to Makefile that makes it easy to create
1716        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1717        Makefile.am that cause it to include this special target, without
1718        affecting anything else. Note that the same mingw target plus all
1719        the other distribution libraries and programs are now supported
1720        when configuring with CMake (see 6 below) instead of with
1721        Configure/Make.
1723    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1724        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1725        exported in the Windows port". It has not yet been confirmed that the patch
1726        solves the problem, but it does no harm.
1728    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1729        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1730        with CMake, and also correct the comment about stack recursion.
1732    7.  Remove the automatic building of pcredemo from the ./configure system and
1733        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1734        of a program that users should build themselves after PCRE is installed, so
1735        building it automatically is not really right. What is more, it gave
1736        trouble in some build environments.
1738    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1741    Version 7.5 10-Jan-08
1742    ---------------------
1744    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1745        values in parens when parsing an RE using the C++ wrapper."
1747    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1748        Characters greater than 255 were excluded from the class instead of being
1749        included.
1751    3.  The same bug as (2) above applied to negated POSIX classes such as
1752        [:^space:].
1754    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1755        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1756        I have changed it.
1758    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1759        first named subpattern) and a construct such as (?&a) would reference the
1760        first named subpattern whose name started with "a" (in other words, the
1761        length check was missing). Both these problems are fixed. "Subpattern name
1762        expected" is now given for (?&) (a zero-length name), and this patch also
1763        makes it give the same error for \k'' (previously it complained that that
1764        was a reference to a non-existent subpattern).
1766    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1767        this is right because (?- can be followed by option settings as well as by
1768        digits. I have, however, made the messages clearer.
1770    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1771        than the number used in the conditional) now cause a compile-time error.
1772        This is actually not compatible with Perl, which accepts such patterns, but
1773        treats the conditional as always being FALSE (as PCRE used to), but it
1774        seems to me that giving a diagnostic is better.
1776    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1777        and messages.
1779    9.  Fix two occurrences of "backslash" in comments that should have been
1780        "backspace".
1782    10. Remove two redundant lines of code that can never be obeyed (their function
1783        was moved elsewhere).
1785    11. The program that makes PCRE's Unicode character property table had a bug
1786        which caused it to generate incorrect table entries for sequences of
1787        characters that have the same character type, but are in different scripts.
1788        It amalgamated them into a single range, with the script of the first of
1789        them. In other words, some characters were in the wrong script. There were
1790        thirteen such cases, affecting characters in the following ranges:
1792          U+002b0 - U+002c1
1793          U+0060c - U+0060d
1794          U+0061e - U+00612
1795          U+0064b - U+0065e
1796          U+0074d - U+0076d
1797          U+01800 - U+01805
1798          U+01d00 - U+01d77
1799          U+01d9b - U+01dbf
1800          U+0200b - U+0200f
1801          U+030fc - U+030fe
1802          U+03260 - U+0327f
1803          U+0fb46 - U+0fbb1
1804          U+10450 - U+1049d
1806    12. The -o option (show only the matching part of a line) for pcregrep was not
1807        compatible with GNU grep in that, if there was more than one match in a
1808        line, it showed only the first of them. It now behaves in the same way as
1809        GNU grep.
1811    13. If the -o and -v options were combined for pcregrep, it printed a blank
1812        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1813        does the same. The return code can be used to tell if there were any
1814        non-matching lines.
1816    14. Added --file-offsets and --line-offsets to pcregrep.
1818    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1819        infinitely looping recursion. The bug was that positive lookaheads were not
1820        being skipped when checking for a possible empty match (negative lookaheads
1821        and both kinds of lookbehind were skipped).
1823    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1824        inclusion of <windows.h> to before rather than after the definition of
1825        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1827    17. Specifying a possessive quantifier with a specific limit for a Unicode
1828        character property caused pcre_compile() to compile bad code, which led at
1829        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1830        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1831        caused the error; without that there was no problem.
1833    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1835    19. Added --enable-pcretest-libreadline.
1837    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1838        RE::GlobalReplace(). As a result, the number of replacements returned was
1839        double what it should be. I removed one of the increments, but Craig sent a
1840        later patch that removed the other one (the right fix) and added unit tests
1841        that check the return values (which was not done before).
1843    21. Several CMake things:
1845        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1846            the names libpcre and libpcreposix, not just pcre and pcreposix.
1848        (2) The above change means that pcretest and pcregrep are now correctly
1849            linked with the newly-built libraries, not previously installed ones.
1853    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1854        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1855        UTF-8 newline character). The key issue is that the pattern starts .*;
1856        this means that the match must be either at the beginning, or after a
1857        newline. The bug was in the code for advancing after a failed match and
1858        checking that the new position followed a newline. It was not taking
1859        account of UTF-8 characters correctly.
1861    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1862        character classes. PCRE was not treating the sequence [:...:] as a
1863        character class unless the ... were all letters. Perl, however, seems to
1864        allow any characters between [: and :], though of course it rejects as
1865        unknown any "names" that contain non-letters, because all the known class
1866        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1867        for example, whereas PCRE did not - it did not recognize a POSIX character
1868        class. This seemed a bit dangerous, so the code has been changed to be
1869        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1870        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1871        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1872        Perl does, and where it didn't before.
1874    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1875        Windows environments %n is disabled by default.
1878    Version 7.4 21-Sep-07
1879    ---------------------
1881    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1882        means that a class such as [\s] counted as "explicit reference to CR or
1883        LF". That isn't really right - the whole point of the change was to try to
1884        help when there was an actual mention of one of the two characters. So now
1885        the change happens only if \r or \n (or a literal CR or LF) character is
1886        encountered.
1888    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1889        of both had grown to the point where there were only 3 bits left.
1890        Fortunately, there was spare space in the data structure, and so I have
1891        moved the internal flags into a new 16-bit field to free up more option
1892        bits.
1894    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1895        but did not set the internal JCHANGED flag - either of these is enough to
1896        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1897        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1898        start sets both bits.
1900    4.  Added options (at build time, compile time, exec time) to change \R from
1901        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1903    5.  doc/pcresyntax.html was missing from the distribution.
1905    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1906        compatibility, even though it is no longer used.
1908    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1909        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1910        windows.h file is present (where different names are used). [This was
1911        reversed later after testing - see 16 below.]
1913    8.  Changed all #include <config.h> to #include "config.h". There were also
1914        some further <pcre.h> cases that I changed to "pcre.h".
1916    9.  When pcregrep was used with the --colour option, it missed the line ending
1917        sequence off the lines that it output.
1919    10. It was pointed out to me that arrays of string pointers cause lots of
1920        relocations when a shared library is dynamically loaded. A technique of
1921        using a single long string with a table of offsets can drastically reduce
1922        these. I have refactored PCRE in four places to do this. The result is
1923        dramatic:
1925          Originally:                          290
1926          After changing UCP table:            187
1927          After changing error message table:   43
1928          After changing table of "verbs"       36
1929          After changing table of Posix names   22
1931        Thanks to the folks working on Gregex for glib for this insight.
1933    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1934        unicode-properties was also set.
1936    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1938    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1939        checked only for CRLF.
1941    14. Added casts to pcretest.c to avoid compiler warnings.
1943    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1945    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1946        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1947        entirely. This removes changes made in 7 above.
1949    17. The CMake files have been updated, and there is now more information about
1950        building with CMake in the NON-UNIX-USE document.
1953    Version 7.3 28-Aug-07
1954    ---------------------
1956     1. In the rejigging of the build system that eventually resulted in 7.1, the
1957        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1958        brackets there is not right, since it causes compilers to look for an
1959        installed pcre.h, not the version that is in the source that is being
1960        compiled (which of course may be different). I have changed it back to:
1962          #include "pcre.h"
1964        I have a vague recollection that the change was concerned with compiling in
1965        different directories, but in the new build system, that is taken care of
1966        by the VPATH setting the Makefile.
1968     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1969        when the subject happened to end in the byte 0x85 (e.g. if the last
1970        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1971        characters but of course it shouldn't be taken as a newline when it is part
1972        of another character. The bug was that, for an unlimited repeat of . in
1973        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1974        characters when looking for a newline.
1976     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1978     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1979        in debug output.
1981     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1982        long printing in the pcrecpp unittest when running under MinGW.
1984     6. ESC_K was left out of the EBCDIC table.
1986     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1987        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1988        limit also applies to "virtual nesting" when a pattern is recursive, and in
1989        this case 1000 isn't so big. I have been able to remove this limit at the
1990        expense of backing off one optimization in certain circumstances. Normally,
1991        when pcre_exec() would call its internal match() function recursively and
1992        immediately return the result unconditionally, it uses a "tail recursion"
1993        feature to save stack. However, when a subpattern that can match an empty
1994        string has an unlimited repetition quantifier, it no longer makes this
1995        optimization. That gives it a stack frame in which to save the data for
1996        checking that an empty string has been matched. Previously this was taken
1997        from the 1000-entry workspace that had been reserved. So now there is no
1998        explicit limit, but more stack is used.
2000     8. Applied Daniel's patches to solve problems with the import/export magic
2001        syntax that is required for Windows, and which was going wrong for the
2002        pcreposix and pcrecpp parts of the library. These were overlooked when this
2003        problem was solved for the main library.
2005     9. There were some crude static tests to avoid integer overflow when computing
2006        the size of patterns that contain repeated groups with explicit upper
2007        limits. As the maximum quantifier is 65535, the maximum group length was
2008        set at 30,000 so that the product of these two numbers did not overflow a
2009        32-bit integer. However, it turns out that people want to use groups that
2010        are longer than 30,000 bytes (though not repeat them that many times).
2011        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2012        made it possible to implement the integer overflow checks in a much more
2013        dynamic way, which I have now done. The artificial limitation on group
2014        length has been removed - we now have only the limit on the total length of
2015        the compiled pattern, which depends on the LINK_SIZE setting.
2017    10. Fixed a bug in the documentation for get/copy named substring when
2018        duplicate names are permitted. If none of the named substrings are set, the
2019        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2020        empty string.
2022    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2023        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2024        because the ] is interpreted as the first data character and the
2025        terminating ] is not found. PCRE has been made compatible with Perl in this
2026        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2027        cause memory overwriting.
2029    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2030        string has been matched (to stop an infinite loop). It was not recognizing
2031        a conditional subpattern that could match an empty string if that
2032        subpattern was within another subpattern. For example, it looped when
2033        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2034        condition was not nested. This bug has been fixed.
2036    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2037        past the start of the subject in the presence of bytes with the top bit
2038        set, for example "\x8aBCD".
2040    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2041        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2043    14. Optimized (?!) to (*FAIL).
2045    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2046        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2047        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2048        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2049        does: it's just the validity check that is more restrictive.
2051    16. Inserted checks for integer overflows during escape sequence (backslash)
2052        processing, and also fixed erroneous offset values for syntax errors during
2053        backslash processing.
2055    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2056        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2058    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2059        caused an overrun.
2061    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2062        something other than just ASCII characters) inside a group that had an
2063        unlimited repeat caused a loop at compile time (while checking to see
2064        whether the group could match an empty string).
2066    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2067        [\P{Any}] did so. (Error in the code for printing property names.)
2069    21. An orphan \E inside a character class could cause a crash.
2071    22. A repeated capturing bracket such as (A)? could cause a wild memory
2072        reference during compilation.
2074    23. There are several functions in pcre_compile() that scan along a compiled
2075        expression for various reasons (e.g. to see if it's fixed length for look
2076        behind). There were bugs in these functions when a repeated \p or \P was
2077        present in the pattern. These operators have additional parameters compared
2078        with \d, etc, and these were not being taken into account when moving along
2079        the compiled data. Specifically:
2081        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2082            length.
2084        (b) An item such as \pL+ within a repeated group could cause crashes or
2085            loops.
2087        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2088            "reference to non-existent subpattern" error.
2090        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2092    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2093        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2095    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2097    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2098        character were causing crashes (broken optimization).
2100    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2101        \p or \P) caused a compile-time loop.
2103    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2104        break. For example, the unstudied pattern [\r\n]A does not match the string
2105        "\r\nA" because change 7.0/46 below moves the current point on by two
2106        characters after failing to match at the start. However, the pattern \nA
2107        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2108        the same is true. There doesn't seem any very clean way out of this, but
2109        what I have chosen to do makes the common cases work: PCRE now takes note
2110        of whether there can be an explicit match for \r or \n anywhere in the
2111        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2112        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2113        pattern has explicit CR or LF references.
2115    29. Added (*CR) etc for changing newline setting at start of pattern.
2118    Version 7.2 19-Jun-07
2119    ---------------------
2121     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2122        which is apparently normally available under Windows.
2124     2. Re-jig the pcregrep tests with different newline settings in an attempt
2125        to make them independent of the local environment's newline setting.
2127     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2129     4. Some of the "internals" tests were previously cut out when the link size
2130        was not 2, because the output contained actual offsets. The recent new
2131        "Z" feature of pcretest means that these can be cut out, making the tests
2132        usable with all link sizes.
2134     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2135        stack recursion. This gives a massive performance boost under BSD, but just
2136        a small improvement under Linux. However, it saves one field in the frame
2137        in all cases.
2139     6. Added more features from the forthcoming Perl 5.10:
2141        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2142            recursion call. It refers to the nth most recently opened parentheses.
2144        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2145            to be opened parentheses.
2147        (c) Conditions that refer to capturing parentheses can be specified
2148            relatively, for example, (?(-2)... or (?(+3)...
2150        (d) \K resets the start of the current match so that everything before
2151            is not part of it.
2153        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2155        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2156            reference syntax.
2158        (g) (?| introduces a group in which the numbering of parentheses in each
2159            alternative starts with the same number.
2161        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2163     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2166     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2167        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2168        for detecting groups that can match an empty string.
2170     9. A pattern with a very large number of alternatives (more than several
2171        hundred) was running out of internal workspace during the pre-compile
2172        phase, where pcre_compile() figures out how much memory will be needed. A
2173        bit of new cunning has reduced the workspace needed for groups with
2174        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2175        workspace instead of running out of the 4096 that are available.
2177    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2179    11. Applied patch from Google to remove an optimization that didn't quite work.
2180        The report of the bug said:
2182          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2183          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2184          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2186    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2187        it matched the wrong number of bytes.
2190    Version 7.1 24-Apr-07
2191    ---------------------
2193     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2194        that is more "standard", making use of automake and other Autotools. There
2195        is some re-arrangement of the files and adjustment of comments consequent
2196        on this.
2198     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2199        for recursive directory scanning broke on some systems because the files
2200        are not scanned in any specific order and on different systems the order
2201        was different. A call to "sort" has been inserted into RunGrepTest for the
2202        approprate test as a short-term fix. In the longer term there may be an
2203        alternative.
2205     3. I had an email from Eric Raymond about problems translating some of PCRE's
2206        man pages to HTML (despite the fact that I distribute HTML pages, some
2207        people do their own conversions for various reasons). The problems
2208        concerned the use of low-level troff macros .br and .in. I have therefore
2209        removed all such uses from the man pages (some were redundant, some could
2210        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2211        HTML has been updated to handle .nf/.fi and to complain if it encounters
2212        .br or .in.
2214     4. Updated comments in configure.ac that get placed in config.h.in and also
2215        arranged for config.h to be included in the distribution, with the name
2216        config.h.generic, for the benefit of those who have to compile without
2217        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2219     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2220        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2221        makevp.bat for the current PCRE, using the additional files
2222        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2224     6. A Windows user reported a minor discrepancy with test 2, which turned out
2225        to be caused by a trailing space on an input line that had got lost in his
2226        copy. The trailing space was an accident, so I've just removed it.
2228     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2229        that is needed.
2231     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2232        as "const" (a) because they are and (b) because it helps the PHP
2233        maintainers who have recently made a script to detect big data structures
2234        in the php code that should be moved to the .rodata section. I remembered
2235        to update Builducptable as well, so it won't revert if ucptable.h is ever
2236        re-created.
2238     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2239        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2240        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2241        support is not required. This saves 1.5-2K of code, which is important in
2242        some applications.
2244        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2245        so as not to refer to the tables, even though these functions will never be
2246        called when UTF-8 support is disabled. Otherwise there are problems with a
2247        shared library.
2249    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2251        (a) It was defining its arguments as char * instead of void *.
2253        (b) It was assuming that all moves were upwards in memory; this was true
2254            a long time ago when I wrote it, but is no longer the case.
2256        The emulated memove() is provided for those environments that have neither
2257        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2258        is clearly not the case, as these two bugs were recently reported.
2260    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2261        and Detrail to create the HTML documentation, the .txt form of the man
2262        pages, and it removes trailing spaces from listed files. It also creates
2263        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2264        case, it wraps all the #defines with #ifndefs. This script should be run
2265        before "make dist".
2267    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2268        with Unicode property support.
2270        (a) For a maximizing quantifier, if the two different cases of the
2271            character were of different lengths in their UTF-8 codings (there are
2272            some cases like this - I found 11), and the matching function had to
2273            back up over a mixture of the two cases, it incorrectly assumed they
2274            were both the same length.
2276        (b) When PCRE was configured to use the heap rather than the stack for
2277            recursion during matching, it was not correctly preserving the data for
2278            the other case of a UTF-8 character when checking ahead for a match
2279            while processing a minimizing repeat. If the check also involved
2280            matching a wide character, but failed, corruption could cause an
2281            erroneous result when trying to check for a repeat of the original
2282            character.
2284    13. Some tidying changes to the testing mechanism:
2286        (a) The RunTest script now detects the internal link size and whether there
2287            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2288            values substituted by "configure". (The RunGrepTest script already did
2289            this for UTF-8.) The configure.ac script no longer substitutes the
2290            relevant variables.
2292        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2293            with length and offset values. This means that the output is different
2294            for different internal link sizes. Test 2 is skipped for link sizes
2295            other than 2 because of this, bypassing the problem. Unfortunately,
2296            there was also a test in test 3 (the locale tests) that used /B and
2297            failed for link sizes other than 2. Rather than cut the whole test out,
2298            I have added a new /Z option to pcretest that replaces the length and
2299            offset values with spaces. This is now used to make test 3 independent
2300            of link size. (Test 2 will be tidied up later.)
2302    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2303        segmentation fault instead of returning the appropriate error message.
2305    15. In multiline mode when the newline sequence was set to "any", the pattern
2306        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2307        This doesn't seem right; it now treats the CRLF combination as the line
2308        ending, and so does not match in that case. It's only a pattern such as ^$
2309        that would hit this one: something like ^ABC$ would have failed after \r
2310        and then tried again after \r\n.
2312    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2313        in an attempt to make files that differ only in their line terminators
2314        compare equal. This works on Linux.
2316    17. Under certain error circumstances pcregrep might try to free random memory
2317        as it exited. This is now fixed, thanks to valgrind.
2319    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2320        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2321        was because its rules for how to advance for /g after matching an empty
2322        string at the end of a line did not allow for this case. They now check for
2323        it specially.
2325    20. pcretest is supposed to handle patterns and data of any length, by
2326        extending its buffers when necessary. It was getting this wrong when the
2327        buffer for a data line had to be extended.
2329    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2330        CRLF as a newline sequence.
2332    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2333        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2334        I have nevertheless tidied it up.
2336    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2338    24. Added a man page for pcre-config.
2341    Version 7.0 19-Dec-06
2342    ---------------------
2344     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2345        moving to gcc 4.1.1.
2347     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2348        sys/time.h, which is documented as needed for this function. It doesn't
2349        seem to matter on Linux, but it showed up on some releases of OS X.
2351     3. It seems that there are systems where bytes whose values are greater than
2352        127 match isprint() in the "C" locale. The "C" locale should be the
2353        default when a C program starts up. In most systems, only ASCII printing
2354        characters match isprint(). This difference caused the output from pcretest
2355        to vary, making some of the tests fail. I have changed pcretest so that:
2357        (a) When it is outputting text in the compiled version of a pattern, bytes
2358            other than 32-126 are always shown as hex escapes.
2360        (b) When it is outputting text that is a matched part of a subject string,
2361            it does the same, unless a different locale has been set for the match
2362            (using the /L modifier). In this case, it uses isprint() to decide.
2364     4. Fixed a major bug that caused incorrect computation of the amount of memory
2365        required for a compiled pattern when options that changed within the
2366        pattern affected the logic of the preliminary scan that determines the
2367        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2368        that the computed length was too small. The symptoms of this bug were
2369        either the PCRE error "internal error: code overflow" from pcre_compile(),
2370        or a glibc crash with a message such as "pcretest: free(): invalid next
2371        size (fast)". Examples of patterns that provoked this bug (shown in
2372        pcretest format) are:
2374          /(?-x: )/x
2375          /(?x)(?-x: \s*#\s*)/
2376          /((?i)[\x{c0}])/8
2377          /(?i:[\x{c0}])/8
2379        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2380        is now done differently.
2382     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2383        wrapper classes; (b) implement a new function in the C++ scanner that is
2384        more efficient than the old way of doing things because it avoids levels of
2385        recursion in the regex matching; (c) add a paragraph to the documentation
2386        for the FullMatch() function.
2388     6. The escape sequence \n was being treated as whatever was defined as
2389        "newline". Not only was this contrary to the documentation, which states
2390        that \n is character 10 (hex 0A), but it also went horribly wrong when
2391        "newline" was defined as CRLF. This has been fixed.
2393     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2394        was being set to -1 for the "end of line" case (supposedly a value that no
2395        character can have). Though this value is never used (the check for end of
2396        line is "zero bytes in current character"), it caused compiler complaints.
2397        I've changed it to 0xffffffff.
2399     8. In pcre_version.c, the version string was being built by a sequence of
2400        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2401        string (as it is for production releases) called a macro with an empty
2402        argument. The C standard says the result of this is undefined. The gcc
2403        compiler treats it as an empty string (which was what was wanted) but it is
2404        reported that Visual C gives an error. The source has been hacked around to
2405        avoid this problem.
2407     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2408        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2409        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2410        of them did).
2412    10. Originally, pcretest opened its input and output without "b"; then I was
2413        told that "b" was needed in some environments, so it was added for release
2414        5.0 to both the input and output. (It makes no difference on Unix-like
2415        systems.) Later I was told that it is wrong for the input on Windows. I've
2416        now abstracted the modes into two macros, to make it easier to fiddle with
2417        them, and removed "b" from the input mode under Windows.
2419    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2421    12. Added -help and --help to pcretest as an official way of being reminded
2422        of the options.
2424    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2425        and pcrecpp.cc because they annoy compilers at high warning levels.
2427    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2429    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2430        scripts are not C programs :-) and which was not noticed because it works
2431        on Linux.
2433    16. pcretest is supposed to handle any length of pattern and data line (as one
2434        line or as a continued sequence of lines) by extending its input buffer if
2435        necessary. This feature was broken for very long pattern lines, leading to
2436        a string of junk being passed to pcre_compile() if the pattern was longer
2437        than about 50K.
2439    17. I have done a major re-factoring of the way pcre_compile() computes the
2440        amount of memory needed for a compiled pattern. Previously, there was code
2441        that made a preliminary scan of the pattern in order to do this. That was
2442        OK when PCRE was new, but as the facilities have expanded, it has become
2443        harder and harder to keep it in step with the real compile phase, and there
2444        have been a number of bugs (see for example, 4 above). I have now found a
2445        cunning way of running the real compile function in a "fake" mode that
2446        enables it to compute how much memory it would need, while actually only
2447        ever using a few hundred bytes of working memory and without too many
2448        tests of the mode. This should make future maintenance and development
2449        easier. A side effect of this work is that the limit of 200 on the nesting
2450        depth of parentheses has been removed (though this was never a serious
2451        limitation, I suspect). However, there is a downside: pcre_compile() now
2452        runs more slowly than before (30% or more, depending on the pattern). I
2453        hope this isn't a big issue. There is no effect on runtime performance.
2455    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2456        newline (only possible for the last line of a file) and it was a
2457        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2459    19. Added additional timing features to pcretest. (1) The -tm option now times
2460        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2461        separate command line item, by a number that specifies the number of
2462        repeats to use when timing. The default is 50000; this gives better
2463        precision, but takes uncomfortably long for very large patterns.
2465    20. Extended pcre_study() to be more clever in cases where a branch of a
2466        subpattern has no definite first character. For example, (a*|b*)[cd] would
2467        previously give no result from pcre_study(). Now it recognizes that the
2468        first character must be a, b, c, or d.
2470    21. There was an incorrect error "recursive call could loop indefinitely" if
2471        a subpattern (or the entire pattern) that was being tested for matching an
2472        empty string contained only one non-empty item after a nested subpattern.
2473        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2474        incorrectly, because the \d was being skipped in the check.
2476    22. The pcretest program now has a new pattern option /B and a command line
2477        option -b, which is equivalent to adding /B to every pattern. This causes
2478        it to show the compiled bytecode, without the additional information that
2479        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2480        is the same as /B/I).
2482    23. A new optimization is now able automatically to treat some sequences such
2483        as a*b as a*+b. More specifically, if something simple (such as a character
2484        or a simple class like \d) has an unlimited quantifier, and is followed by
2485        something that cannot possibly match the quantified thing, the quantifier
2486        is automatically "possessified".
2488    24. A recursive reference to a subpattern whose number was greater than 39
2489        went wrong under certain circumstances in UTF-8 mode. This bug could also
2490        have affected the operation of pcre_study().
2492    25. Realized that a little bit of performance could be had by replacing
2493        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2495    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2497    27. Possessive quantifiers such as a++ were previously implemented by turning
2498        them into atomic groups such as ($>a+). Now they have their own opcodes,
2499        which improves performance. This includes the automatically created ones
2500        from 23 above.
2502    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2503        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2504        the first matched character to be a colon. This applied both to named and
2505        numbered groups.
2507    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2509    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2510        building PCRE on Windows easier, so I have included it in the distribution.
2512    31. There is now a check in pcretest against a ridiculously large number being
2513        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2514        loop, the loop is abandoned.
2516    32. Forward references to subpatterns in conditions such as (?(2)...) where
2517        subpattern 2 is defined later cause pcre_compile() to search forwards in
2518        the pattern for the relevant set of parentheses. This search went wrong
2519        when there were unescaped parentheses in a character class, parentheses
2520        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2522    33. "Subroutine" calls and backreferences were previously restricted to
2523        referencing subpatterns earlier in the regex. This restriction has now
2524        been removed.
2526    34. Added a number of extra features that are going to be in Perl 5.10. On the
2527        whole, these are just syntactic alternatives for features that PCRE had
2528        previously implemented using the Python syntax or my own invention. The
2529        other formats are all retained for compatibility.
2531        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2532            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2533            also .NET compatible.
2535        (b) A recursion or subroutine call to a named group can now be defined as
2536            (?&name) as well as (?P>name).
2538        (c) A backreference to a named group can now be defined as \k<name> or
2539            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2540            5.10, are also .NET compatible.
2542        (d) A conditional reference to a named group can now use the syntax
2543            (?(<name>) or (?('name') as well as (?(name).
2545        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2546            groups (named and numbered) that are never evaluated inline, but can be
2547            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2548            is always false. There may be only one alternative in such a group.
2550        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2551            as the simple (?(R). The condition is true only if the most recent
2552            recursion is that of the given number or name. It does not search out
2553            through the entire recursion stack.
2555        (g) The escape \gN or \g{N} has been added, where N is a positive or
2556            negative number, specifying an absolute or relative reference.
2558    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2559        some "unreachable code" warnings.
2561    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2562        things, this adds five new scripts.
2564    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2565        There were also incompatibilities regarding the handling of \Q..\E inside
2566        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2567        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2569    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2570        matches an empty string, and forcibly breaks the loop. There were bugs in
2571        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2572        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2573        separate and independent bugs (that affected different cases) have been
2574        fixed.
2576    39. Refactored the code to abolish the use of different opcodes for small
2577        capturing bracket numbers. This is a tidy that I avoided doing when I
2578        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2579        The new approach is not only tidier, it makes it possible to reduce the
2580        memory needed to fix the previous bug (38).
2582    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2583        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2584        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2585        mode.
2587    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2588        report.
2590    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2591        copy construction and assignment in the C++ wrapper.
2593    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2594        couple of bugs that could have given wrong results in the "--newline=crlf"
2595        case.
2597    44. Added a number of casts and did some reorganization of signed/unsigned int
2598        variables following suggestions from Dair Grant. Also renamed the variable
2599        "this" as "item" because it is a C++ keyword.
2601    45. Arranged for dftables to add
2603          #include "pcre_internal.h"
2605        to pcre_chartables.c because without it, gcc 4.x may remove the array
2606        definition from the final binary if PCRE is built into a static library and
2607        dead code stripping is activated.
2609    46. For an unanchored pattern, if a match attempt fails at the start of a
2610        newline sequence, and the newline setting is CRLF or ANY, and the next two
2611        characters are CRLF, advance by two characters instead of one.
2614    Version 6.7 04-Jul-06
2615    ---------------------
2617     1. In order to handle tests when input lines are enormously long, pcretest has
2618        been re-factored so that it automatically extends its buffers when
2619        necessary. The code is crude, but this _is_ just a test program. The
2620        default size has been increased from 32K to 50K.
2622     2. The code in pcre_study() was using the value of the re argument before
2623        testing it for NULL. (Of course, in any sensible call of the function, it
2624        won't be NULL.)
2626     3. The memmove() emulation function in pcre_internal.h, which is used on
2627        systems that lack both memmove() and bcopy() - that is, hardly ever -
2628        was missing a "static" storage class specifier.
2630     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2631        containing an extended class (one that cannot be represented by a bitmap
2632        because it contains high-valued characters or Unicode property items, e.g.
2633        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2634        pattern, but PCRE should not loop if you do not (it no longer does).
2635        [Detail: two cases were found: (a) a repeated subpattern containing an
2636        extended class; (b) a recursive reference to a subpattern that followed a
2637        previous extended class. It wasn't skipping over the extended class
2638        correctly when UTF-8 mode was not set.]
2640     5. A negated single-character class was not being recognized as fixed-length
2641        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2642        compile error "lookbehind assertion is not fixed length".
2644     6. The RunPerlTest auxiliary script was showing an unexpected difference
2645        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2646        write a Perl script that can interpret lines of an input file either as
2647        byte characters or as UTF-8, which is what "perltest" was being required to
2648        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2649        can't do is switch easily at run time between having the "use utf8;" pragma
2650        or not. In the end, I fudged it by using the RunPerlTest script to insert
2651        "use utf8;" explicitly for the UTF-8 tests.
2653     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2654        the end of the subject string, contrary to the documentation and to what
2655        Perl does. This was true of both matching functions. Now it matches only at
2656        the start of the subject and immediately after *internal* newlines.
2658     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2659        a pointer to an int instead of a pointer to an unsigned long int. This
2660        caused problems on 64-bit systems.
2662     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2663        instance of the 'standard' template library not being so standard".
2665    10. There was no check on the number of named subpatterns nor the maximum
2666        length of a subpattern name. The product of these values is used to compute
2667        the size of the memory block for a compiled pattern. By supplying a very
2668        long subpattern name and a large number of named subpatterns, the size
2669        computation could be caused to overflow. This is now prevented by limiting
2670        the length of names to 32 characters, and the number of named subpatterns
2671        to 10,000.
2673    11. Subpatterns that are repeated with specific counts have to be replicated in
2674        the compiled pattern. The size of memory for this was computed from the
2675        length of the subpattern and the repeat count. The latter is limited to
2676        65535, but there was no limit on the former, meaning that integer overflow
2677        could in principle occur. The compiled length of a repeated subpattern is
2678        now limited to 30,000 bytes in order to prevent this.
2680    12. Added the optional facility to have named substrings with the same name.
2682    13. Added the ability to use a named substring as a condition, using the
2683        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2684        are numbers (not recommended). Forward references are permitted.
2686    14. Added forward references in named backreferences (if you see what I mean).
2688    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2689        pattern could run off the end of the subject. For example, the pattern
2690        "(?s)(.{1,5})"8 did this with the subject "ab".
2692    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2693        PCRE_CASELESS was set when matching characters that were quantified with ?
2694        or *.
2696    17. A character class other than a single negated character that had a minimum
2697        but no maximum quantifier - for example [ab]{6,} - was not handled
2698        correctly by pce_dfa_exec(). It would match only one character.
2700    18. A valid (though odd) pattern that looked like a POSIX character
2701        class but used an invalid character after [ (for example [[,abc,]]) caused
2702        pcre_compile() to give the error "Failed: internal error: code overflow" or
2703        in some cases to crash with a glibc free() error. This could even happen if
2704        the pattern terminated after [[ but there just happened to be a sequence of
2705        letters, a binary zero, and a closing ] in the memory that followed.
2707    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2708        over the years. Originally (before any Unicode support), just the bottom 8
2709        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2710        output from "man perlunicode" includes this:
2712          The regular expression compiler produces polymorphic opcodes.  That
2713          is, the pattern adapts to the data and automatically switches to
2714          the Unicode character scheme when presented with Unicode data--or
2715          instead uses a traditional byte scheme when presented with byte
2716          data.
2718        Sadly, a wide octal escape does not cause a switch, and in a string with
2719        no other multibyte characters, these octal escapes are treated as before.
2720        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2721        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2722        Unicode string.
2724        I have not perpetrated such confusion in PCRE. Up till now, it took just
2725        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2726        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2727        translate to the appropriate multibyte character.
2729    29. Applied some refactoring to reduce the number of warnings from Microsoft
2730        and Borland compilers. This has included removing the fudge introduced
2731        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2732        a warning about an unused variable.
2734    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2735        characters since release 4.0, because Perl (from release 5.004) does not.
2736        [Or at least, is documented not to: some releases seem to be in conflict
2737        with the documentation.] However, when a pattern was studied with
2738        pcre_study() and all its branches started with \s, PCRE still included VT
2739        as a possible starting character. Of course, this did no harm; it just
2740        caused an unnecessary match attempt.
2742    22. Removed a now-redundant internal flag bit that recorded the fact that case
2743        dependency changed within the pattern. This was once needed for "required
2744        byte" processing, but is no longer used. This recovers a now-scarce options
2745        bit. Also moved the least significant internal flag bit to the most-
2746        significant bit of the word, which was not previously used (hangover from
2747        the days when it was an int rather than a uint) to free up another bit for
2748        the future.
2750    23. Added support for CRLF line endings as well as CR and LF. As well as the
2751        default being selectable at build time, it can now be changed at runtime
2752        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2753        specify that it is scanning data with non-default line endings.
2755    24. Changed the definition of CXXLINK to make it agree with the definition of
2756        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2758    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2759        recursions. This makes a big different to stack usage for some patterns.
2761    26. If a subpattern containing a named recursion or subroutine reference such
2762        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2763        the space required for the compiled pattern went wrong and gave too small a
2764        value. Depending on the environment, this could lead to "Failed: internal
2765        error: code overflow at offset 49" or "glibc detected double free or
2766        corruption" errors.
2768    27. Applied patches from Google (a) to support the new newline modes and (b) to
2769        advance over multibyte UTF-8 characters in GlobalReplace.
2771    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2772        difference for some implementation of PCRE in some Windows version.
2774    29. Added some extra testing facilities to pcretest:
2776        \q<number>   in a data line sets the "match limit" value
2777        \Q<number>   in a data line sets the "match recursion limt" value
2778        -S <number>  sets the stack size, where <number> is in megabytes
2780        The -S option isn't available for Windows.
2783    Version 6.6 06-Feb-06
2784    ---------------------
2786     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2787        in pcreposix.h. I have copied the definition from pcre.h.
2789     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2790        because pcre.h is no longer a built file.
2792     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2793        not normally included in the compiled code.
2796    Version 6.5 01-Feb-06
2797    ---------------------
2799     1. When using the partial match feature with pcre_dfa_exec(), it was not
2800        anchoring the second and subsequent partial matches at the new starting
2801        point. This could lead to incorrect results. For example, with the pattern
2802        /1234/, partially matching against "123" and then "a4" gave a match.
2804     2. Changes to pcregrep:
2806        (a) All non-match returns from pcre_exec() were being treated as failures
2807            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2808            error message is output. Some extra information is given for the
2809            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2810            probably the only errors that are likely to be caused by users (by
2811            specifying a regex that has nested indefinite repeats, for instance).
2812            If there are more than 20 of these errors, pcregrep is abandoned.
2814        (b) A binary zero was treated as data while matching, but terminated the
2815            output line if it was written out. This has been fixed: binary zeroes
2816            are now no different to any other data bytes.
2818        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2819            used to set a locale for matching. The --locale=xxxx long option has
2820            been added (no short equivalent) to specify a locale explicitly on the
2821            pcregrep command, overriding the environment variables.
2823        (d) When -B was used with -n, some line numbers in the output were one less
2824            than they should have been.
2826        (e) Added the -o (--only-matching) option.
2828        (f) If -A or -C was used with -c (count only), some lines of context were
2829            accidentally printed for the final match.
2831        (g) Added the -H (--with-filename) option.
2833        (h) The combination of options -rh failed to suppress file names for files
2834            that were found from directory arguments.
2836        (i) Added the -D (--devices) and -d (--directories) options.
2838        (j) Added the -F (--fixed-strings) option.
2840        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2842        (l) Added the --colo(u)r option.
2844        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2845            is not present by default.
2847     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2848        items such as (?R) or (?1), when the recursion could match a number of
2849        alternatives. If it matched one of the alternatives, but subsequently,
2850        outside the recursion, there was a failure, the code tried to back up into
2851        the recursion. However, because of the way PCRE is implemented, this is not
2852        possible, and the result was an incorrect result from the match.
2854        In order to prevent this happening, the specification of recursion has
2855        been changed so that all such subpatterns are automatically treated as
2856        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2858     4. I had overlooked the fact that, in some locales, there are characters for
2859        which isalpha() is true but neither isupper() nor islower() are true. In
2860        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2861        and ordfeminine) are like this. This affected the treatment of \w and \W
2862        when they appeared in character classes, but not when they appeared outside
2863        a character class. The bit map for "word" characters is now created
2864        separately from the results of isalnum() instead of just taking it from the
2865        upper, lower, and digit maps. (Plus the underscore character, of course.)
2867     5. The above bug also affected the handling of POSIX character classes such as
2868        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2869        permanent tables. Instead, the bit maps for such a class were previously
2870        created as the appropriate unions of the upper, lower, and digit bitmaps.
2871        Now they are created by subtraction from the [[:word:]] class, which has
2872        its own bitmap.
2874     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2875        It is created by subtracting the vertical space characters (\x09, \x0a,
2876        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2877        subtraction was done in the overall bitmap for a character class, meaning
2878        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2879        be recognized. This bug has been fixed.
2881     7. Patches from the folks at Google:
2883          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2884          real life, but is still worth protecting against".
2886          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2887          regular expressions".
2889          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2890          have it.
2892          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2893          "configure" and the latter not, in order to fix a problem somebody had
2894          with compiling the Arg class on HP-UX.
2896          (e) Improve the error-handling of the C++ wrapper a little bit.
2898          (f) New tests for checking recursion limiting.
2900     8. The pcre_memmove() function, which is used only if the environment does not
2901        have a standard memmove() function (and is therefore rarely compiled),
2902        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2903        returning a result (though PCRE never actually uses the result).
2905     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2906        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2907        returned instead of calling malloc() with an overflowing number that would
2908        most likely cause subsequent chaos.
2910    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2912    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2913        with this option is matched, the nmatch and pmatch options of regexec() are
2914        ignored.
2916    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2917        provided in case anyone wants to the the POSIX interface with UTF-8
2918        strings.
2920    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2921        C++ linking (needed for some HP-UX environments).
2923    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2924        (unused parameter) and in the pcre_printint() function (omitted "default"
2925        switch label when the default is to do nothing).
2927    15. Added some code to make it possible, when PCRE is compiled as a C++
2928        library, to replace subject pointers for pcre_exec() with a smart pointer
2929        class, thus making it possible to process discontinuous strings.
2931    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2932        much the same function. They were added by different people who were trying
2933        to make PCRE easy to compile on non-Unix systems. It has been suggested
2934        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2935        for compiling on Windows systems. I have therefore replaced it with
2936        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2937        defaults to "extern" for C or "extern C" for C++, which works fine on
2938        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2939        SCOPE with something explicit in config.h. In addition:
2941        (a) pcreposix.h still had just "extern" instead of either of these macros;
2942            I have replaced it with PCRE_DATA_SCOPE.
2944        (b) Functions such as _pcre_xclass(), which are internal to the library,
2945            but external in the C sense, all had PCRE_EXPORT in their definitions.
2946            This is apparently wrong for the Windows case, so I have removed it.
2947            (It makes no difference on Unix-like systems.)
2949    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2950        of recursive calls to match(). This is different to MATCH_LIMIT because
2951        that limits the total number of calls to match(), not all of which increase
2952        the depth of recursion. Limiting the recursion depth limits the amount of
2953        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2954        when PCRE is compiled, and changed at run time. A patch from Google adds
2955        this functionality to the C++ interface.
2957    18. Changes to the handling of Unicode character properties:
2959        (a) Updated the table to Unicode 4.1.0.
2961        (b) Recognize characters that are not in the table as "Cn" (undefined).
2963        (c) I revised the way the table is implemented to a much improved format
2964            which includes recognition of ranges. It now supports the ranges that
2965            are defined in UnicodeData.txt, and it also amalgamates other
2966            characters into ranges. This has reduced the number of entries in the
2967            table from around 16,000 to around 3,000, thus reducing its size
2968            considerably. I realized I did not need to use a tree structure after
2969            all - a binary chop search is just as efficient. Having reduced the
2970            number of entries, I extended their size from 6 bytes to 8 bytes to
2971            allow for more data.
2973        (d) Added support for Unicode script names via properties such as \p{Han}.
2975    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2976        matching that character.
2978    20. When matching a repeated Unicode property with a minimum greater than zero,
2979        (for example \pL{2,}), PCRE could look past the end of the subject if it
2980        reached it while seeking the minimum number of characters. This could
2981        happen only if some of the characters were more than one byte long, because
2982        there is a check for at least the minimum number of bytes.
2984    21. Refactored the implementation of \p and \P so as to be more general, to
2985        allow for more different types of property in future. This has changed the
2986        compiled form incompatibly. Anybody with saved compiled patterns that use
2987        \p or \P will have to recompile them.
2989    22. Added "Any" and "L&" to the supported property types.
2991    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2992        but give a compile time error if the value is greater than 0xff.
2994    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2995        accidentally not being installed or uninstalled.
2997    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2998        made were to insert the current release number. This seemed silly, because
2999        it made things harder for people building PCRE on systems that don't run
3000        "configure". I have turned pcre.h into a distributed file, no longer built
3001        by "configure", with the version identification directly included. There is
3002        no longer a pcre.h.in file.
3004        However, this change necessitated a change to the pcre-config script as
3005        well. It is built from pcre-config.in, and one of the substitutions was the
3006        release number. I have updated configure.ac so that ./configure now finds
3007        the release number by grepping pcre.h.
3009    26. Added the ability to run the tests under valgrind.
3012    Version 6.4 05-Sep-05
3013    ---------------------
3015     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3016        "--" to be printed when multiple files were scanned, even when none of the
3017        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3018        consider it to be a bug, and have restored the previous behaviour.
3020     2. A couple of code tidies to get rid of compiler warnings.
3022     3. The pcretest program used to cheat by referring to symbols in the library
3023        whose names begin with _pcre_. These are internal symbols that are not
3024        really supposed to be visible externally, and in some environments it is
3025        possible to suppress them. The cheating is now confined to including
3026        certain files from the library's source, which is a bit cleaner.
3028     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3029        file's purpose clearer.
3031     5. Reorganized pcre_ucp_findchar().
3034    Version 6.3 15-Aug-05
3035    ---------------------
3037     1. The file libpcre.pc.in did not have general read permission in the tarball.
3039     2. There were some problems when building without C++ support:
3041        (a) If C++ support was not built, "make install" and "make test" still
3042            tried to test it.
3044        (b) There were problems when the value of CXX was explicitly set. Some
3045            changes have been made to try to fix these, and ...
3047        (c) --disable-cpp can now be used to explicitly disable C++ support.
3049        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3050            backslash in a target when C++ was disabled. This confuses some
3051            versions of "make", apparently. Using an intermediate variable solves
3052            this. (Same for CPP_LOBJ.)
3054     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3055        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3056        necessary on certain architectures.
3058     4. Added a setting of -export-symbols-regex to the link command to remove
3059        those symbols that are exported in the C sense, but actually are local
3060        within the library, and not documented. Their names all begin with
3061        "_pcre_". This is not a perfect job, because (a) we have to except some
3062        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3063        available (and never for static libraries). I have made a note to try to
3064        find a way round (a) in the future.
3067    Version 6.2 01-Aug-05
3068    ---------------------
3070     1. There was no test for integer overflow of quantifier values. A construction
3071        such as {1111111111111111} would give undefined results. What is worse, if
3072        a minimum quantifier for a parenthesized subpattern overflowed and became
3073        negative, the calculation of the memory size went wrong. This could have
3074        led to memory overwriting.
3076     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3078     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3079        operating environments where this matters.
3081     4. Applied Giuseppe Maxia's patch to add additional features for controlling
3082        PCRE options from within the C++ wrapper.
3084     5. Named capturing subpatterns were not being correctly counted when a pattern
3085        was compiled. This caused two problems: (a) If there were more than 100
3086        such subpatterns, the calculation of the memory needed for the whole
3087        compiled pattern went wrong, leading to an overflow error. (b) Numerical
3088        back references of the form \12, where the number was greater than 9, were
3089        not recognized as back references, even though there were sufficient
3090        previous subpatterns.
3092     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3093        versions of gcc, e.g. 2.95.4.
3096    Version 6.1 21-Jun-05
3097    ---------------------
3099     1. There was one reference to the variable "posix" in pcretest.c that was not
3100        surrounded by "#if !defined NOPOSIX".
3102     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3103        the cross-check on the old pcre_info() function, for the benefit of the
3104        cut-down version of PCRE that is currently imported into Exim.
3106     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3107        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3108        patterns that start (?i) but I don't think that matters. The use of (?i) is
3109        just an example; this all applies to the other options as well.
3111     4. Since libtool seems to echo the compile commands it is issuing, the output
3112        from "make" can be reduced a bit by putting "@" in front of each libtool
3113        compile command.
3115     5. Patch from the folks at Google for configure.in to be a bit more thorough
3116        in checking for a suitable C++ installation before trying to compile the
3117        C++ stuff. This should fix a reported problem when a compiler was present,
3118        but no suitable headers.
3120     6. The man pages all had just "PCRE" as their title. I have changed them to
3121        be the relevant file name. I have also arranged that these names are
3122        retained in the file doc/pcre.txt, which is a concatenation in text format
3123        of all the man pages except the little individual ones for each function.
3125     7. The NON-UNIX-USE file had not been updated for the different set of source
3126        files that come with release 6. I also added a few comments about the C++
3127        wrapper.
3130    Version 6.0 07-Jun-05
3131    ---------------------
3133     1. Some minor internal re-organization to help with my DFA experiments.
3135     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3136        didn't matter for the library itself when fully configured, but did matter
3137        when compiling without UCP support, or within Exim, where the ucp files are
3138        not imported.
3140     3. Refactoring of the library code to split up the various functions into
3141        different source modules. The addition of the new DFA matching code (see
3142        below) to a single monolithic source would have made it really too
3143        unwieldy, quite apart from causing all the code to be include in a
3144        statically linked application, when only some functions are used. This is
3145        relevant even without the DFA addition now that patterns can be compiled in
3146        one application and matched in another.
3148        The downside of splitting up is that there have to be some external
3149        functions and data tables that are used internally in different modules of
3150        the library but which are not part of the API. These have all had their
3151        names changed to start with "_pcre_" so that they are unlikely to clash
3152        with other external names.
3154     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3155        a different (DFA) algorithm. Although it is slower than the original
3156        function, it does have some advantages for certain types of matching
3157        problem.
3159     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3160        including restarting after a partial match.
3162     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3163        defined when compiling for Windows was sent to me. I have put it into the
3164        code, though I have no means of testing or verifying it.
3166     7. Added the pcre_refcount() auxiliary function.
3168     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3169        match before or at the first newline in the subject string. In pcretest,
3170        the /f option on a pattern can be used to set this.
3172     9. A repeated \w when used in UTF-8 mode with characters greater than 256
3173        would behave wrongly. This has been present in PCRE since release 4.0.
3175    10. A number of changes to the pcregrep command:
3177        (a) Refactored how -x works; insert ^(...)$ instead of setting
3178            PCRE_ANCHORED and checking the length, in preparation for adding
3179            something similar for -w.
3181        (b) Added the -w (match as a word) option.
3183        (c) Refactored the way lines are read and buffered so as to have more
3184            than one at a time available.
3186        (d) Implemented a pcregrep test script.
3188        (e) Added the -M (multiline match) option. This allows patterns to match
3189            over several lines of the subject. The buffering ensures that at least
3190            8K, or the rest of the document (whichever is the shorter) is available
3191            for matching (and similarly the previous 8K for lookbehind assertions).
3193        (f) Changed the --help output so that it now says
3195              -w, --word-regex(p)
3197            instead of two lines, one with "regex" and the other with "regexp"
3198            because that confused at least one person since the short forms are the
3199            same. (This required a bit of code, as the output is generated
3200            automatically from a table. It wasn't just a text change.)
3202        (g) -- can be used to terminate pcregrep options if the next thing isn't an
3203            option but starts with a hyphen. Could be a pattern or a path name
3204            starting with a hyphen, for instance.
3206        (h) "-" can be given as a file name to represent stdin.
3208        (i) When file names are being printed, "(standard input)" is used for
3209            the standard input, for compatibility with GNU grep. Previously
3210            "<stdin>" was used.
3212        (j) The option --label=xxx can be used to supply a name to be used for
3213            stdin when file names are being printed. There is no short form.
3215        (k) Re-factored the options decoding logic because we are going to add
3216            two more options that take data. Such options can now be given in four
3217            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3219        (l) Added the -A, -B, and -C options for requesting that lines of context
3220            around matches be printed.
3222        (m) Added the -L option to print the names of files that do not contain
3223            any matching lines, that is, the complement of -l.
3225        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3226            continue to scan other files.
3228        (o) The -s option was incorrectly implemented. For compatibility with other
3229            greps, it now suppresses the error message for a non-existent or non-
3230            accessible file (but not the return code). There is a new option called
3231            -q that suppresses the output of matching lines, which was what -s was
3232            previously doing.
3234        (p) Added --include and --exclude options to specify files for inclusion
3235            and exclusion when recursing.
3237    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3238        Hopefully, it now does.
3240    12. Missing cast in pcre_study().
3242    13. Added an "uninstall" target to the makefile.
3244    14. Replaced "extern" in the function prototypes in Makefile.in with
3245        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3246        world, but is set differently for Windows.
3248    15. Added a second compiling function called pcre_compile2(). The only
3249        difference is that it has an extra argument, which is a pointer to an
3250        integer error code. When there is a compile-time failure, this is set
3251        non-zero, in addition to the error test pointer being set to point to an
3252        error message. The new argument may be NULL if no error number is required
3253        (but then you may as well call pcre_compile(), which is now just a
3254        wrapper). This facility is provided because some applications need a
3255        numeric error indication, but it has also enabled me to tidy up the way
3256        compile-time errors are handled in the POSIX wrapper.
3258    16. Added VPATH=.libs to the makefile; this should help when building with one
3259        prefix path and installing with another. (Or so I'm told by someone who
3260        knows more about this stuff than I do.)
3262    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3263        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3264        match everything, including newlines. This is not POSIX-compatible, but
3265        somebody wanted the feature. From pcretest it can be activated by using
3266        both the P and the s flags.
3268    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3270    19. libpcre.pc was being incorrectly installed as executable.
3272    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3273        it now also looks for '\r' so that it will work unmodified on Windows.
3275    21. Added Google's contributed C++ wrapper to the distribution.
3277    22. Added some untidy missing memory free() calls in pcretest, to keep
3278        Electric Fence happy when testing.
3282    Version 5.0 13-Sep-04
3283    ---------------------
3285     1. Internal change: literal characters are no longer packed up into items
3286        containing multiple characters in a single byte-string. Each character
3287        is now matched using a separate opcode. However, there may be more than one
3288        byte in the character in UTF-8 mode.
3290     2. The pcre_callout_block structure has two new fields: pattern_position and
3291        next_item_length. These contain the offset in the pattern to the next match
3292        item, and its length, respectively.
3294     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3295        insertion of callouts before each pattern item. Added the /C option to
3296        pcretest to make use of this.
3298     4. On the advice of a Windows user, the lines
3300          #if defined(_WIN32) || defined(WIN32)
3301          _setmode( _fileno( stdout ), 0x8000 );
3302          #endif  /* defined(_WIN32) || defined(WIN32) */
3304        have been added to the source of pcretest. This apparently does useful
3305        magic in relation to line terminators.
3307     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3308        for the benefit of those environments where the "b" makes a difference.
3310     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3311        to know about it. I have put a hack into configure.in that adds in code
3312        to set GCC=yes if CC=icc. This seems to end up at a point in the
3313        generated configure script that is early enough to affect the setting of
3314        compiler options, which is what is needed, but I have no means of testing
3315        whether it really works. (The user who reported this had patched the
3316        generated configure script, which of course I cannot do.)
3318        LATER: After change 22 below (new libtool files), the configure script
3319        seems to know about icc (and also ecc). Therefore, I have commented out
3320        this hack in configure.in.
3322     7. Added support for pkg-config (2 patches were sent in).
3324     8. Negated POSIX character classes that used a combination of internal tables
3325        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3326        [[:^ascii]]. Typically, they would match almost any characters. The other
3327        POSIX classes were not broken in this way.
3329     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3330        to find the match, as PCRE was deluded into thinking that the match had to
3331        start at the start point or following a newline. The same bug applied to
3332        patterns with negative forward assertions or any backward assertions
3333        preceding ".*" at the start, unless the pattern required a fixed first
3334        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3336    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3337        starting at the last subject character, bytes beyond the end of the subject
3338        string were read.
3340    11. Renamed the variable "class" as "classbits" to make life easier for C++
3341        users. (Previously there was a macro definition, but it apparently wasn't
3342        enough.)
3344    12. Added the new field "tables" to the extra data so that tables can be passed
3345        in at exec time, or the internal tables can be re-selected. This allows
3346        a compiled regex to be saved and re-used at a later time by a different
3347        program that might have everything at different addresses.
3349    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3350        -R library as well as a -L library.
3352    14. The debugging options of pcretest (-d on the command line or D on a
3353        pattern) showed incorrect output for anything following an extended class
3354        that contained multibyte characters and which was followed by a quantifier.
3356    15. Added optional support for general category Unicode character properties
3357        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3358        support. It adds about 90K to the size of the library. The meanings of the
3359        inbuilt class escapes such as \d and \s have NOT been changed.
3361    16. Updated pcredemo.c to include calls to free() to release the memory for the
3362        compiled pattern.
3364    17. The generated file chartables.c was being created in the source directory
3365        instead of in the building directory. This caused the build to fail if the
3366        source directory was different from the building directory, and was
3367        read-only.
3369    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3370        file. No doubt somebody will tell me if they don't make sense... Also added
3371        Dan Mooney's comments about building on OpenVMS.
3373    19. Added support for partial matching via the PCRE_PARTIAL option for
3374        pcre_exec() and the \P data escape in pcretest.
3376    20. Extended pcretest with 3 new pattern features:
3378        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3379              write the compiled pattern to the file whose name is "rest-of-line".
3380              This is a straight binary dump of the data, with the saved pointer to
3381              the character tables forced to be NULL. The study data, if any, is
3382              written too. After writing, pcretest reads a new pattern.
3384        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3385              compiled pattern from the given file. There must not be any
3386              occurrences of "<" in the file name (pretty unlikely); if there are,
3387              pcretest will instead treat the initial "<" as a pattern delimiter.
3388              After reading in the pattern, pcretest goes on to read data lines as
3389              usual.
3391        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3392              and 16-bit fields in a compiled pattern, to simulate a pattern that
3393              was compiled on a host of opposite endianness.
3395    21. The pcre-exec() function can now cope with patterns that were compiled on
3396        hosts of opposite endianness, with this restriction:
3398          As for any compiled expression that is saved and used later, the tables
3399          pointer field cannot be preserved; the extra_data field in the arguments
3400          to pcre_exec() should be used to pass in a tables address if a value
3401          other than the default internal tables were used at compile time.
3403    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3404        now diagnosed as an error. Previously, most of the time, a negative number
3405        would have been treated as zero, but if in addition "ovector" was passed as
3406        NULL, a crash could occur.
3408    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3409        new versions from the libtool 1.5 distribution (the last one is a copy of
3410        a file called libtool.m4). This seems to have fixed the need to patch
3411        "configure" to support Darwin 1.3 (which I used to do). However, I still
3412        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3413        workstation).
3415    24. Changed the PCRE licence to be the more standard "BSD" licence.
3418    Version 4.5 01-Dec-03
3419    ---------------------
3421     1. There has been some re-arrangement of the code for the match() function so
3422        that it can be compiled in a version that does not call itself recursively.
3423        Instead, it keeps those local variables that need separate instances for
3424        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3425        needs to "recurse". Keeping track of where control must go is done by means
3426        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3427        hide most of the details from the main code, and operates only if
3428        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3429        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3430        operating.
3432        To make it easier for callers to provide specially tailored get/free
3433        functions for this usage, two new functions, pcre_stack_malloc, and
3434        pcre_stack_free, are used. They are always called in strict stacking order,
3435        and the size of block requested is always the same.
3437        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3438        PCRE has been compiled to use the stack or the heap for recursion. The
3439        -C option of pcretest uses this to show which version is compiled.
3441        A new data escape \S, is added to pcretest; it causes the amounts of store
3442        obtained and freed by both kinds of malloc/free at match time to be added
3443        to the output.
3445     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3446        what's available on my current Linux desktop machine.
3448     3. When matching a UTF-8 string, the test for a valid string at the start has
3449        been extended. If start_offset is not zero, PCRE now checks that it points
3450        to a byte that is the start of a UTF-8 character. If not, it returns
3451        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3452        this is necessary because there may be backward assertions in the pattern.
3453        When matching the same subject several times, it may save resources to use
3454        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3456     4. The code for checking the validity of UTF-8 strings has been tightened so
3457        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3458        containing "overlong sequences".
3460     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3461        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3462        should have been "&", but it just so happened that all the cases this let
3463        through by mistake were picked up later in the function.
3465     6. I had used a variable called "isblank" - this is a C99 function, causing
3466        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3468     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3469        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3470        so the version is shown; (c) stop "make test" echoing "./RunTest".
3472     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3474     9. The prototype for memmove() for systems that don't have it was using
3475        size_t, but the inclusion of the header that defines size_t was later. I've
3476        moved the #includes for the C headers earlier to avoid this.
3478    10. Added some adjustments to the code to make it easier to compiler on certain
3479        special systems:
3481          (a) Some "const" qualifiers were missing.
3482          (b) Added the macro EXPORT before all exported functions; by default this
3483              is defined to be empty.
3484          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3485              that it reads its output file name as an argument instead of writing
3486              to the standard output and assuming this can be redirected.
3488    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3489        class containing characters with values greater than 255, PCRE compilation
3490        went into a loop.
3492    12. A recursive reference to a subpattern that was within another subpattern
3493        that had a minimum quantifier of zero caused PCRE to crash. For example,
3494        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3495        recursion. If the recursively-called subpattern itself had a zero repeat,
3496        that was OK.
3498    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3499        buffer into which it was copied (for escape processing) was still set at
3500        1024, so long lines caused crashes.
3502    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3503        "internal error: code overflow...". This applied to any character class
3504        that was followed by a possessive quantifier.
3506    15. Modified the Makefile to add libpcre.la as a prerequisite for
3507        libpcreposix.la because I was told this is needed for a parallel build to
3508        work.
3510    16. If a pattern that contained .* following optional items at the start was
3511        studied, the wrong optimizing data was generated, leading to matching
3512        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3513        matching string must start with a or b or c. The correct conclusion for
3514        this pattern is that a match can start with any character.
3517    Version 4.4 13-Aug-03
3518    ---------------------
3520     1. In UTF-8 mode, a character class containing characters with values between
3521        127 and 255 was not handled correctly if the compiled pattern was studied.
3522        In fixing this, I have also improved the studying algorithm for such
3523        classes (slightly).
3525     2. Three internal functions had redundant arguments passed to them. Removal
3526        might give a very teeny performance improvement.
3528     3. Documentation bug: the value of the capture_top field in a callout is *one
3529        more than* the number of the hightest numbered captured substring.
3531     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3532        in incorrectly linking with a previously installed version. They now link
3533        explicitly with libpcre.la.
3535     5. configure.in no longer needs to recognize Cygwin specially.
3537     6. A problem in pcre.in for Windows platforms is fixed.
3539     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3540        pcretest, it used to include the size of the study block as part of its
3541        output. Unfortunately, the structure contains a field that has a different
3542        size on different hardware architectures. This meant that the tests that
3543        showed this size failed. As the block is currently always of a fixed size,
3544        this information isn't actually particularly useful in pcretest output, so
3545        I have just removed it.
3547     8. Three pre-processor statements accidentally did not start in column 1.
3548        Sadly, there are *still* compilers around that complain, even though
3549        standard C has not required this for well over a decade. Sigh.
3551     9. In pcretest, the code for checking callouts passed small integers in the
3552        callout_data field, which is a void * field. However, some picky compilers
3553        complained about the casts involved for this on 64-bit systems. Now
3554        pcretest passes the address of the small integer instead, which should get
3555        rid of the warnings.
3557    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3558        both compile and run time, and gives an error if an invalid UTF-8 sequence
3559        is found. There is a option for disabling this check in cases where the
3560        string is known to be correct and/or the maximum performance is wanted.
3562    11. In response to a bug report, I changed one line in Makefile.in from
3564            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3565        to
3566            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3568        to look similar to other lines, but I have no way of telling whether this
3569        is the right thing to do, as I do not use Windows. No doubt I'll get told
3570        if it's wrong...
3573    Version 4.3 21-May-03
3574    ---------------------
3576    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3577       Makefile.
3579    2. Some refactoring to improve the quality of the code:
3581       (i)   The utf8_table... variables are now declared "const".
3583       (ii)  The code for \cx, which used the "case flipping" table to upper case
3584             lower case letters, now just substracts 32. This is ASCII-specific,
3585             but the whole concept of \cx is ASCII-specific, so it seems
3586             reasonable.
3588       (iii) PCRE was using its character types table to recognize decimal and
3589             hexadecimal digits in the pattern. This is silly, because it handles
3590             only 0-9, a-f, and A-F, but the character types table is locale-
3591             specific, which means strange things might happen. A private
3592             table is now used for this - though it costs 256 bytes, a table is
3593             much faster than multiple explicit tests. Of course, the standard
3594             character types table is still used for matching digits in subject
3595             strings against \d.
3597       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3598             ending in _t are). So I've renamed it as ESC_tee.
3600    3. The first argument for regexec() in the POSIX wrapper should have been
3601       defined as "const".
3603    4. Changed pcretest to use malloc() for its buffers so that they can be
3604       Electric Fenced for debugging.
3606    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3607       to read one or more bytes before the start of the subject string. Often this
3608       had no effect on PCRE's behaviour, but in some circumstances it could
3609       provoke a segmentation fault.
3611    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3612       to try to read one or more bytes before the start of the subject string.
3614    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3615       UTF-8 support could misbehave in various ways if the subject string
3616       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3617       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3618       back over UTF-8 characters.)
3621    Version 4.2 14-Apr-03
3622    ---------------------
3624    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3626    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3627         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3628         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3629         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3630         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3631           and BUILD_EXEEXT
3632         Note: automatic setting of the BUILD variables is not yet working
3633         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3634           compile-time but not at link-time
3635         [LINK]: use for linking executables only
3636         make different versions for Windows and non-Windows
3637         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3638           libraries
3639         [LINK_FOR_BUILD]: new variable
3640         [OBJEXT]: use throughout
3641         [EXEEXT]: use throughout
3642         <winshared>: new target
3643         <wininstall>: new target
3644         <dftables.o>: use native compiler
3645         <dftables>: use native linker
3646         <install>: handle Windows platform correctly
3647         <clean>: ditto
3648         <check>: ditto
3649         copy DLL to top builddir before testing
3651       As part of these changes, -no-undefined was removed again. This was reported
3652       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3653       in any case.
3655    3. Some tidies to get rid of compiler warnings:
3657       . In the match_data structure, match_limit was an unsigned long int, whereas
3658         match_call_count was an int. I've made them both unsigned long ints.
3660       . In pcretest the fact that a const uschar * doesn't automatically cast to
3661         a void * provoked a warning.
3663       . Turning on some more compiler warnings threw up some "shadow" variables
3664         and a few more missing casts.
3666    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3667       option, a class that contained a single character with a value between 128
3668       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3670    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3671       option, a class that contained several characters, but with at least one
3672       whose value was between 128 and 255 caused PCRE to crash.
3675    Version 4.1 12-Mar-03
3676    ---------------------
3678    1. Compiling with gcc -pedantic found a couple of places where casts were
3679    needed, and a string in dftables.c that was longer than standard compilers are
3680    required to support.
3682    2. Compiling with Sun's compiler found a few more places where the code could
3683    be tidied up in order to avoid warnings.
3685    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3686    first of these names is deprecated in the latest Autoconf in favour of the name
3687    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3688    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3689    analogy I have changed it to CFLAGS_FOR_BUILD.
3691    4. Added -no-undefined to the linking command in the Makefile, because this is
3692    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3693    linking step for the pcreposix library.
3695    5. PCRE was failing to diagnose the case of two named groups with the same
3696    name.
3698    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3699    literal character that is needed in the subject for a match, and scans along to
3700    ensure that it is present before embarking on the full matching process. This
3701    saves time in cases of nested unlimited repeats that are never going to match.
3702    Problem: the scan can take a lot of time if the subject is very long (e.g.
3703    megabytes), thus penalizing straightforward matches. It is now done only if the
3704    amount of subject to be scanned is less than 1000 bytes.
3706    7. A lesser problem with the same optimization is that it was recording the
3707    first character of an anchored pattern as "needed", thus provoking a search
3708    right along the subject, even when the first match of the pattern was going to
3709    fail. The "needed" character is now not set for anchored patterns, unless it
3710    follows something in the pattern that is of non-fixed length. Thus, it still
3711    fulfils its original purpose of finding quick non-matches in cases of nested
3712    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3715    Version 4.0 17-Feb-03
3716    ---------------------
3718    1. If a comment in an extended regex that started immediately after a meta-item
3719    extended to the end of string, PCRE compiled incorrect data. This could lead to
3720    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3722    2. Moved to autoconf 2.53 and libtool 1.4.2.
3724    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3725    the special perltest8 script is no longer needed - all the tests can be run
3726    from a single perltest script.
3728    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3729    by \s. It has now been removed in PCRE. This means it isn't recognized as
3730    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3731    class [:space:] *does* include VT, thereby creating a mess.
3733    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3734    space and tab.
3736    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3737    its new features into the main test script, reducing the number of scripts.
3739    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3740    were backward compatible, and made the (?i) apply to the whole pattern, as if
3741    /i were given. Now it behaves more logically, and applies the option setting
3742    only to what follows. PCRE has been changed to follow suit. However, if it
3743    finds options settings right at the start of the pattern, it extracts them into
3744    the global options, as before. Thus, they show up in the info data.
3746    8. Added support for the \Q...\E escape sequence. Characters in between are
3747    treated as literals. This is slightly different from Perl in that $ and @ are
3748    also handled as literals inside the quotes. In Perl, they will cause variable
3749    interpolation. Note the following examples:
3751        Pattern            PCRE matches      Perl matches
3753        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3754        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3755        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3757    For compatibility with Perl, \Q...\E sequences are recognized inside character
3758    classes as well as outside them.
3760    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3761    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3762    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3763    signed/unsigned warnings.
3765    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3766    option for pcretest, so I've replaced it by a simple function that does just
3767    that job.
3769    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3770    "pcregrep -".
3772    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3773    Java package. This provides some syntactic sugar for simple cases of what my
3774    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3775    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3776    item, you can use this simplified notation. Note that only makes sense with
3777    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3778    greediness, whatever the setting of the PCRE_UNGREEDY option.
3780    13. A change of greediness default within a pattern was not taking effect at
3781    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3782    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3783    was abstracted outside.
3785    14. PCRE now supports the \G assertion. It is true when the current matching
3786    position is at the start point of the match. This differs from \A when the
3787    starting offset is non-zero. Used with the /g option of pcretest (or similar
3788    code), it works in the same way as it does for Perl's /g opti