ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 45 by nigel, Sat Feb 24 21:39:25 2007 UTC revision 1324 by ph10, Fri May 10 11:40:06 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.33 28-April-2013
5    --------------------------
7    1.  Added 'U' to some constants that are compared to unsigned integers, to
8        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
9        variables that are added to signed variables, to ensure the result is
10        signed and can be negated.
12    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
13        CMake config files.
15    3.  Revise the creation of config.h.generic so that all boolean macros are
16        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
17        overriding via -D on the command line possible.
19    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
20        to unsigned int is reported to make a quite noticeable speed difference in
21        a specific Windows environment. Testing on Linux did also appear to show
22        some benefit (and it is clearly not harmful). Also fixed the definition of
23        Xop which should be unsigned.
25    5.  Related to (4), changing the definition of the intermediate variable cc
26        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
27        improvements.
29    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
30        unnecessary spaces.
32    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
33        and later.
35    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
37    9.  Optimizing fast_forward_start_bits in JIT.
39    10. Adding experimental support for callouts in JIT, and fixing some
40        issues revealed during this work. Namely:
42        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
44        (b) Minimum length was not checked before the matching is started.
46    11. The value of capture_last that is passed to callouts was incorrect in some
47        cases when there was a capture on one path that was subsequently abandoned
48        after a backtrack. Also, the capture_last value is now reset after a
49        recursion, since all captures are also reset in this case.
51    12. The interpreter no longer returns the "too many substrings" error in the
52        case when an overflowing capture is in a branch that is subsequently
53        abandoned after a backtrack.
55    13. In the pathological case when an offset vector of size 2 is used, pcretest
56        now prints out the matched string after a yield of 0 or 1.
58    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
59        Only supported by the JIT compiler at the moment.
61    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
63    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
64        offset of the starting point of the matching process, provided the offsets
65        vector is large enough.
67    17. The \A escape now records a lookbehind value of 1, though its execution
68        does not actually inspect the previous character. This is to ensure that,
69        in partial multi-segment matching, at least one character from the old
70        segment is retained when a new segment is processed. Otherwise, if there
71        are no lookbehinds in the pattern, \A might match incorrectly at the start
72        of a new segment.
74    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
76    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
77        gives some modest performance improvement in 8-bit mode.
79    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
80        be expressed in certain programming languages using Universal Character
81        Names.
83    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
84        which points out that "non characters" are not "characters that may not
85        appear in Unicode strings" but rather "characters that are reserved for
86        internal use and have only local meaning".
88    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
89        there was a conditional group that depended on an assertion, if the
90        assertion was false, the callout that immediately followed the alternation
91        in the condition was skipped when pcre_exec() was used for matching.
93    23. Allow an explicit callout to be inserted before an assertion that is the
94        condition for a conditional group, for compatibility with automatic
95        callouts, which always insert a callout at this point.
97    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
98        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
100    25. (*PRUNE) is now supported by the JIT compiler.
102    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
104    27. Fix the case where there are two or more SKIPs with arguments that may be
105        ignored.
107    28. (*SKIP) is now supported by the JIT compiler.
109    29. (*THEN) is now supported by the JIT compiler.
111    30. Update RunTest with additional test selector options.
113    31. The way PCRE handles backtracking verbs has been changed in two ways.
115        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
116        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
117        backtracking. In some cases this makes it more Perl-compatible, but Perl's
118        rather obscure rules do not always do the same thing.
120        (2) Previously, backtracking verbs were confined within assertions. This is
121        no longer the case for positive assertions, except for (*ACCEPT). Again,
122        this sometimes improves Perl compatibility, and sometimes does not.
124    32. A number of tests that were in test 2 because Perl did things differently
125        have been moved to test 1, because either Perl or PCRE has changed, and
126        these tests are now compatible.
128    32. Control verbs are handled in the same way in JIT and interpreter.
130    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
131        contained a forward subroutine reference caused a compile error.
133    34. Auto-detect and optimize limited repetitions in JIT.
135    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
136        blocking (*UTF) etc.
138    36. In the interpreter, maximizing pattern repetitions for characters and
139        character types now use tail recursion, which reduces stack usage.
141    37. The value of the max lookbehind was not correctly preserved if a compiled
142        and saved regex was reloaded on a host of different endianness.
144    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
145        of the compiled pattern block, expand the flags field from 16 to 32 bits
146        because it was almost full.
148    39. Try madvise first before posix_madvise.
150    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
151        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
152        original change locked out all matches of empty strings. This has been
153        changed so that one match of an empty string per line is recognized.
154        Subsequent searches on the same line (for colouring or for --only-matching,
155        for example) do not recognize empty strings.
158    Version 8.32 30-November-2012
159    -----------------------------
161    1.  Improved JIT compiler optimizations for first character search and single
162        character iterators.
164    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
165        Patch by Daniel Richard G.
167    3.  Single character iterator optimizations in the JIT compiler.
169    4.  Improved JIT compiler optimizations for character ranges.
171    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
172        Reported by Giuseppe D'Angelo.
174    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
175        of a line, was being set incorrectly in cases where .* appeared inside
176        atomic brackets at the start of a pattern, or where there was a subsequent
177        *PRUNE or *SKIP.
179    7.  Improved instruction cache flush for POWER/PowerPC.
180        Patch by Daniel Richard G.
182    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
183        grep:
185        (a) There is now no limit to the number of patterns to be matched.
187        (b) An error is given if a pattern is too long.
189        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
190            are now supported.
192        (d) --exclude-from and --include-from (multiple use) have been added.
194        (e) Exclusions and inclusions now apply to all files and directories, not
195            just to those obtained from scanning a directory recursively.
197        (f) Multiple uses of -f and --file-list are now supported.
199        (g) In a Windows environment, the default for -d has been changed from
200            "read" (the GNU grep default) to "skip", because otherwise the presence
201            of a directory in the file list provokes an error.
203        (h) The documentation has been revised and clarified in places.
205    9.  Improve the matching speed of capturing brackets.
207    10. Changed the meaning of \X so that it now matches a Unicode extended
208        grapheme cluster.
210    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
211        out POSIX threads when JIT support is configured.
213    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
215    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
216        structure could go wrong in environments where size_t is not the same size
217        as int.
219    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
220        set.
222    15. The EBCDIC support had decayed; later updates to the code had included
223        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
224        general tidy up of EBCDIC-related issues, and the documentation was also
225        not quite right. There is now a test that can be run on ASCII systems to
226        check some of the EBCDIC-related things (but is it not a full test).
228    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
229        in a small tidy to the code.
231    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
233    18. If the --only-matching (-o) option in pcregrep is specified multiple
234        times, each one causes appropriate output. For example, -o1 -o2 outputs the
235        substrings matched by the 1st and 2nd capturing parentheses. A separating
236        string can be specified by --om-separator (default empty).
238    19. Improving the first n character searches.
240    20. Turn case lists for horizontal and vertical white space into macros so that
241        they are defined only once.
243    21. This set of changes together give more compatible Unicode case-folding
244        behaviour for characters that have more than one other case when UCP
245        support is available.
247        (a) The Unicode property table now has offsets into a new table of sets of
248            three or more characters that are case-equivalent. The MultiStage2.py
249            script that generates these tables (the pcre_ucd.c file) now scans
250            CaseFolding.txt instead of UnicodeData.txt for character case
251            information.
253        (b) The code for adding characters or ranges of characters to a character
254            class has been abstracted into a generalized function that also handles
255            case-independence. In UTF-mode with UCP support, this uses the new data
256            to handle characters with more than one other case.
258        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
259            whose other case is greater than 256 are now correctly matched
260            caselessly. Previously, the high codepoint matched the low one, but not
261            vice versa.
263        (d) The processing of \h, \H, \v, and \ in character classes now makes use
264            of the new class addition function, using character lists defined as
265            macros alongside the case definitions of 20 above.
267        (e) Caseless back references now work with characters that have more than
268            one other case.
270        (f) General caseless matching of characters with more than one other case
271            is supported.
273    22. Unicode character properties were updated from Unicode 6.2.0
275    23. Improved CMake support under Windows. Patch by Daniel Richard G.
277    24. Add support for 32-bit character strings, and UTF-32
279    25. Major JIT compiler update (code refactoring and bugfixing).
280        Experimental Sparc 32 support is added.
282    26. Applied a modified version of Daniel Richard G's patch to create
283        pcre.h.generic and config.h.generic by "make" instead of in the
284        PrepareRelease script.
286    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
287        pcre_compile.c when checking for a zero character.
289    28. Introducing a native interface for JIT. Through this interface, the compiled
290        machine code can be directly executed. The purpose of this interface is to
291        provide fast pattern matching, so several sanity checks are not performed.
292        However, feature tests are still performed. The new interface provides
293        1.4x speedup compared to the old one.
295    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
296        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
297        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
299    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
300        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
301        the "old" RFC 2279). Instead, it ended up passing a negative length to
302        pcre_exec().
304    31. Add support for GCC's visibility feature to hide internal functions.
306    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
307        "unknown -C option" after outputting 0 or 1.
309    33. There is now support for generating a code coverage report for the test
310        suite in environments where gcc is the compiler and lcov is installed. This
311        is mainly for the benefit of the developers.
313    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
314        unaddressable using valgrind annotations, allowing valgrind to detect
315        invalid memory accesses. This is mainly for the benefit of the developers.
317    25. (*UTF) can now be used to start a pattern in any of the three libraries.
319    26. Give configure error if --enable-cpp but no C++ compiler found.
322    Version 8.31 06-July-2012
323    -------------------------
325    1.  Fixing a wrong JIT test case and some compiler warnings.
327    2.  Removed a bashism from the RunTest script.
329    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
330        to unsigned type, result still unsigned" that was given by an MS compiler
331        on encountering the code "-sizeof(xxx)".
333    4.  Partial matching support is added to the JIT compiler.
335    5.  Fixed several bugs concerned with partial matching of items that consist
336        of more than one character:
338        (a) /^(..)\1/ did not partially match "aba" because checking references was
339            done on an "all or nothing" basis. This also applied to repeated
340            references.
342        (b) \R did not give a hard partial match if \r was found at the end of the
343            subject.
345        (c) \X did not give a hard partial match after matching one or more
346            characters at the end of the subject.
348        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
349            a partial match for the string "\r".
351        (e) When newline was set to CRLF, the metacharacter "." did not recognize
352            a partial match for a CR character at the end of the subject string.
354    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
355        running pcretest, the text "(JIT)" added to the output whenever JIT is
356        actually used to run the match.
358    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
359        or /S+[+] with a digit between 1 and 7.
361    8.  OP_NOT now supports any UTF character not just single-byte ones.
363    9.  (*MARK) control verb is now supported by the JIT compiler.
365    10. The command "./RunTest list" lists the available tests without actually
366        running any of them. (Because I keep forgetting what they all are.)
370    12. Applied a (slightly modified) user-supplied patch that improves performance
371        when the heap is used for recursion (compiled with --disable-stack-for-
372        recursion). Instead of malloc and free for each heap frame each time a
373        logical recursion happens, frames are retained on a chain and re-used where
374        possible. This sometimes gives as much as 30% improvement.
376    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
377        call.
379    14. As documented, (*COMMIT) is now confined to within a positive assertion.
381    15. It is now possible to link pcretest with libedit as an alternative to
382        libreadline.
384    16. (*COMMIT) control verb is now supported by the JIT compiler.
386    17. The Unicode data tables have been updated to Unicode 6.1.0.
388    18. Added --file-list option to pcregrep.
390    19. Added binary file support to pcregrep, including the -a, --binary-files,
391        -I, and --text options.
393    20. The madvise function is renamed for posix_madvise for QNX compatibility
394        reasons. Fixed by Giuseppe D'Angelo.
396    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
398    22. Changed the option for creating symbolic links for 16-bit man pages from
399        -s to -sf so that re-installing does not cause issues.
401    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
403    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
404        matches in certain environments (the workspace was not being correctly
405        retained). Also added to pcre_dfa_exec() a simple plausibility check on
406        some of the workspace data at the beginning of a restart.
408    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
409        was not doing so when it should - probably a typo introduced by SVN 528
410        (change 8.10/14).
412    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
413        \w+ when the character tables indicated that \x{c4} was a word character.
414        There were several related cases, all because the tests for doing a table
415        lookup were testing for characters less than 127 instead of 255.
417    27. If a pattern contains capturing parentheses that are not used in a match,
418        their slots in the ovector are set to -1. For those that are higher than
419        any matched groups, this happens at the end of processing. In the case when
420        there were back references that the ovector was too small to contain
421        (causing temporary malloc'd memory to be used during matching), and the
422        highest capturing number was not used, memory off the end of the ovector
423        was incorrectly being set to -1. (It was using the size of the temporary
424        memory instead of the true size.)
426    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
427        ovector size, it uses memory at the end of the block that it has got.
429    29. Check for an overlong MARK name and give an error at compile time. The
430        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
432    30. JIT compiler update.
434    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
435        Rill for the patch.
437    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
439    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
441    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
442        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
443        was enabled.
445    35. Improve JIT code generation for greedy plus quantifier.
447    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
448        1 to "aa" instead of to an empty string. The bug affected repeated groups
449        that could potentially match an empty string.
451    37. Optimizing single character iterators in JIT.
453    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
454        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
455        codepoints that are too big for the mode are faulted, and in a UTF mode,
456        disallowed codepoints are also faulted.
458    39. If PCRE was compiled with UTF support, in three places in the DFA
459        matcher there was code that should only have been obeyed in UTF mode, but
460        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
461        processing when bytes with values greater than 127 were present. In 16-bit
462        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
463        both cases the values are those that cannot be the first data item in a UTF
464        character. The three items that might have provoked this were recursions,
465        possessively repeated groups, and atomic groups.
467    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
468        and pcregrep, because some OS require shared objects to be explicitly
469        passed to ld, causing the link step to fail if they are not.
471    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
472        mode, patterns that started with \h* or \R* might be incorrectly matched.
475    Version 8.30 04-February-2012
476    -----------------------------
478    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
479        name is defined in ctype.h.
481    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
482        only in quite long subpatterns.
484    3.  Removed the function pcre_info(), which has been obsolete and deprecated
485        since it was replaced by pcre_fullinfo() in February 2000.
487    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
488        match a (*MARK), and the match failed at the start of the subject, a
489        reference to memory before the start of the subject could occur. This bug
490        was introduced by fix 17 of release 8.21.
492    5.  A reference to an unset group with zero minimum repetition was giving
493        totally wrong answers (in non-JavaScript-compatibility mode). For example,
494        /(another)?(\1?)test/ matched against "hello world test". This bug was
495        introduced in release 8.13.
497    6.  Add support for 16-bit character strings (a large amount of work involving
498        many changes and refactorings).
500    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
501        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
502        from a file.
504    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
505        rounding is not applied in this particular case).
507    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
508        if they appear, or are escaped, in patterns.
510    10. Get rid of a number of -Wunused-but-set-variable warnings.
512    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
513        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
514        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
515        also returns the mark "x". This bug applied to capturing parentheses,
516        non-capturing parentheses, and atomic parentheses. It also applied to some
517        assertions.
519    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
520        information out of configure.ac instead of relying on pcre.h.generic, which
521        is not stored in the repository.
523    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
524        -lreadline.
526    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
528    15. Applied Graycode's patch to put the top-level frame on the stack rather
529        than the heap when not using the stack for recursion. This gives a
530        performance improvement in many cases when recursion is not deep.
532    16. Experimental code added to "pcretest -C" to output the stack frame size.
535    Version 8.21 12-Dec-2011
536    ------------------------
538    1.  Updating the JIT compiler.
540    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
541        are added as well.
543    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
544        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
545        calling _pcre_jit_exec. Some extra comments are added.
547    4.  (*MARK) settings inside atomic groups that do not contain any capturing
548        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
549        was introduced by change 18 for 8.20.
551    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
552        ECMA-262 standard.
554    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
555        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
556        This bug was probably introduced by change 9 of 8.13.
558    7.  While fixing 6 above, I noticed that a number of other items were being
559        incorrectly rejected as "not fixed length". This arose partly because newer
560        opcodes had not been added to the fixed-length checking code. I have (a)
561        corrected the bug and added tests for these items, and (b) arranged for an
562        error to occur if an unknown opcode is encountered while checking for fixed
563        length instead of just assuming "not fixed length". The items that were
564        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
565        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
566        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
568    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
569        being incorrectly compiled and would have given unpredicatble results.
571    9.  A possessively repeated subpattern with minimum repeat count greater than
572        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
573        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
574        the first (A) could occur when it should not.
576    10. Add a cast and remove a redundant test from the code.
578    11. JIT should use pcre_malloc/pcre_free for allocation.
580    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
581        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
582        is anything other than /usr, -L is still shown).
584    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
586    14. Perl does not support \N without a following name in a [] class; PCRE now
587        also gives an error.
589    15. If a forward reference was repeated with an upper limit of around 2000,
590        it caused the error "internal error: overran compiling workspace". The
591        maximum number of forward references (including repeats) was limited by the
592        internal workspace, and dependent on the LINK_SIZE. The code has been
593        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
594        the default depends on LINK_SIZE. There is a new upper limit (for safety)
595        of around 200,000 forward references. While doing this, I also speeded up
596        the filling in of repeated forward references.
598    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
599        incorrectly expecting the subject to contain another "a" after the start.
601    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
602        in the match, the SKIP should be ignored. This was not happening; instead
603        the SKIP was being treated as NOMATCH. For patterns such as
604        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
605        tested.
607    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
608        now much more compatible with Perl, in particular in cases where the result
609        is a non-match for a non-anchored pattern. For example, if
610        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
611        "m", where previously it did not return a name. A side effect of this
612        change is that for partial matches, the last encountered mark name is
613        returned, as for non matches. A number of tests that were previously not
614        Perl-compatible have been moved into the Perl-compatible test files. The
615        refactoring has had the pleasing side effect of removing one argument from
616        the match() function, thus reducing its stack requirements.
618    19. If the /S+ option was used in pcretest to study a pattern using JIT,
619        subsequent uses of /S (without +) incorrectly behaved like /S+.
621    21. Retrieve executable code size support for the JIT compiler and fixing
622        some warnings.
624    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
625        not work when the shorter character appeared right at the end of the
626        subject string.
628    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
629        systems.
631    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
632        output it when the /M option is used in pcretest.
634    25. The CheckMan script was not being included in the distribution. Also, added
635        an explicit "perl" to run Perl scripts from the PrepareRelease script
636        because this is reportedly needed in Windows.
638    26. If study data was being save in a file and studying had not found a set of
639        "starts with" bytes for the pattern, the data written to the file (though
640        never used) was taken from uninitialized memory and so caused valgrind to
641        complain.
643    27. Updated RunTest.bat as provided by Sheri Pierce.
645    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
647    29. Computation of memory usage for the table of capturing group names was
648        giving an unnecessarily large value.
651    Version 8.20 21-Oct-2011
652    ------------------------
654    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
655        a POSIX class. After further experiments with Perl, which convinced me that
656        Perl has bugs and confusions, a closing square bracket is no longer allowed
657        in a POSIX name. This bug also affected patterns with classes that started
658        with full stops.
660    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
661        captured substring, but while checking the failing first alternative,
662        substring 1 is temporarily captured. If the output vector supplied to
663        pcre_exec() was not big enough for this capture, the yield of the function
664        was still zero ("insufficient space for captured substrings"). This cannot
665        be totally fixed without adding another stack variable, which seems a lot
666        of expense for a edge case. However, I have improved the situation in cases
667        such as /(a)(b)x|abc/ matched against "abc", where the return code
668        indicates that fewer than the maximum number of slots in the ovector have
669        been set.
671    3.  Related to (2) above: when there are more back references in a pattern than
672        slots in the output vector, pcre_exec() uses temporary memory during
673        matching, and copies in the captures as far as possible afterwards. It was
674        using the entire output vector, but this conflicts with the specification
675        that only 2/3 is used for passing back captured substrings. Now it uses
676        only the first 2/3, for compatibility. This is, of course, another edge
677        case.
679    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
680        main code base, and can be used by building with --enable-jit. When this is
681        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
682        runtime --no-jit option is given.
684    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
685        ovector, the return from the function was zero, implying that there were
686        other matches that did not fit. The correct "exactly full" value is now
687        returned.
689    6.  If a subpattern that was called recursively or as a subroutine contained
690        (*PRUNE) or any other control that caused it to give a non-standard return,
691        invalid errors such as "Error -26 (nested recursion at the same subject
692        position)" or even infinite loops could occur.
694    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
695        computing the minimum length on reaching *ACCEPT, and so ended up with the
696        wrong value of 1 rather than 0. Further investigation indicates that
697        computing a minimum subject length in the presence of *ACCEPT is difficult
698        (think back references, subroutine calls), and so I have changed the code
699        so that no minimum is registered for a pattern that contains *ACCEPT.
701    8.  If (*THEN) was present in the first (true) branch of a conditional group,
702        it was not handled as intended. [But see 16 below.]
704    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
705        Sheri Pierce.
707    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
708        the first byte in a match must be "a".
710    11. Change 17 for 8.13 increased the recursion depth for patterns like
711        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
712        pattern contains any instances of (*THEN). If it does not, the old
713        optimizations are restored. It would be nice to do this on a per-group
714        basis, but at the moment that is not feasible.
716    12. In some environments, the output of pcretest -C is CRLF terminated. This
717        broke RunTest's code that checks for the link size. A single white space
718        character after the value is now allowed for.
720    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
721        For "fr", it uses the Windows-specific input and output files.
723    14. If (*THEN) appeared in a group that was called recursively or as a
724        subroutine, it did not work as intended. [But see next item.]
726    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
727        pattern fragments (but not containing any | characters). If A and B are
728        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
729        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
730        D. In other words, Perl considers parentheses that do not contain any |
731        characters to be part of a surrounding alternative, whereas PCRE was
732        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
733        differently. PCRE now behaves in the same way as Perl, except in the case
734        of subroutine/recursion calls such as (?1) which have in any case always
735        been different (but PCRE had them first :-).
737    16. Related to 15 above: Perl does not treat the | in a conditional group as
738        creating alternatives. Such a group is treated in the same way as an
739        ordinary group without any | characters when processing (*THEN). PCRE has
740        been changed to match Perl's behaviour.
742    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
743        RunGrepTest script failed.
745    18. Change 22 for version 13 caused atomic groups to use more stack. This is
746        inevitable for groups that contain captures, but it can lead to a lot of
747        stack use in large patterns. The old behaviour has been restored for atomic
748        groups that do not contain any capturing parentheses.
750    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
751        suppress the check for a minimum subject length at run time. (If it was
752        given to pcre_exec() or pcre_dfa_exec() it did work.)
754    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
755        fail to work when decoding hex characters in data strings in EBCDIC
756        environments.
758    21. It appears that in at least one Mac OS environment, the isxdigit() function
759        is implemented as a macro that evaluates to its argument more than once,
760        contravening the C 90 Standard (I haven't checked a later standard). There
761        was an instance in pcretest which caused it to go wrong when processing
762        \x{...} escapes in subject strings. The has been rewritten to avoid using
763        things like p++ in the argument of isxdigit().
766    Version 8.13 16-Aug-2011
767    ------------------------
769    1.  The Unicode data tables have been updated to Unicode 6.0.0.
771    2.  Two minor typos in pcre_internal.h have been fixed.
773    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
774        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
775        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
777    4.  There were a number of related bugs in the code for matching backrefences
778        caselessly in UTF-8 mode when codes for the characters concerned were
779        different numbers of bytes. For example, U+023A and U+2C65 are an upper
780        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
781        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
782        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
783        2-byte code at the end of the subject (it thought there wasn't enough data
784        left).
786    5.  Comprehensive information about what went wrong is now returned by
787        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
788        as the output vector has at least 2 elements. The offset of the start of
789        the failing character and a reason code are placed in the vector.
791    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
792        now returned is for the first byte of the failing character, instead of the
793        last byte inspected. This is an incompatible change, but I hope it is small
794        enough not to be a problem. It makes the returned offset consistent with
795        pcre_exec() and pcre_dfa_exec().
797    7.  pcretest now gives a text phrase as well as the error number when
798        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
799        failure, the offset and reason code are output.
801    8.  When \R was used with a maximizing quantifier it failed to skip backwards
802        over a \r\n pair if the subsequent match failed. Instead, it just skipped
803        back over a single character (\n). This seems wrong (because it treated the
804        two characters as a single entity when going forwards), conflicts with the
805        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
806        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
807        has been changed.
809    9.  Some internal refactoring has changed the processing so that the handling
810        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
811        time (the PCRE_DOTALL option was changed this way some time ago: version
812        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
813        which was always a bit of a fudge. It also means that there is one less
814        argument for the match() function, which reduces its stack requirements
815        slightly. This change also fixes an incompatibility with Perl: the pattern
816        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
818    10. More internal refactoring has drastically reduced the number of recursive
819        calls to match() for possessively repeated groups such as (abc)++ when
820        using pcre_exec().
822    11. While implementing 10, a number of bugs in the handling of groups were
823        discovered and fixed:
825        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
826        (a|)*(?1) gave a compile-time internal error.
827        ((a|)+)+  did not notice that the outer group could match an empty string.
828        (^a|^)+   was not marked as anchored.
829        (.*a|.*)+ was not marked as matching at start or after a newline.
831    12. Yet more internal refactoring has removed another argument from the match()
832        function. Special calls to this function are now indicated by setting a
833        value in a variable in the "match data" data block.
835    13. Be more explicit in pcre_study() instead of relying on "default" for
836        opcodes that mean there is no starting character; this means that when new
837        ones are added and accidentally left out of pcre_study(), testing should
838        pick them up.
840    14. The -s option of pcretest has been documented for ages as being an old
841        synonym of -m (show memory usage). I have changed it to mean "force study
842        for every regex", that is, assume /S for every regex. This is similar to -i
843        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
844        using it. It makes it easier to run collections of tests with and without
845        study enabled, and thereby test pcre_study() more easily. All the standard
846        tests are now run with and without -s (but some patterns can be marked as
847        "never study" - see 20 below).
849    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
850        restoration of the capturing data to the outer values was not happening
851        correctly.
853    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
854        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
855        pattern had matched an empty string, and so incorrectly returned a no
856        match.
858    17. There was optimizing code for the last branch of non-capturing parentheses,
859        and also for the obeyed branch of a conditional subexpression, which used
860        tail recursion to cut down on stack usage. Unfortunately, now that there is
861        the possibility of (*THEN) occurring in these branches, tail recursion is
862        no longer possible because the return has to be checked for (*THEN). These
863        two optimizations have therefore been removed. [But see 8.20/11 above.]
865    18. If a pattern containing \R was studied, it was assumed that \R always
866        matched two bytes, thus causing the minimum subject length to be
867        incorrectly computed because \R can also match just one byte.
869    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
870        was incorrectly computed.
872    20. If /S is present twice on a test pattern in pcretest input, it now
873        *disables* studying, thereby overriding the use of -s on the command line
874        (see 14 above). This is necessary for one or two tests to keep the output
875        identical in both cases.
877    21. When (*ACCEPT) was used in an assertion that matched an empty string and
878        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
880    22. When an atomic group that contained a capturing parenthesis was
881        successfully matched, but the branch in which it appeared failed, the
882        capturing was not being forgotten if a higher numbered group was later
883        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
884        group 1 to "a", when in fact it should be unset. This applied to multi-
885        branched capturing and non-capturing groups, repeated or not, and also to
886        positive assertions (capturing in negative assertions does not happen
887        in PCRE) and also to nested atomic groups.
889    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
890        subject after a captured substring, to make it easier to tell which of a
891        number of identical substrings has been captured.
893    24. The way atomic groups are processed by pcre_exec() has been changed so that
894        if they are repeated, backtracking one repetition now resets captured
895        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
896        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
897        "aaa". Previously, it would have been "a". As part of this code
898        refactoring, the way recursive calls are handled has also been changed.
900    25. If an assertion condition captured any substrings, they were not passed
901        back unless some other capturing happened later. For example, if
902        (?(?=(a))a) was matched against "a", no capturing was returned.
904    26. When studying a pattern that contained subroutine calls or assertions,
905        the code for finding the minimum length of a possible match was handling
906        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
907        group 1 called group 2 while simultaneously a separate group 2 called group
908        1). A stack overflow occurred in this case. I have fixed this by limiting
909        the recursion depth to 10.
911    27. Updated RunTest.bat in the distribution to the version supplied by Tom
912        Fortmann. This supports explicit test numbers on the command line, and has
913        argument validation and error reporting.
915    28. An instance of \X with an unlimited repeat could fail if at any point the
916        first character it looked at was a mark character.
918    29. Some minor code refactoring concerning Unicode properties and scripts
919        should reduce the stack requirement of match() slightly.
921    30. Added the '=' option to pcretest to check the setting of unused capturing
922        slots at the end of the pattern, which are documented as being -1, but are
923        not included in the return count.
925    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
926        compiled something random. Now it gives a compile-time error (as does
927        Perl).
929    32. A *MARK encountered during the processing of a positive assertion is now
930        recorded and passed back (compatible with Perl).
932    33. If --only-matching or --colour was set on a pcregrep call whose pattern
933        had alternative anchored branches, the search for a second match in a line
934        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
935        matched the line "0102" twice. The same bug affected patterns that started
936        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
937        twice.
939    34. Previously, PCRE did not allow quantification of assertions. However, Perl
940        does, and because of capturing effects, quantifying parenthesized
941        assertions may at times be useful. Quantifiers are now allowed for
942        parenthesized assertions.
944    35. A minor code tidy in pcre_compile() when checking options for \R usage.
946    36. \g was being checked for fancy things in a character class, when it should
947        just be a literal "g".
949    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
950        appearance of a nested POSIX class supersedes an apparent external class.
951        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
952        unescaped square brackets may also appear as part of class names. For
953        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
954        more like Perl. (But see 8.20/1 above.)
956    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
957        was because it thought it was \N{name}, which is not supported).
959    39. Add minix to OS list not supporting the -S option in pcretest.
961    40. PCRE tries to detect cases of infinite recursion at compile time, but it
962        cannot analyze patterns in sufficient detail to catch mutual recursions
963        such as ((?1))((?2)). There is now a runtime test that gives an error if a
964        subgroup is called recursively as a subpattern for a second time at the
965        same position in the subject string. In previous releases this might have
966        been caught by the recursion limit, or it might have run out of stack.
968    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
969        happen only once. PCRE was, however incorrectly giving a compile time error
970        "recursive call could loop indefinitely" because it cannot analyze the
971        pattern in sufficient detail. The compile time test no longer happens when
972        PCRE is compiling a conditional subpattern, but actual runaway loops are
973        now caught at runtime (see 40 above).
975    42. It seems that Perl allows any characters other than a closing parenthesis
976        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
977        has been changed to be the same.
979    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
980        as not to get warnings when autogen.sh is called. Also changed
981        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
983    44. To help people who use pcregrep to scan files containing exceedingly long
984        lines, the following changes have been made:
986        (a) The default value of the buffer size parameter has been increased from
987            8K to 20K. (The actual buffer used is three times this size.)
989        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
990            PCRE is built.
992        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
993            to be set at run time.
995        (d) Numerical values in pcregrep options can be followed by K or M, for
996            example --buffer-size=50K.
998        (e) If a line being scanned overflows pcregrep's buffer, an error is now
999            given and the return code is set to 2.
1001    45. Add a pointer to the latest mark to the callout data block.
1003    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1004        partial match of an empty string instead of no match. This was specific to
1005        the use of ".".
1007    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1008        complete match instead of a partial match. This bug was dependent on both
1009        the PCRE_UTF8 and PCRE_DOTALL options being set.
1011    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1012        starting byte set, because \b was not being ignored.
1015    Version 8.12 15-Jan-2011
1016    ------------------------
1018    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1019        checks for such things as part of the documentation building process.
1021    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1022        --match-limit and --recursion-limit options (added for 8.11). In
1023        particular, this made one of the standard tests fail. (The integer value
1024        went into the wrong half of a long int.)
1026    3.  If the --colour option was given to pcregrep with -v (invert match), it
1027        did strange things, either producing crazy output, or crashing. It should,
1028        of course, ignore a request for colour when reporting lines that do not
1029        match.
1031    4.  Another pcregrep bug caused similar problems if --colour was specified with
1032        -M (multiline) and the pattern match finished with a line ending.
1034    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1035        matched in multiline mode, the following line was shown as part of the
1036        match. This seems wrong, so I have changed it.
1038    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1039        the check for further matches in the same line (so they could be coloured)
1040        to overrun the end of the current line. If another match was found, it was
1041        incorrectly shown (and then shown again when found in the next line).
1043    7.  If pcregrep was compiled under Windows, there was a reference to the
1044        function pcregrep_exit() before it was defined. I am assuming this was
1045        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1046        reported by a user. I've moved the definition above the reference.
1049    Version 8.11 10-Dec-2010
1050    ------------------------
1052    1.  (*THEN) was not working properly if there were untried alternatives prior
1053        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1054        backtracked to try for "b" instead of moving to the next alternative branch
1055        at the same level (in this case, to look for "c"). The Perl documentation
1056        is clear that when (*THEN) is backtracked onto, it goes to the "next
1057        alternative in the innermost enclosing group".
1059    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1060        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1061        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1062        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1063        (*THEN).
1065    3.  If \s appeared in a character class, it removed the VT character from
1066        the class, even if it had been included by some previous item, for example
1067        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1068        of \s, but is part of the POSIX "space" class.)
1070    4.  A partial match never returns an empty string (because you can always
1071        match an empty string at the end of the subject); however the checking for
1072        an empty string was starting at the "start of match" point. This has been
1073        changed to the "earliest inspected character" point, because the returned
1074        data for a partial match starts at this character. This means that, for
1075        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1076        (previously it gave "no match").
1078    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1079        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1080        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1081        has an implication that the given string is incomplete (because a partial
1082        match is preferred over a full match). For this reason, these items now
1083        give a partial match in this situation. [Aside: previously, the one case
1084        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1085        match rather than a full match, which was wrong by the old rules, but is
1086        now correct.]
1088    6.  There was a bug in the handling of #-introduced comments, recognized when
1089        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1090        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1091        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1092        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1093        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1094        places in pcre_compile().
1096    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1097        comments when looking ahead for named forward references to subpatterns,
1098        the only newline sequence it recognized was NL. It now handles newlines
1099        according to the set newline convention.
1101    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1102        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1103        cater for a lack of strerror(). These oversights have been fixed.
1105    9.  Added --match-limit and --recursion-limit to pcregrep.
1107    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1109    11. When the -o option was used, pcregrep was setting a return code of 1, even
1110        when matches were found, and --line-buffered was not being honoured.
1112    12. Added an optional parentheses number to the -o and --only-matching options
1113        of pcregrep.
1115    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1116        can match an empty string. The code to do it in pcretest and pcredemo
1117        needed fixing:
1119        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1120            only one byte after an empty string match just before CRLF (this case
1121            just got forgotten; "any" and "anycrlf" were OK).
1123        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1124            mode when an empty string match preceded an ASCII character followed by
1125            a non-ASCII character. (The code for advancing by one character rather
1126            than one byte was nonsense.)
1128        (c) The pcredemo.c sample program did not have any code at all to handle
1129            the cases when CRLF is a valid newline sequence.
1131    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1132        as a starting offset was within the subject string. There is now a new
1133        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1134        negative or greater than the length of the string. In order to test this,
1135        pcretest is extended to allow the setting of negative starting offsets.
1137    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1138        starting offset points to the beginning of a UTF-8 character was
1139        unnecessarily clumsy. I tidied it up.
1141    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1142        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1144    17. Nobody had reported that the --include_dir option, which was added in
1145        release 7.7 should have been called --include-dir (hyphen, not underscore)
1146        for compatibility with GNU grep. I have changed it to --include-dir, but
1147        left --include_dir as an undocumented synonym, and the same for
1148        --exclude-dir, though that is not available in GNU grep, at least as of
1149        release 2.5.4.
1151    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1152        characters from a string of bytes) have been redefined so as not to use
1153        loops, in order to improve performance in some environments. At the same
1154        time, I abstracted some of the common code into auxiliary macros to save
1155        repetition (this should not affect the compiled code).
1157    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1158        compile-time error is now given if \c is not followed by an ASCII
1159        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1160        different, and any byte value is allowed.)
1162    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1163        START_OPTIMIZE option, which is now allowed at compile time - but just
1164        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1165        to pcregrep and other applications that have no direct access to PCRE
1166        options. The new /Y option in pcretest sets this option when calling
1167        pcre_compile().
1169    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1170        back references. Groups containing recursive back references were forced to
1171        be atomic by that change, but in the case of named groups, the amount of
1172        memory required was incorrectly computed, leading to "Failed: internal
1173        error: code overflow". This has been fixed.
1175    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1176        pcretest.c, to avoid build problems in some Borland environments.
1179    Version 8.10 25-Jun-2010
1180    ------------------------
1182    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1183        THEN.
1185    2.  (*ACCEPT) was not working when inside an atomic group.
1187    3.  Inside a character class, \B is treated as a literal by default, but
1188        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1189        causes the error). The code is unchanged, but I tidied the documentation.
1191    4.  Inside a character class, PCRE always treated \R and \X as literals,
1192        whereas Perl faults them if its -w option is set. I have changed PCRE so
1193        that it faults them when PCRE_EXTRA is set.
1195    5.  Added support for \N, which always matches any character other than
1196        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1198    6.  When compiling pcregrep with newer versions of gcc which may have
1199        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1200        declared with attribute warn_unused_result" were given. Just casting the
1201        result to (void) does not stop the warnings; a more elaborate fudge is
1202        needed. I've used a macro to implement this.
1204    7.  Minor change to pcretest.c to avoid a compiler warning.
1206    8.  Added four artifical Unicode properties to help with an option to make
1207        \s etc use properties (see next item). The new properties are: Xan
1208        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1210    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1211        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1212        this option. Modified pcretest to add /W to test this facility. Added
1213        REG_UCP to make it available via the POSIX interface.
1215    10. Added --line-buffered to pcregrep.
1217    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1218        studied, and the match started with a letter with a code point greater than
1219        127 whose first byte was different to the first byte of the other case of
1220        the letter, the other case of this starting letter was not recognized
1221        (#976).
1223    12. If a pattern that was studied started with a repeated Unicode property
1224        test, for example, \p{Nd}+, there was the theoretical possibility of
1225        setting up an incorrect bitmap of starting bytes, but fortunately it could
1226        not have actually happened in practice until change 8 above was made (it
1227        added property types that matched character-matching opcodes).
1229    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1230        possible starting bytes for non-anchored patterns.
1232    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1233        \R, and also a number of cases that involve Unicode properties, both
1234        explicit and implicit when PCRE_UCP is set.
1236    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1237        input, it could crash or give wrong results if characters with values
1238        greater than 0xc0 were present in the subject string. (Detail: it assumed
1239        UTF-8 input when processing these items.)
1241    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1242        size_t is 64-bit (#991).
1244    17. Added a check for running out of memory when PCRE is compiled with
1245        --disable-stack-for-recursion (#990).
1247    18. If the last data line in a file for pcretest does not have a newline on
1248        the end, a newline was missing in the output.
1250    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1251        less than 128) in its various bitmaps. However, there is a facility for
1252        generating tables according to the current locale when PCRE is compiled. It
1253        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1254        characters, are recognized by isspace() and therefore were getting set in
1255        these tables, and indeed these tables seem to approximate to ISO 8859. This
1256        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1257        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1258        which of course cannot start UTF-8 characters. I have changed the code so
1259        that only real ASCII characters (less than 128) and the correct starting
1260        bytes for UTF-8 encodings are set for characters greater than 127 when in
1261        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1262        altogether.)
1264    20. Added the /T option to pcretest so as to be able to run tests with non-
1265        standard character tables, thus making it possible to include the tests
1266        used for 19 above in the standard set of tests.
1268    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1269        reference to a subpattern the other side of a comment that contains an
1270        opening parenthesis caused either an internal compiling error, or a
1271        reference to the wrong subpattern.
1274    Version 8.02 19-Mar-2010
1275    ------------------------
1277    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1279    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1280        configured.
1282    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1283        original author of that file, following a query about its status.
1285    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1286        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1288    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1289        quantifier applied to a forward-referencing subroutine call, could compile
1290        incorrect code or give the error "internal error: previously-checked
1291        referenced subpattern not found".
1293    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1294        variables to point to external functions. For these systems, therefore,
1295        pcre_malloc etc. are now initialized to local functions that call the
1296        relevant global functions.
1298    7.  There were two entries missing in the vectors called coptable and poptable
1299        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1300        I've fixed the data, and added a kludgy way of testing at compile time that
1301        the lengths are correct (equal to the number of opcodes).
1303    8.  Following on from 7, I added a similar kludge to check the length of the
1304        eint vector in pcreposix.c.
1306    9.  Error texts for pcre_compile() are held as one long string to avoid too
1307        much relocation at load time. To find a text, the string is searched,
1308        counting zeros. There was no check for running off the end of the string,
1309        which could happen if a new error number was added without updating the
1310        string.
1312    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1314    11. \K was not working if it appeared in an atomic group or in a group that
1315        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1316        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1317        the assertion is positive, but not if it is negative.
1319    12. Change 11 fortuitously reduced the size of the stack frame used in the
1320        "match()" function of pcre_exec.c by one pointer. Forthcoming
1321        implementation of support for (*MARK) will need an extra pointer on the
1322        stack; I have reserved it now, so that the stack frame size does not
1323        decrease.
1325    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1326        item in branch that calls a recursion is a subroutine call - as in the
1327        second branch in the above example - was incorrectly given the compile-
1328        time error "recursive call could loop indefinitely" because pcre_compile()
1329        was not correctly checking the subroutine for matching a non-empty string.
1331    14. The checks for overrunning compiling workspace could trigger after an
1332        overrun had occurred. This is a "should never occur" error, but it can be
1333        triggered by pathological patterns such as hundreds of nested parentheses.
1334        The checks now trigger 100 bytes before the end of the workspace.
1336    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1339    Version 8.01 19-Jan-2010
1340    ------------------------
1342    1.  If a pattern contained a conditional subpattern with only one branch (in
1343        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1344        computed the wrong minimum data length (which is of course zero for such
1345        subpatterns). This could cause incorrect "no match" results.
1347    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1348        the pattern is reset in the first branch, pcre_compile() failed with
1349        "internal error: code overflow at offset...". This happened only when
1350        the reset was to the original external option setting. (An optimization
1351        abstracts leading options settings into an external setting, which was the
1352        cause of this.)
1354    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1355        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1356        assertion pattern did not match (meaning that the assertion was true), it
1357        was incorrectly treated as false if the SKIP had been reached during the
1358        matching. This also applied to assertions used as conditions.
1360    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1361        assertion subpattern, including such a pattern used as a condition,
1362        unpredictable results occurred, instead of the error return
1365    5.  The C++ GlobalReplace function was not working like Perl for the special
1366        situation when an empty string is matched. It now does the fancy magic
1367        stuff that is necessary.
1369    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1370        removed. (These were left over from very, very early versions of PCRE.)
1372    7.  Some cosmetic changes to the code to make life easier when compiling it
1373        as part of something else:
1375        (a) Change DEBUG to PCRE_DEBUG.
1377        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1378            called "current" as "current_branch", to prevent a collision with the
1379            Linux macro when compiled as a kernel module.
1381        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1382            prevent a collision with the Linux macro when compiled as a kernel
1383            module.
1385    8.  In pcre_compile() there are some checks for integer overflows that used to
1386        cast potentially large values to (double). This has been changed to that
1387        when building, a check for int64_t is made, and if it is found, it is used
1388        instead, thus avoiding the use of floating point arithmetic. (There is no
1389        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1390        double.
1392    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1393        2005 (difference between two addresses compared to an unsigned value).
1395    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1396        custom one, because of the following reported problem in Windows:
1398          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1399              under Win32.
1400          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1401              therefore missing the function definition.
1402          - The compiler thus generates a "C" signature for the test function.
1403          - The linker fails to find the "C" function.
1404          - PCRE fails to configure if asked to do so against libbz2.
1406    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1407        messages were output:
1409          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1410          rerunning libtoolize, to keep the correct libtool macros in-tree.
1411          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1413        I have done both of these things.
1415    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1416        most of the time, it *can* run out if it is given a pattern that contains a
1417        runaway infinite recursion. I updated the discussion in the pcrestack man
1418        page.
1420    13. Now that we have gone to the x.xx style of version numbers, the minor
1421        version may start with zero. Using 08 or 09 is a bad idea because users
1422        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1423        interpreted as invalid octal numbers. I've updated the previous comment in
1424        configure.ac, and also added a check that gives an error if 08 or 09 are
1425        used.
1427    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1428        causing partial matching to fail when the end of the subject matched \W
1429        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1431    15. There were some discrepancies between the declarations in pcre_internal.h
1432        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1433        their definitions. The declarations used "const uschar *" and the
1434        definitions used USPTR. Even though USPTR is normally defined as "const
1435        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1436        reported that: "This difference in casting confuses some C++ compilers, for
1437        example, SunCC recognizes above declarations as different functions and
1438        generates broken code for hbpcre." I have changed the declarations to use
1439        USPTR.
1441    16. GNU libtool is named differently on some systems. The autogen.sh script now
1442        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1443        (FreeBSD).
1445    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1446        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1447        comment: "Figure out how to create a longlong from a string: strtoll and
1448        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1449        instance, but it only takes 2 args instead of 3!"
1451    18. A subtle bug concerned with back references has been fixed by a change of
1452        specification, with a corresponding code fix. A pattern such as
1453        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1454        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1455        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1456        same bug. Such groups have to be quantified to be useful, or contained
1457        inside another quantified group. (If there's no repetition, the reference
1458        can never match.) The problem arises because, having left the group and
1459        moved on to the rest of the pattern, a later failure that backtracks into
1460        the group uses the captured value from the final iteration of the group
1461        rather than the correct earlier one. I have fixed this in PCRE by forcing
1462        any group that contains a reference to itself to be an atomic group; that
1463        is, there cannot be any backtracking into it once it has completed. This is
1464        similar to recursive and subroutine calls.
1467    Version 8.00 19-Oct-09
1468    ----------------------
1470    1.  The table for translating pcre_compile() error codes into POSIX error codes
1471        was out-of-date, and there was no check on the pcre_compile() error code
1472        being within the table. This could lead to an OK return being given in
1473        error.
1475    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1476        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1477        in a Windows environment.
1479    3.  The pcregrep --count option prints the count for each file even when it is
1480        zero, as does GNU grep. However, pcregrep was also printing all files when
1481        --files-with-matches was added. Now, when both options are given, it prints
1482        counts only for those files that have at least one match. (GNU grep just
1483        prints the file name in this circumstance, but including the count seems
1484        more useful - otherwise, why use --count?) Also ensured that the
1485        combination -clh just lists non-zero counts, with no names.
1487    4.  The long form of the pcregrep -F option was incorrectly implemented as
1488        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1489        but it seems right to fix it, and I didn't think it was worth preserving
1490        the old behaviour.
1492    5.  The command line items --regex=pattern and --regexp=pattern were not
1493        recognized by pcregrep, which required --regex pattern or --regexp pattern
1494        (with a space rather than an '='). The man page documented the '=' forms,
1495        which are compatible with GNU grep; these now work.
1497    6.  No libpcreposix.pc file was created for pkg-config; there was just
1498        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1500    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1501        when UCP support is not needed, by modifying the Python script that
1502        generates it from Unicode data files. This should not matter if the module
1503        is correctly used as a library, but I received one complaint about 50K of
1504        unwanted data. My guess is that the person linked everything into his
1505        program rather than using a library. Anyway, it does no harm.
1507    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1508        was a minimum greater than 1 for a wide character in a possessive
1509        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1510        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1511        character. Chaos in the form of incorrect output or a compiling loop could
1512        result.
1514    9.  The restrictions on what a pattern can contain when partial matching is
1515        requested for pcre_exec() have been removed. All patterns can now be
1516        partially matched by this function. In addition, if there are at least two
1517        slots in the offset vector, the offset of the earliest inspected character
1518        for the match and the offset of the end of the subject are set in them when
1519        PCRE_ERROR_PARTIAL is returned.
1521    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1522        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1523        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1524        and may be more useful for multi-segment matching.
1526    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1527        used to be given if ever the end of the subject was reached; now it is
1528        given only if matching could not proceed because another character was
1529        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1530        string "Z", which now yields "no match" instead of "partial match". In the
1531        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1532        final character ended with (*FAIL).
1534    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1535        if the pattern had a "must contain" character that was already found in the
1536        earlier partial match, unless partial matching was again requested. For
1537        example, with the pattern /dog.(body)?/, the "must contain" character is
1538        "g". If the first part-match was for the string "dog", restarting with
1539        "sbody" failed. This bug has been fixed.
1541    13. The string returned by pcre_dfa_exec() after a partial match has been
1542        changed so that it starts at the first inspected character rather than the
1543        first character of the match. This makes a difference only if the pattern
1544        starts with a lookbehind assertion or \b or \B (\K is not supported by
1545        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1546        matching functions compatible, and I think it's the right thing to do.
1548    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1549        so that the demonstration program is easily available in environments where
1550        PCRE has not been installed from source.
1552    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1553        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1554        library.
1556    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1557        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1558        is not the first non-POSIX option to be added. Clearly some people find
1559        these options useful.
1561    17. If a caller to the POSIX matching function regexec() passes a non-zero
1562        value for nmatch with a NULL value for pmatch, the value of
1563        nmatch is forced to zero.
1565    18. RunGrepTest did not have a test for the availability of the -u option of
1566        the diff command, as RunTest does. It now checks in the same way as
1567        RunTest, and also checks for the -b option.
1569    19. If an odd number of negated classes containing just a single character
1570        interposed, within parentheses, between a forward reference to a named
1571        subpattern and the definition of the subpattern, compilation crashed with
1572        an internal error, complaining that it could not find the referenced
1573        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1574        [The bug was that it was starting one character too far in when skipping
1575        over the character class, thus treating the ] as data rather than
1576        terminating the class. This meant it could skip too much.]
1578    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1579        /g option in pcretest when the pattern contains \K, which makes it possible
1580        to have an empty string match not at the start, even when the pattern is
1581        anchored. Updated pcretest and pcredemo to use this option.
1583    21. If the maximum number of capturing subpatterns in a recursion was greater
1584        than the maximum at the outer level, the higher number was returned, but
1585        with unset values at the outer level. The correct (outer level) value is
1586        now given.
1588    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1589        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1590        make it do so. The string so far is captured, making this feature
1591        compatible with Perl.
1593    23. The tests have been re-organized, adding tests 11 and 12, to make it
1594        possible to check the Perl 5.10 features against Perl 5.10.
1596    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1597        pattern matches a fixed length string. PCRE did not allow this; now it
1598        does. Neither allows recursion.
1600    25. I finally figured out how to implement a request to provide the minimum
1601        length of subject string that was needed in order to match a given pattern.
1602        (It was back references and recursion that I had previously got hung up
1603        on.) This code has now been added to pcre_study(); it finds a lower bound
1604        to the length of subject needed. It is not necessarily the greatest lower
1605        bound, but using it to avoid searching strings that are too short does give
1606        some useful speed-ups. The value is available to calling programs via
1607        pcre_fullinfo().
1609    26. While implementing 25, I discovered to my embarrassment that pcretest had
1610        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1611        study optimizations had never been tested with that matching function.
1612        Oops. What is worse, even when it was passed study data, there was a bug in
1613        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1614        were also very few tests of studied patterns with pcre_dfa_exec().
1616    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1617        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1618        on the other side of the coin, they are no longer allowed to have different
1619        names, because these cannot be distinguished in PCRE, and this has caused
1620        confusion. (This is a difference from Perl.)
1622    28. When duplicate subpattern names are present (necessarily with different
1623        numbers, as required by 27 above), and a test is made by name in a
1624        conditional pattern, either for a subpattern having been matched, or for
1625        recursion in such a pattern, all the associated numbered subpatterns are
1626        tested, and the overall condition is true if the condition is true for any
1627        one of them. This is the way Perl works, and is also more like the way
1628        testing by number works.
1631    Version 7.9 11-Apr-09
1632    ---------------------
1634    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1635        (pcretest), all targets were linked against these libraries. This included
1636        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1637        libraries. This caused unwanted dependencies to be created. This problem
1638        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1639        pcretest is linked with readline.
1641    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1642        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1643        moved outside it again, because FALSE and TRUE are already defined in AIX,
1644        but BOOL is not.
1646    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1647        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1649    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1650        hyphens) following file names and line numbers when outputting matching
1651        lines. This is not true; no spaces are inserted. I have also clarified the
1652        wording for the --colour (or --color) option.
1654    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1655        was not coloured; this is different to GNU grep, so I have changed it to be
1656        the same.
1658    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1659        each matching line was coloured. Now it goes on to look for further matches
1660        of any of the test patterns, which is the same behaviour as GNU grep.
1662    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1663        doesn't make sense to accept an empty string match in pcregrep, so I have
1664        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1665        seems to be how GNU grep behaves. [But see later change 40 for release
1666        8.33.]
1668    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1669        start or after a newline", because the conditional assertion was not being
1670        correctly handled. The rule now is that both the assertion and what follows
1671        in the first alternative must satisfy the test.
1673    9.  If auto-callout was enabled in a pattern with a conditional group whose
1674        condition was an assertion, PCRE could crash during matching, both with
1675        pcre_exec() and pcre_dfa_exec().
1677    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1678        used for matching.
1680    11. Unicode property support in character classes was not working for
1681        characters (bytes) greater than 127 when not in UTF-8 mode.
1683    12. Added the -M command line option to pcretest.
1685    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1687    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1689    16. Added comments and documentation about mis-use of no_arg in the C++
1690        wrapper.
1692    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1693        from Martin Jerabek that uses macro names for all relevant character and
1694        string constants.
1696    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1697        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1698        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1699        these, but not everybody uses configure.
1701    19. A conditional group that had only one branch was not being correctly
1702        recognized as an item that could match an empty string. This meant that an
1703        enclosing group might also not be so recognized, causing infinite looping
1704        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1705        with the subject "ab", where knowledge that the repeated group can match
1706        nothing is needed in order to break the loop.
1708    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1709        exec(), but without supplying a callout function, matching went wrong.
1711    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1712        leak if the size of the offset vector was greater than 30. When the vector
1713        is smaller, the saved offsets during recursion go onto a local stack
1714        vector, but for larger vectors malloc() is used. It was failing to free
1715        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1716        error, in fact).
1718    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1719        heapframe that is used only when UTF-8 support is enabled. This caused no
1720        problem, but was untidy.
1722    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1723        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1724        included within another project.
1726    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1727        slightly modified by me:
1729          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1730              not building pcregrep.
1732          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1733              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1735    25. Forward references, both numeric and by name, in patterns that made use of
1736        duplicate group numbers, could behave incorrectly or give incorrect errors,
1737        because when scanning forward to find the reference group, PCRE was not
1738        taking into account the duplicate group numbers. A pattern such as
1739        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1741    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1742        the feature of a custom pointer more persuasive (as requested by a user).
1744    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1745        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1746        pre-defined.
1748    28. Added support for (*UTF8) at the start of a pattern.
1750    29. Arrange for flags added by the "release type" setting in CMake to be shown
1751        in the configuration summary.
1754    Version 7.8 05-Sep-08
1755    ---------------------
1757    1.  Replaced UCP searching code with optimized version as implemented for Ad
1758        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1759        stage table and inline lookup instead of a function, giving speed ups of 2
1760        to 5 times on some simple patterns that I tested. Permission was given to
1761        distribute the MultiStage2.py script that generates the tables (it's not in
1762        the tarball, but is in the Subversion repository).
1764    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1765        scripts.
1767    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1768        a group with a zero qualifier. The result of the study could be incorrect,
1769        or the function might crash, depending on the pattern.
1771    4.  Caseless matching was not working for non-ASCII characters in back
1772        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1773        It now works when Unicode Property Support is available.
1775    5.  In pcretest, an escape such as \x{de} in the data was always generating
1776        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1777        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1778        truncation.
1780    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1782    7.  Added two (int) casts to pcregrep when printing the difference of two
1783        pointers, in case they are 64-bit values.
1785    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1786        test 2 if it fails.
1788    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1789        and a #define of that name to empty if it is not externally set. This is to
1790        allow users of MSVC to set it if necessary.
1792    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1793        the convenience functions in the pcre_get.c source file.
1795    11. An option change at the start of a pattern that had top-level alternatives
1796        could cause overwriting and/or a crash. This command provoked a crash in
1797        some environments:
1799          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1801        This potential security problem was recorded as CVE-2008-2371.
1803    12. For a pattern where the match had to start at the beginning or immediately
1804        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1805        pcre_dfa_exec() could read past the end of the passed subject if there was
1806        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1807        pcretest so that it places the subject at the end of its malloc-ed buffer.
1809    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1810        exec() might read past the end of the data buffer in UTF-8 mode.
1812    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1813        the data contained the byte 0x85 as part of a UTF-8 character within its
1814        first line. This applied both to normal and DFA matching.
1816    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1817        /^[^d]*?$/8 failed to match "abc".
1819    16. Added a missing copyright notice to pcrecpp_internal.h.
1821    17. Make it more clear in the documentation that values returned from
1822        pcre_exec() in ovector are byte offsets, not character counts.
1824    18. Tidied a few places to stop certain compilers from issuing warnings.
1826    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1827        supplied by Stefan Weber. I made a further small update for 7.8 because
1828        there is a change of source arrangements: the pcre_searchfuncs.c module is
1829        replaced by pcre_ucd.c.
1832    Version 7.7 07-May-08
1833    ---------------------
1835    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1836        a string to a long long, pretend we don't even have a long long." This is
1837        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1839    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1840        pre-7.6 versions, which defined a global no_arg variable instead of putting
1841        it in the RE class. (See also #8 below.)
1843    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1844        Lopes.
1846    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1848        (1) The include/exclude patterns were being applied to the whole pathnames
1849            of files, instead of just to the final components.
1851        (2) If there was more than one level of directory, the subdirectories were
1852            skipped unless they satisfied the include/exclude conditions. This is
1853            inconsistent with GNU grep (and could even be seen as contrary to the
1854            pcregrep specification - which I improved to make it absolutely clear).
1855            The action now is always to scan all levels of directory, and just
1856            apply the include/exclude patterns to regular files.
1858    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1859        --exclude_dir in the tests to avoid scanning .svn directories.
1861    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1862        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1863        doesn't support NULs in patterns.
1865    7.  Added some missing "const"s to declarations of static tables in
1866        pcre_compile.c and pcre_dfa_exec.c.
1868    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1869        caused by fix #2  above. (Subsequently also a second patch to fix the
1870        first patch. And a third patch - this was a messy problem.)
1872    9.  Applied Craig's patch to remove the use of push_back().
1874    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1875        matching function regexec().
1877    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1878        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1879        references. PCRE supports relative numbers with this syntax (I don't think
1880        Oniguruma does).
1882    12. Previously, a group with a zero repeat such as (...){0} was completely
1883        omitted from the compiled regex. However, this means that if the group
1884        was called as a subroutine from elsewhere in the pattern, things went wrong
1885        (an internal error was given). Such groups are now left in the compiled
1886        pattern, with a new opcode that causes them to be skipped at execution
1887        time.
1889    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1890        to the way PCRE behaves:
1892        (a) A lone ] character is dis-allowed (Perl treats it as data).
1894        (b) A back reference to an unmatched subpattern matches an empty string
1895            (Perl fails the current match path).
1897        (c) A data ] in a character class must be notated as \] because if the
1898            first data character in a class is ], it defines an empty class. (In
1899            Perl it is not possible to have an empty class.) The empty class []
1900            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1901            The negative empty class [^] matches any one character, independently
1902            of the DOTALL setting.
1904    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1905        non-existent subpattern following a character class starting with ']' and
1906        containing () gave an internal compiling error instead of "reference to
1907        non-existent subpattern". Fortunately, when the pattern did exist, the
1908        compiled code was correct. (When scanning forwards to check for the
1909        existencd of the subpattern, it was treating the data ']' as terminating
1910        the class, so got the count wrong. When actually compiling, the reference
1911        was subsequently set up correctly.)
1913    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1914        it was being rejected as not supported by pcre_dfa_exec(), even though
1915        other assertions are supported. I have made pcre_dfa_exec() support
1916        (*FAIL).
1918    16. The implementation of 13c above involved the invention of a new opcode,
1919        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1920        cannot be changed at match time, I realized I could make a small
1921        improvement to matching performance by compiling OP_ALLANY instead of
1922        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1923        on the OP_ANY path.
1925    17. Compiling pcretest on Windows with readline support failed without the
1926        following two fixes: (1) Make the unistd.h include conditional on
1927        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1929    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1930        ncurses library to be included for pcretest when ReadLine support is
1931        requested, but also to allow for it to be overridden. This patch came from
1932        Daniel Bergström.
1934    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1935        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1936        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1937        spotting this.
1940    Version 7.6 28-Jan-08
1941    ---------------------
1943    1.  A character class containing a very large number of characters with
1944        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1945        overflow.
1947    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1948        HAVE_LONG_LONG is not defined.
1950    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1951        bring them up to date and include new features. This patch includes:
1953        - Fixed PH's badly added libz and libbz2 support.
1954        - Fixed a problem with static linking.
1955        - Added pcredemo. [But later removed - see 7 below.]
1956        - Fixed dftables problem and added an option.
1957        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1958            HAVE_LONG_LONG.
1959        - Added readline support for pcretest.
1960        - Added an listing of the option settings after cmake has run.
1962    4.  A user submitted a patch to Makefile that makes it easy to create
1963        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1964        Makefile.am that cause it to include this special target, without
1965        affecting anything else. Note that the same mingw target plus all
1966        the other distribution libraries and programs are now supported
1967        when configuring with CMake (see 6 below) instead of with
1968        Configure/Make.
1970    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1971        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1972        exported in the Windows port". It has not yet been confirmed that the patch
1973        solves the problem, but it does no harm.
1975    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1976        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1977        with CMake, and also correct the comment about stack recursion.
1979    7.  Remove the automatic building of pcredemo from the ./configure system and
1980        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1981        of a program that users should build themselves after PCRE is installed, so
1982        building it automatically is not really right. What is more, it gave
1983        trouble in some build environments.
1985    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1988    Version 7.5 10-Jan-08
1989    ---------------------
1991    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1992        values in parens when parsing an RE using the C++ wrapper."
1994    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1995        Characters greater than 255 were excluded from the class instead of being
1996        included.
1998    3.  The same bug as (2) above applied to negated POSIX classes such as
1999        [:^space:].
2001    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2002        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2003        I have changed it.
2005    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2006        first named subpattern) and a construct such as (?&a) would reference the
2007        first named subpattern whose name started with "a" (in other words, the
2008        length check was missing). Both these problems are fixed. "Subpattern name
2009        expected" is now given for (?&) (a zero-length name), and this patch also
2010        makes it give the same error for \k'' (previously it complained that that
2011        was a reference to a non-existent subpattern).
2013    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2014        this is right because (?- can be followed by option settings as well as by
2015        digits. I have, however, made the messages clearer.
2017    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2018        than the number used in the conditional) now cause a compile-time error.
2019        This is actually not compatible with Perl, which accepts such patterns, but
2020        treats the conditional as always being FALSE (as PCRE used to), but it
2021        seems to me that giving a diagnostic is better.
2023    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2024        and messages.
2026    9.  Fix two occurrences of "backslash" in comments that should have been
2027        "backspace".
2029    10. Remove two redundant lines of code that can never be obeyed (their function
2030        was moved elsewhere).
2032    11. The program that makes PCRE's Unicode character property table had a bug
2033        which caused it to generate incorrect table entries for sequences of
2034        characters that have the same character type, but are in different scripts.
2035        It amalgamated them into a single range, with the script of the first of
2036        them. In other words, some characters were in the wrong script. There were
2037        thirteen such cases, affecting characters in the following ranges:
2039          U+002b0 - U+002c1
2040          U+0060c - U+0060d
2041          U+0061e - U+00612
2042          U+0064b - U+0065e
2043          U+0074d - U+0076d
2044          U+01800 - U+01805
2045          U+01d00 - U+01d77
2046          U+01d9b - U+01dbf
2047          U+0200b - U+0200f
2048          U+030fc - U+030fe
2049          U+03260 - U+0327f
2050          U+0fb46 - U+0fbb1
2051          U+10450 - U+1049d
2053    12. The -o option (show only the matching part of a line) for pcregrep was not
2054        compatible with GNU grep in that, if there was more than one match in a
2055        line, it showed only the first of them. It now behaves in the same way as
2056        GNU grep.
2058    13. If the -o and -v options were combined for pcregrep, it printed a blank
2059        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2060        does the same. The return code can be used to tell if there were any
2061        non-matching lines.
2063    14. Added --file-offsets and --line-offsets to pcregrep.
2065    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2066        infinitely looping recursion. The bug was that positive lookaheads were not
2067        being skipped when checking for a possible empty match (negative lookaheads
2068        and both kinds of lookbehind were skipped).
2070    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2071        inclusion of <windows.h> to before rather than after the definition of
2072        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2074    17. Specifying a possessive quantifier with a specific limit for a Unicode
2075        character property caused pcre_compile() to compile bad code, which led at
2076        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2077        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2078        caused the error; without that there was no problem.
2080    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2082    19. Added --enable-pcretest-libreadline.
2084    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2085        RE::GlobalReplace(). As a result, the number of replacements returned was
2086        double what it should be. I removed one of the increments, but Craig sent a
2087        later patch that removed the other one (the right fix) and added unit tests
2088        that check the return values (which was not done before).
2090    21. Several CMake things:
2092        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2093            the names libpcre and libpcreposix, not just pcre and pcreposix.
2095        (2) The above change means that pcretest and pcregrep are now correctly
2096            linked with the newly-built libraries, not previously installed ones.
2100    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2101        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2102        UTF-8 newline character). The key issue is that the pattern starts .*;
2103        this means that the match must be either at the beginning, or after a
2104        newline. The bug was in the code for advancing after a failed match and
2105        checking that the new position followed a newline. It was not taking
2106        account of UTF-8 characters correctly.
2108    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2109        character classes. PCRE was not treating the sequence [:...:] as a
2110        character class unless the ... were all letters. Perl, however, seems to
2111        allow any characters between [: and :], though of course it rejects as
2112        unknown any "names" that contain non-letters, because all the known class
2113        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2114        for example, whereas PCRE did not - it did not recognize a POSIX character
2115        class. This seemed a bit dangerous, so the code has been changed to be
2116        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2117        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2118        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2119        Perl does, and where it didn't before.
2121    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2122        Windows environments %n is disabled by default.
2125    Version 7.4 21-Sep-07
2126    ---------------------
2128    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2129        means that a class such as [\s] counted as "explicit reference to CR or
2130        LF". That isn't really right - the whole point of the change was to try to
2131        help when there was an actual mention of one of the two characters. So now
2132        the change happens only if \r or \n (or a literal CR or LF) character is
2133        encountered.
2135    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2136        of both had grown to the point where there were only 3 bits left.
2137        Fortunately, there was spare space in the data structure, and so I have
2138        moved the internal flags into a new 16-bit field to free up more option
2139        bits.
2141    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2142        but did not set the internal JCHANGED flag - either of these is enough to
2143        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2144        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2145        start sets both bits.
2147    4.  Added options (at build time, compile time, exec time) to change \R from
2148        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2150    5.  doc/pcresyntax.html was missing from the distribution.
2152    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2153        compatibility, even though it is no longer used.
2155    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2156        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2157        windows.h file is present (where different names are used). [This was
2158        reversed later after testing - see 16 below.]
2160    8.  Changed all #include <config.h> to #include "config.h". There were also
2161        some further <pcre.h> cases that I changed to "pcre.h".
2163    9.  When pcregrep was used with the --colour option, it missed the line ending
2164        sequence off the lines that it output.
2166    10. It was pointed out to me that arrays of string pointers cause lots of
2167        relocations when a shared library is dynamically loaded. A technique of
2168        using a single long string with a table of offsets can drastically reduce
2169        these. I have refactored PCRE in four places to do this. The result is
2170        dramatic:
2172          Originally:                          290
2173          After changing UCP table:            187
2174          After changing error message table:   43
2175          After changing table of "verbs"       36
2176          After changing table of Posix names   22
2178        Thanks to the folks working on Gregex for glib for this insight.
2180    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2181        unicode-properties was also set.
2183    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2185    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2186        checked only for CRLF.
2188    14. Added casts to pcretest.c to avoid compiler warnings.
2190    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2192    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2193        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2194        entirely. This removes changes made in 7 above.
2196    17. The CMake files have been updated, and there is now more information about
2197        building with CMake in the NON-UNIX-USE document.
2200    Version 7.3 28-Aug-07
2201    ---------------------
2203     1. In the rejigging of the build system that eventually resulted in 7.1, the
2204        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2205        brackets there is not right, since it causes compilers to look for an
2206        installed pcre.h, not the version that is in the source that is being
2207        compiled (which of course may be different). I have changed it back to:
2209          #include "pcre.h"
2211        I have a vague recollection that the change was concerned with compiling in
2212        different directories, but in the new build system, that is taken care of
2213        by the VPATH setting the Makefile.
2215     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2216        when the subject happened to end in the byte 0x85 (e.g. if the last
2217        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2218        characters but of course it shouldn't be taken as a newline when it is part
2219        of another character. The bug was that, for an unlimited repeat of . in
2220        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2221        characters when looking for a newline.
2223     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2225     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2226        in debug output.
2228     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2229        long printing in the pcrecpp unittest when running under MinGW.
2231     6. ESC_K was left out of the EBCDIC table.
2233     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2234        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2235        limit also applies to "virtual nesting" when a pattern is recursive, and in
2236        this case 1000 isn't so big. I have been able to remove this limit at the
2237        expense of backing off one optimization in certain circumstances. Normally,
2238        when pcre_exec() would call its internal match() function recursively and
2239        immediately return the result unconditionally, it uses a "tail recursion"
2240        feature to save stack. However, when a subpattern that can match an empty
2241        string has an unlimited repetition quantifier, it no longer makes this
2242        optimization. That gives it a stack frame in which to save the data for
2243        checking that an empty string has been matched. Previously this was taken
2244        from the 1000-entry workspace that had been reserved. So now there is no
2245        explicit limit, but more stack is used.
2247     8. Applied Daniel's patches to solve problems with the import/export magic
2248        syntax that is required for Windows, and which was going wrong for the
2249        pcreposix and pcrecpp parts of the library. These were overlooked when this
2250        problem was solved for the main library.
2252     9. There were some crude static tests to avoid integer overflow when computing
2253        the size of patterns that contain repeated groups with explicit upper
2254        limits. As the maximum quantifier is 65535, the maximum group length was
2255        set at 30,000 so that the product of these two numbers did not overflow a
2256        32-bit integer. However, it turns out that people want to use groups that
2257        are longer than 30,000 bytes (though not repeat them that many times).
2258        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2259        made it possible to implement the integer overflow checks in a much more
2260        dynamic way, which I have now done. The artificial limitation on group
2261        length has been removed - we now have only the limit on the total length of
2262        the compiled pattern, which depends on the LINK_SIZE setting.
2264    10. Fixed a bug in the documentation for get/copy named substring when
2265        duplicate names are permitted. If none of the named substrings are set, the
2266        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2267        empty string.
2269    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2270        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2271        because the ] is interpreted as the first data character and the
2272        terminating ] is not found. PCRE has been made compatible with Perl in this
2273        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2274        cause memory overwriting.
2276    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2277        string has been matched (to stop an infinite loop). It was not recognizing
2278        a conditional subpattern that could match an empty string if that
2279        subpattern was within another subpattern. For example, it looped when
2280        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2281        condition was not nested. This bug has been fixed.
2283    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2284        past the start of the subject in the presence of bytes with the top bit
2285        set, for example "\x8aBCD".
2287    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2288        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2290    14. Optimized (?!) to (*FAIL).
2292    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2293        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2294        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2295        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2296        does: it's just the validity check that is more restrictive.
2298    16. Inserted checks for integer overflows during escape sequence (backslash)
2299        processing, and also fixed erroneous offset values for syntax errors during
2300        backslash processing.
2302    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2303        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2305    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2306        caused an overrun.
2308    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2309        something other than just ASCII characters) inside a group that had an
2310        unlimited repeat caused a loop at compile time (while checking to see
2311        whether the group could match an empty string).
2313    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2314        [\P{Any}] did so. (Error in the code for printing property names.)
2316    21. An orphan \E inside a character class could cause a crash.
2318    22. A repeated capturing bracket such as (A)? could cause a wild memory
2319        reference during compilation.
2321    23. There are several functions in pcre_compile() that scan along a compiled
2322        expression for various reasons (e.g. to see if it's fixed length for look
2323        behind). There were bugs in these functions when a repeated \p or \P was
2324        present in the pattern. These operators have additional parameters compared
2325        with \d, etc, and these were not being taken into account when moving along
2326        the compiled data. Specifically:
2328        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2329            length.
2331        (b) An item such as \pL+ within a repeated group could cause crashes or
2332            loops.
2334        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2335            "reference to non-existent subpattern" error.
2337        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2339    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2340        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2342    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2344    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2345        character were causing crashes (broken optimization).
2347    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2348        \p or \P) caused a compile-time loop.
2350    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2351        break. For example, the unstudied pattern [\r\n]A does not match the string
2352        "\r\nA" because change 7.0/46 below moves the current point on by two
2353        characters after failing to match at the start. However, the pattern \nA
2354        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2355        the same is true. There doesn't seem any very clean way out of this, but
2356        what I have chosen to do makes the common cases work: PCRE now takes note
2357        of whether there can be an explicit match for \r or \n anywhere in the
2358        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2359        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2360        pattern has explicit CR or LF references.
2362    29. Added (*CR) etc for changing newline setting at start of pattern.
2365    Version 7.2 19-Jun-07
2366    ---------------------
2368     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2369        which is apparently normally available under Windows.
2371     2. Re-jig the pcregrep tests with different newline settings in an attempt
2372        to make them independent of the local environment's newline setting.
2374     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2376     4. Some of the "internals" tests were previously cut out when the link size
2377        was not 2, because the output contained actual offsets. The recent new
2378        "Z" feature of pcretest means that these can be cut out, making the tests
2379        usable with all link sizes.
2381     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2382        stack recursion. This gives a massive performance boost under BSD, but just
2383        a small improvement under Linux. However, it saves one field in the frame
2384        in all cases.
2386     6. Added more features from the forthcoming Perl 5.10:
2388        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2389            recursion call. It refers to the nth most recently opened parentheses.
2391        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2392            to be opened parentheses.
2394        (c) Conditions that refer to capturing parentheses can be specified
2395            relatively, for example, (?(-2)... or (?(+3)...
2397        (d) \K resets the start of the current match so that everything before
2398            is not part of it.
2400        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2402        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2403            reference syntax.
2405        (g) (?| introduces a group in which the numbering of parentheses in each
2406            alternative starts with the same number.
2408        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2410     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2413     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2414        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2415        for detecting groups that can match an empty string.
2417     9. A pattern with a very large number of alternatives (more than several
2418        hundred) was running out of internal workspace during the pre-compile
2419        phase, where pcre_compile() figures out how much memory will be needed. A
2420        bit of new cunning has reduced the workspace needed for groups with
2421        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2422        workspace instead of running out of the 4096 that are available.
2424    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2426    11. Applied patch from Google to remove an optimization that didn't quite work.
2427        The report of the bug said:
2429          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2430          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2431          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2433    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2434        it matched the wrong number of bytes.
2437    Version 7.1 24-Apr-07
2438    ---------------------
2440     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2441        that is more "standard", making use of automake and other Autotools. There
2442        is some re-arrangement of the files and adjustment of comments consequent
2443        on this.
2445     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2446        for recursive directory scanning broke on some systems because the files
2447        are not scanned in any specific order and on different systems the order
2448        was different. A call to "sort" has been inserted into RunGrepTest for the
2449        approprate test as a short-term fix. In the longer term there may be an
2450        alternative.
2452     3. I had an email from Eric Raymond about problems translating some of PCRE's
2453        man pages to HTML (despite the fact that I distribute HTML pages, some
2454        people do their own conversions for various reasons). The problems
2455        concerned the use of low-level troff macros .br and .in. I have therefore
2456        removed all such uses from the man pages (some were redundant, some could
2457        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2458        HTML has been updated to handle .nf/.fi and to complain if it encounters
2459        .br or .in.
2461     4. Updated comments in configure.ac that get placed in config.h.in and also
2462        arranged for config.h to be included in the distribution, with the name
2463        config.h.generic, for the benefit of those who have to compile without
2464        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2466     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2467        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2468        makevp.bat for the current PCRE, using the additional files
2469        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2471     6. A Windows user reported a minor discrepancy with test 2, which turned out
2472        to be caused by a trailing space on an input line that had got lost in his
2473        copy. The trailing space was an accident, so I've just removed it.
2475     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2476        that is needed.
2478     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2479        as "const" (a) because they are and (b) because it helps the PHP
2480        maintainers who have recently made a script to detect big data structures
2481        in the php code that should be moved to the .rodata section. I remembered
2482        to update Builducptable as well, so it won't revert if ucptable.h is ever
2483        re-created.
2485     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2486        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2487        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2488        support is not required. This saves 1.5-2K of code, which is important in
2489        some applications.
2491        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2492        so as not to refer to the tables, even though these functions will never be
2493        called when UTF-8 support is disabled. Otherwise there are problems with a
2494        shared library.
2496    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2498        (a) It was defining its arguments as char * instead of void *.
2500        (b) It was assuming that all moves were upwards in memory; this was true
2501            a long time ago when I wrote it, but is no longer the case.
2503        The emulated memove() is provided for those environments that have neither
2504        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2505        is clearly not the case, as these two bugs were recently reported.
2507    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2508        and Detrail to create the HTML documentation, the .txt form of the man
2509        pages, and it removes trailing spaces from listed files. It also creates
2510        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2511        case, it wraps all the #defines with #ifndefs. This script should be run
2512        before "make dist".
2514    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2515        with Unicode property support.
2517        (a) For a maximizing quantifier, if the two different cases of the
2518            character were of different lengths in their UTF-8 codings (there are
2519            some cases like this - I found 11), and the matching function had to
2520            back up over a mixture of the two cases, it incorrectly assumed they
2521            were both the same length.
2523        (b) When PCRE was configured to use the heap rather than the stack for
2524            recursion during matching, it was not correctly preserving the data for
2525            the other case of a UTF-8 character when checking ahead for a match
2526            while processing a minimizing repeat. If the check also involved
2527            matching a wide character, but failed, corruption could cause an
2528            erroneous result when trying to check for a repeat of the original
2529            character.
2531    13. Some tidying changes to the testing mechanism:
2533        (a) The RunTest script now detects the internal link size and whether there
2534            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2535            values substituted by "configure". (The RunGrepTest script already did
2536            this for UTF-8.) The configure.ac script no longer substitutes the
2537            relevant variables.
2539        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2540            with length and offset values. This means that the output is different
2541            for different internal link sizes. Test 2 is skipped for link sizes
2542            other than 2 because of this, bypassing the problem. Unfortunately,
2543            there was also a test in test 3 (the locale tests) that used /B and
2544            failed for link sizes other than 2. Rather than cut the whole test out,
2545            I have added a new /Z option to pcretest that replaces the length and
2546            offset values with spaces. This is now used to make test 3 independent
2547            of link size. (Test 2 will be tidied up later.)
2549    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2550        segmentation fault instead of returning the appropriate error message.
2552    15. In multiline mode when the newline sequence was set to "any", the pattern
2553        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2554        This doesn't seem right; it now treats the CRLF combination as the line
2555        ending, and so does not match in that case. It's only a pattern such as ^$
2556        that would hit this one: something like ^ABC$ would have failed after \r
2557        and then tried again after \r\n.
2559    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2560        in an attempt to make files that differ only in their line terminators
2561        compare equal. This works on Linux.
2563    17. Under certain error circumstances pcregrep might try to free random memory
2564        as it exited. This is now fixed, thanks to valgrind.
2566    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2567        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2568        was because its rules for how to advance for /g after matching an empty
2569        string at the end of a line did not allow for this case. They now check for
2570        it specially.
2572    20. pcretest is supposed to handle patterns and data of any length, by
2573        extending its buffers when necessary. It was getting this wrong when the
2574        buffer for a data line had to be extended.
2576    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2577        CRLF as a newline sequence.
2579    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2580        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2581        I have nevertheless tidied it up.
2583    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2585    24. Added a man page for pcre-config.
2588    Version 7.0 19-Dec-06
2589    ---------------------
2591     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2592        moving to gcc 4.1.1.
2594     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2595        sys/time.h, which is documented as needed for this function. It doesn't
2596        seem to matter on Linux, but it showed up on some releases of OS X.
2598     3. It seems that there are systems where bytes whose values are greater than
2599        127 match isprint() in the "C" locale. The "C" locale should be the
2600        default when a C program starts up. In most systems, only ASCII printing
2601        characters match isprint(). This difference caused the output from pcretest
2602        to vary, making some of the tests fail. I have changed pcretest so that:
2604        (a) When it is outputting text in the compiled version of a pattern, bytes
2605            other than 32-126 are always shown as hex escapes.
2607        (b) When it is outputting text that is a matched part of a subject string,
2608            it does the same, unless a different locale has been set for the match
2609            (using the /L modifier). In this case, it uses isprint() to decide.
2611     4. Fixed a major bug that caused incorrect computation of the amount of memory
2612        required for a compiled pattern when options that changed within the
2613        pattern affected the logic of the preliminary scan that determines the
2614        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2615        that the computed length was too small. The symptoms of this bug were
2616        either the PCRE error "internal error: code overflow" from pcre_compile(),
2617        or a glibc crash with a message such as "pcretest: free(): invalid next
2618        size (fast)". Examples of patterns that provoked this bug (shown in
2619        pcretest format) are:
2621          /(?-x: )/x
2622          /(?x)(?-x: \s*#\s*)/
2623          /((?i)[\x{c0}])/8
2624          /(?i:[\x{c0}])/8
2626        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2627        is now done differently.
2629     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2630        wrapper classes; (b) implement a new function in the C++ scanner that is
2631        more efficient than the old way of doing things because it avoids levels of
2632        recursion in the regex matching; (c) add a paragraph to the documentation
2633        for the FullMatch() function.
2635     6. The escape sequence \n was being treated as whatever was defined as
2636        "newline". Not only was this contrary to the documentation, which states
2637        that \n is character 10 (hex 0A), but it also went horribly wrong when
2638        "newline" was defined as CRLF. This has been fixed.
2640     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2641        was being set to -1 for the "end of line" case (supposedly a value that no
2642        character can have). Though this value is never used (the check for end of
2643        line is "zero bytes in current character"), it caused compiler complaints.
2644        I've changed it to 0xffffffff.
2646     8. In pcre_version.c, the version string was being built by a sequence of
2647        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2648        string (as it is for production releases) called a macro with an empty
2649        argument. The C standard says the result of this is undefined. The gcc
2650        compiler treats it as an empty string (which was what was wanted) but it is
2651        reported that Visual C gives an error. The source has been hacked around to
2652        avoid this problem.
2654     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2655        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2656        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2657        of them did).
2659    10. Originally, pcretest opened its input and output without "b"; then I was
2660        told that "b" was needed in some environments, so it was added for release
2661        5.0 to both the input and output. (It makes no difference on Unix-like
2662        systems.) Later I was told that it is wrong for the input on Windows. I've
2663        now abstracted the modes into two macros, to make it easier to fiddle with
2664        them, and removed "b" from the input mode under Windows.
2666    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2668    12. Added -help and --help to pcretest as an official way of being reminded
2669        of the options.
2671    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2672        and pcrecpp.cc because they annoy compilers at high warning levels.
2674    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2676    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2677        scripts are not C programs :-) and which was not noticed because it works
2678        on Linux.
2680    16. pcretest is supposed to handle any length of pattern and data line (as one
2681        line or as a continued sequence of lines) by extending its input buffer if
2682        necessary. This feature was broken for very long pattern lines, leading to
2683        a string of junk being passed to pcre_compile() if the pattern was longer
2684        than about 50K.
2686    17. I have done a major re-factoring of the way pcre_compile() computes the
2687        amount of memory needed for a compiled pattern. Previously, there was code
2688        that made a preliminary scan of the pattern in order to do this. That was
2689        OK when PCRE was new, but as the facilities have expanded, it has become
2690        harder and harder to keep it in step with the real compile phase, and there
2691        have been a number of bugs (see for example, 4 above). I have now found a
2692        cunning way of running the real compile function in a "fake" mode that
2693        enables it to compute how much memory it would need, while actually only
2694        ever using a few hundred bytes of working memory and without too many
2695        tests of the mode. This should make future maintenance and development
2696        easier. A side effect of this work is that the limit of 200 on the nesting
2697        depth of parentheses has been removed (though this was never a serious
2698        limitation, I suspect). However, there is a downside: pcre_compile() now
2699        runs more slowly than before (30% or more, depending on the pattern). I
2700        hope this isn't a big issue. There is no effect on runtime performance.
2702    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2703        newline (only possible for the last line of a file) and it was a
2704        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2706    19. Added additional timing features to pcretest. (1) The -tm option now times
2707        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2708        separate command line item, by a number that specifies the number of
2709        repeats to use when timing. The default is 50000; this gives better
2710        precision, but takes uncomfortably long for very large patterns.
2712    20. Extended pcre_study() to be more clever in cases where a branch of a
2713        subpattern has no definite first character. For example, (a*|b*)[cd] would
2714        previously give no result from pcre_study(). Now it recognizes that the
2715        first character must be a, b, c, or d.
2717    21. There was an incorrect error "recursive call could loop indefinitely" if
2718        a subpattern (or the entire pattern) that was being tested for matching an
2719        empty string contained only one non-empty item after a nested subpattern.
2720        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2721        incorrectly, because the \d was being skipped in the check.
2723    22. The pcretest program now has a new pattern option /B and a command line
2724        option -b, which is equivalent to adding /B to every pattern. This causes
2725        it to show the compiled bytecode, without the additional information that
2726        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2727        is the same as /B/I).
2729    23. A new optimization is now able automatically to treat some sequences such
2730        as a*b as a*+b. More specifically, if something simple (such as a character
2731        or a simple class like \d) has an unlimited quantifier, and is followed by
2732        something that cannot possibly match the quantified thing, the quantifier
2733        is automatically "possessified".
2735    24. A recursive reference to a subpattern whose number was greater than 39
2736        went wrong under certain circumstances in UTF-8 mode. This bug could also
2737        have affected the operation of pcre_study().
2739    25. Realized that a little bit of performance could be had by replacing
2740        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2742    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2744    27. Possessive quantifiers such as a++ were previously implemented by turning
2745        them into atomic groups such as ($>a+). Now they have their own opcodes,
2746        which improves performance. This includes the automatically created ones
2747        from 23 above.
2749    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2750        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2751        the first matched character to be a colon. This applied both to named and
2752        numbered groups.
2754    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2756    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2757        building PCRE on Windows easier, so I have included it in the distribution.
2759    31. There is now a check in pcretest against a ridiculously large number being
2760        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2761        loop, the loop is abandoned.
2763    32. Forward references to subpatterns in conditions such as (?(2)...) where
2764        subpattern 2 is defined later cause pcre_compile() to search forwards in
2765        the pattern for the relevant set of parentheses. This search went wrong
2766        when there were unescaped parentheses in a character class, parentheses
2767        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2769    33. "Subroutine" calls and backreferences were previously restricted to
2770        referencing subpatterns earlier in the regex. This restriction has now
2771        been removed.
2773    34. Added a number of extra features that are going to be in Perl 5.10. On the
2774        whole, these are just syntactic alternatives for features that PCRE had
2775        previously implemented using the Python syntax or my own invention. The
2776        other formats are all retained for compatibility.
2778        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2779            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2780            also .NET compatible.
2782        (b) A recursion or subroutine call to a named group can now be defined as
2783            (?&name) as well as (?P>name).
2785        (c) A backreference to a named group can now be defined as \k<name> or
2786            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2787            5.10, are also .NET compatible.
2789        (d) A conditional reference to a named group can now use the syntax
2790            (?(<name>) or (?('name') as well as (?(name).
2792        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2793            groups (named and numbered) that are never evaluated inline, but can be
2794            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2795            is always false. There may be only one alternative in such a group.
2797        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2798            as the simple (?(R). The condition is true only if the most recent
2799            recursion is that of the given number or name. It does not search out
2800            through the entire recursion stack.
2802        (g) The escape \gN or \g{N} has been added, where N is a positive or
2803            negative number, specifying an absolute or relative reference.
2805    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2806        some "unreachable code" warnings.
2808    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2809        things, this adds five new scripts.
2811    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2812        There were also incompatibilities regarding the handling of \Q..\E inside
2813        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2814        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2816    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2817        matches an empty string, and forcibly breaks the loop. There were bugs in
2818        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2819        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2820        separate and independent bugs (that affected different cases) have been
2821        fixed.
2823    39. Refactored the code to abolish the use of different opcodes for small
2824        capturing bracket numbers. This is a tidy that I avoided doing when I
2825        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2826        The new approach is not only tidier, it makes it possible to reduce the
2827        memory needed to fix the previous bug (38).
2829    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2830        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2831        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2832        mode.
2834    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2835        report.
2837    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2838        copy construction and assignment in the C++ wrapper.
2840    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2841        couple of bugs that could have given wrong results in the "--newline=crlf"
2842        case.
2844    44. Added a number of casts and did some reorganization of signed/unsigned int
2845        variables following suggestions from Dair Grant. Also renamed the variable
2846        "this" as "item" because it is a C++ keyword.
2848    45. Arranged for dftables to add
2850          #include "pcre_internal.h"
2852        to pcre_chartables.c because without it, gcc 4.x may remove the array
2853        definition from the final binary if PCRE is built into a static library and
2854        dead code stripping is activated.
2856    46. For an unanchored pattern, if a match attempt fails at the start of a
2857        newline sequence, and the newline setting is CRLF or ANY, and the next two
2858        characters are CRLF, advance by two characters instead of one.
2861    Version 6.7 04-Jul-06
2862    ---------------------
2864     1. In order to handle tests when input lines are enormously long, pcretest has
2865        been re-factored so that it automatically extends its buffers when
2866        necessary. The code is crude, but this _is_ just a test program. The
2867        default size has been increased from 32K to 50K.
2869     2. The code in pcre_study() was using the value of the re argument before
2870        testing it for NULL. (Of course, in any sensible call of the function, it
2871        won't be NULL.)
2873     3. The memmove() emulation function in pcre_internal.h, which is used on
2874        systems that lack both memmove() and bcopy() - that is, hardly ever -
2875        was missing a "static" storage class specifier.
2877     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2878        containing an extended class (one that cannot be represented by a bitmap
2879        because it contains high-valued characters or Unicode property items, e.g.
2880        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2881        pattern, but PCRE should not loop if you do not (it no longer does).
2882        [Detail: two cases were found: (a) a repeated subpattern containing an
2883        extended class; (b) a recursive reference to a subpattern that followed a
2884        previous extended class. It wasn't skipping over the extended class
2885        correctly when UTF-8 mode was not set.]
2887     5. A negated single-character class was not being recognized as fixed-length
2888        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2889        compile error "lookbehind assertion is not fixed length".
2891     6. The RunPerlTest auxiliary script was showing an unexpected difference
2892        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2893        write a Perl script that can interpret lines of an input file either as
2894        byte characters or as UTF-8, which is what "perltest" was being required to
2895        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2896        can't do is switch easily at run time between having the "use utf8;" pragma
2897        or not. In the end, I fudged it by using the RunPerlTest script to insert
2898        "use utf8;" explicitly for the UTF-8 tests.
2900     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2901        the end of the subject string, contrary to the documentation and to what
2902        Perl does. This was true of both matching functions. Now it matches only at
2903        the start of the subject and immediately after *internal* newlines.
2905     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2906        a pointer to an int instead of a pointer to an unsigned long int. This
2907        caused problems on 64-bit systems.
2909     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2910        instance of the 'standard' template library not being so standard".
2912    10. There was no check on the number of named subpatterns nor the maximum
2913        length of a subpattern name. The product of these values is used to compute
2914        the size of the memory block for a compiled pattern. By supplying a very
2915        long subpattern name and a large number of named subpatterns, the size
2916        computation could be caused to overflow. This is now prevented by limiting
2917        the length of names to 32 characters, and the number of named subpatterns
2918        to 10,000.
2920    11. Subpatterns that are repeated with specific counts have to be replicated in
2921        the compiled pattern. The size of memory for this was computed from the
2922        length of the subpattern and the repeat count. The latter is limited to
2923        65535, but there was no limit on the former, meaning that integer overflow
2924        could in principle occur. The compiled length of a repeated subpattern is
2925        now limited to 30,000 bytes in order to prevent this.
2927    12. Added the optional facility to have named substrings with the same name.
2929    13. Added the ability to use a named substring as a condition, using the
2930        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2931        are numbers (not recommended). Forward references are permitted.
2933    14. Added forward references in named backreferences (if you see what I mean).
2935    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2936        pattern could run off the end of the subject. For example, the pattern
2937        "(?s)(.{1,5})"8 did this with the subject "ab".
2939    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2940        PCRE_CASELESS was set when matching characters that were quantified with ?
2941        or *.
2943    17. A character class other than a single negated character that had a minimum
2944        but no maximum quantifier - for example [ab]{6,} - was not handled
2945        correctly by pce_dfa_exec(). It would match only one character.
2947    18. A valid (though odd) pattern that looked like a POSIX character
2948        class but used an invalid character after [ (for example [[,abc,]]) caused
2949        pcre_compile() to give the error "Failed: internal error: code overflow" or
2950        in some cases to crash with a glibc free() error. This could even happen if
2951        the pattern terminated after [[ but there just happened to be a sequence of
2952        letters, a binary zero, and a closing ] in the memory that followed.
2954    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2955        over the years. Originally (before any Unicode support), just the bottom 8
2956        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2957        output from "man perlunicode" includes this:
2959          The regular expression compiler produces polymorphic opcodes.  That
2960          is, the pattern adapts to the data and automatically switches to
2961          the Unicode character scheme when presented with Unicode data--or
2962          instead uses a traditional byte scheme when presented with byte
2963          data.
2965        Sadly, a wide octal escape does not cause a switch, and in a string with
2966        no other multibyte characters, these octal escapes are treated as before.
2967        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2968        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2969        Unicode string.
2971        I have not perpetrated such confusion in PCRE. Up till now, it took just
2972        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2973        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2974        translate to the appropriate multibyte character.
2976    29. Applied some refactoring to reduce the number of warnings from Microsoft
2977        and Borland compilers. This has included removing the fudge introduced
2978        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2979        a warning about an unused variable.
2981    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2982        characters since release 4.0, because Perl (from release 5.004) does not.
2983        [Or at least, is documented not to: some releases seem to be in conflict
2984        with the documentation.] However, when a pattern was studied with
2985        pcre_study() and all its branches started with \s, PCRE still included VT
2986        as a possible starting character. Of course, this did no harm; it just
2987        caused an unnecessary match attempt.
2989    22. Removed a now-redundant internal flag bit that recorded the fact that case
2990        dependency changed within the pattern. This was once needed for "required
2991        byte" processing, but is no longer used. This recovers a now-scarce options
2992        bit. Also moved the least significant internal flag bit to the most-
2993        significant bit of the word, which was not previously used (hangover from
2994        the days when it was an int rather than a uint) to free up another bit for
2995        the future.
2997    23. Added support for CRLF line endings as well as CR and LF. As well as the
2998        default being selectable at build time, it can now be changed at runtime
2999        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3000        specify that it is scanning data with non-default line endings.
3002    24. Changed the definition of CXXLINK to make it agree with the definition of
3003        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3005    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3006        recursions. This makes a big different to stack usage for some patterns.
3008    26. If a subpattern containing a named recursion or subroutine reference such
3009        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3010        the space required for the compiled pattern went wrong and gave too small a
3011        value. Depending on the environment, this could lead to "Failed: internal
3012        error: code overflow at offset 49" or "glibc detected double free or
3013        corruption" errors.
3015    27. Applied patches from Google (a) to support the new newline modes and (b) to
3016        advance over multibyte UTF-8 characters in GlobalReplace.
3018    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3019        difference for some implementation of PCRE in some Windows version.
3021    29. Added some extra testing facilities to pcretest:
3023        \q<number>   in a data line sets the "match limit" value
3024        \Q<number>   in a data line sets the "match recursion limt" value
3025        -S <number>  sets the stack size, where <number> is in megabytes
3027        The -S option isn't available for Windows.
3030    Version 6.6 06-Feb-06
3031    ---------------------
3033     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3034        in pcreposix.h. I have copied the definition from pcre.h.
3036     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3037        because pcre.h is no longer a built file.
3039     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3040        not normally included in the compiled code.
3043    Version 6.5 01-Feb-06
3044    ---------------------
3046     1. When using the partial match feature with pcre_dfa_exec(), it was not
3047        anchoring the second and subsequent partial matches at the new starting
3048        point. This could lead to incorrect results. For example, with the pattern
3049        /1234/, partially matching against "123" and then "a4" gave a match.
3051     2. Changes to pcregrep:
3053        (a) All non-match returns from pcre_exec() were being treated as failures
3054            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3055            error message is output. Some extra information is given for the
3056            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
3057            probably the only errors that are likely to be caused by users (by
3058            specifying a regex that has nested indefinite repeats, for instance).
3059            If there are more than 20 of these errors, pcregrep is abandoned.
3061        (b) A binary zero was treated as data while matching, but terminated the
3062            output line if it was written out. This has been fixed: binary zeroes
3063            are now no different to any other data bytes.
3065        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3066            used to set a locale for matching. The --locale=xxxx long option has
3067            been added (no short equivalent) to specify a locale explicitly on the
3068            pcregrep command, overriding the environment variables.
3070        (d) When -B was used with -n, some line numbers in the output were one less
3071            than they should have been.
3073        (e) Added the -o (--only-matching) option.
3075        (f) If -A or -C was used with -c (count only), some lines of context were
3076            accidentally printed for the final match.
3078        (g) Added the -H (--with-filename) option.
3080        (h) The combination of options -rh failed to suppress file names for files
3081            that were found from directory arguments.
3083        (i) Added the -D (--devices) and -d (--directories) options.
3085        (j) Added the -F (--fixed-strings) option.
3087        (k) Allow "-" to be used as a file name for -f as well as for a data file.
3089        (l) Added the --colo(u)r option.
3091        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3092            is not present by default.
3094     3. A nasty bug was discovered in the handling of recursive patterns, that is,
3095        items such as (?R) or (?1), when the recursion could match a number of
3096        alternatives. If it matched one of the alternatives, but subsequently,
3097        outside the recursion, there was a failure, the code tried to back up into
3098        the recursion. However, because of the way PCRE is implemented, this is not
3099        possible, and the result was an incorrect result from the match.
3101        In order to prevent this happening, the specification of recursion has
3102        been changed so that all such subpatterns are automatically treated as
3103        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3105     4. I had overlooked the fact that, in some locales, there are characters for
3106        which isalpha() is true but neither isupper() nor islower() are true. In
3107        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3108        and ordfeminine) are like this. This affected the treatment of \w and \W
3109        when they appeared in character classes, but not when they appeared outside
3110        a character class. The bit map for "word" characters is now created
3111        separately from the results of isalnum() instead of just taking it from the
3112        upper, lower, and digit maps. (Plus the underscore character, of course.)
3114     5. The above bug also affected the handling of POSIX character classes such as
3115        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3116        permanent tables. Instead, the bit maps for such a class were previously
3117        created as the appropriate unions of the upper, lower, and digit bitmaps.
3118        Now they are created by subtraction from the [[:word:]] class, which has
3119        its own bitmap.
3121     6. The [[:blank:]] character class matches horizontal, but not vertical space.
3122        It is created by subtracting the vertical space characters (\x09, \x0a,
3123        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3124        subtraction was done in the overall bitmap for a character class, meaning
3125        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3126        be recognized. This bug has been fixed.
3128     7. Patches from the folks at Google:
3130          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3131          real life, but is still worth protecting against".
3133          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3134          regular expressions".
3136          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3137          have it.
3139          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3140          "configure" and the latter not, in order to fix a problem somebody had
3141          with compiling the Arg class on HP-UX.
3143          (e) Improve the error-handling of the C++ wrapper a little bit.
3145          (f) New tests for checking recursion limiting.
3147     8. The pcre_memmove() function, which is used only if the environment does not
3148        have a standard memmove() function (and is therefore rarely compiled),
3149        contained two bugs: (a) use of int instead of size_t, and (b) it was not
3150        returning a result (though PCRE never actually uses the result).
3152     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3153        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3154        returned instead of calling malloc() with an overflowing number that would
3155        most likely cause subsequent chaos.
3157    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3159    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3160        with this option is matched, the nmatch and pmatch options of regexec() are
3161        ignored.
3163    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3164        provided in case anyone wants to the the POSIX interface with UTF-8
3165        strings.
3167    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3168        C++ linking (needed for some HP-UX environments).
3170    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3171        (unused parameter) and in the pcre_printint() function (omitted "default"
3172        switch label when the default is to do nothing).
3174    15. Added some code to make it possible, when PCRE is compiled as a C++
3175        library, to replace subject pointers for pcre_exec() with a smart pointer
3176        class, thus making it possible to process discontinuous strings.
3178    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3179        much the same function. They were added by different people who were trying
3180        to make PCRE easy to compile on non-Unix systems. It has been suggested
3181        that PCRE_EXPORT be abolished now that there is more automatic apparatus
3182        for compiling on Windows systems. I have therefore replaced it with
3183        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3184        defaults to "extern" for C or "extern C" for C++, which works fine on
3185        Unix-like systems. It is now possible to override the value of PCRE_DATA_
3186        SCOPE with something explicit in config.h. In addition:
3188        (a) pcreposix.h still had just "extern" instead of either of these macros;
3189            I have replaced it with PCRE_DATA_SCOPE.
3191        (b) Functions such as _pcre_xclass(), which are internal to the library,
3192            but external in the C sense, all had PCRE_EXPORT in their definitions.
3193            This is apparently wrong for the Windows case, so I have removed it.
3194            (It makes no difference on Unix-like systems.)
3196    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3197        of recursive calls to match(). This is different to MATCH_LIMIT because
3198        that limits the total number of calls to match(), not all of which increase
3199        the depth of recursion. Limiting the recursion depth limits the amount of
3200        stack (or heap if NO_RECURSE is set) that is used. The default can be set
3201        when PCRE is compiled, and changed at run time. A patch from Google adds
3202        this functionality to the C++ interface.
3204    18. Changes to the handling of Unicode character properties:
3206        (a) Updated the table to Unicode 4.1.0.
3208        (b) Recognize characters that are not in the table as "Cn" (undefined).
3210        (c) I revised the way the table is implemented to a much improved format
3211            which includes recognition of ranges. It now supports the ranges that
3212            are defined in UnicodeData.txt, and it also amalgamates other
3213            characters into ranges. This has reduced the number of entries in the
3214            table from around 16,000 to around 3,000, thus reducing its size
3215            considerably. I realized I did not need to use a tree structure after
3216            all - a binary chop search is just as efficient. Having reduced the
3217            number of entries, I extended their size from 6 bytes to 8 bytes to
3218            allow for more data.
3220        (d) Added support for Unicode script names via properties such as \p{Han}.
3222    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3223        matching that character.
3225    20. When matching a repeated Unicode property with a minimum greater than zero,
3226        (for example \pL{2,}), PCRE could look past the end of the subject if it
3227        reached it while seeking the minimum number of characters. This could
3228        happen only if some of the characters were more than one byte long, because
3229        there is a check for at least the minimum number of bytes.
3231    21. Refactored the implementation of \p and \P so as to be more general, to
3232        allow for more different types of property in future. This has changed the
3233        compiled form incompatibly. Anybody with saved compiled patterns that use
3234        \p or \P will have to recompile them.
3236    22. Added "Any" and "L&" to the supported property types.
3238    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3239        but give a compile time error if the value is greater than 0xff.
3241    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3242        accidentally not being installed or uninstalled.
3244    25. The pcre.h file was built from pcre.h.in, but the only changes that were
3245        made were to insert the current release number. This seemed silly, because
3246        it made things harder for people building PCRE on systems that don't run
3247        "configure". I have turned pcre.h into a distributed file, no longer built
3248        by "configure", with the version identification directly included. There is
3249        no longer a pcre.h.in file.
3251        However, this change necessitated a change to the pcre-config script as
3252        well. It is built from pcre-config.in, and one of the substitutions was the
3253        release number. I have updated configure.ac so that ./configure now finds
3254        the release number by grepping pcre.h.
3256    26. Added the ability to run the tests under valgrind.
3259    Version 6.4 05-Sep-05
3260    ---------------------
3262     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3263        "--" to be printed when multiple files were scanned, even when none of the
3264        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3265        consider it to be a bug, and have restored the previous behaviour.
3267     2. A couple of code tidies to get rid of compiler warnings.
3269     3. The pcretest program used to cheat by referring to symbols in the library
3270        whose names begin with _pcre_. These are internal symbols that are not
3271        really supposed to be visible externally, and in some environments it is
3272        possible to suppress them. The cheating is now confined to including
3273        certain files from the library's source, which is a bit cleaner.
3275     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3276        file's purpose clearer.
3278     5. Reorganized pcre_ucp_findchar().
3281    Version 6.3 15-Aug-05
3282    ---------------------
3284     1. The file libpcre.pc.in did not have general read permission in the tarball.
3286     2. There were some problems when building without C++ support:
3288        (a) If C++ support was not built, "make install" and "make test" still
3289            tried to test it.
3291        (b) There were problems when the value of CXX was explicitly set. Some
3292            changes have been made to try to fix these, and ...
3294        (c) --disable-cpp can now be used to explicitly disable C++ support.
3296        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3297            backslash in a target when C++ was disabled. This confuses some
3298            versions of "make", apparently. Using an intermediate variable solves
3299            this. (Same for CPP_LOBJ.)
3301     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3302        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3303        necessary on certain architectures.
3305     4. Added a setting of -export-symbols-regex to the link command to remove
3306        those symbols that are exported in the C sense, but actually are local
3307        within the library, and not documented. Their names all begin with
3308        "_pcre_". This is not a perfect job, because (a) we have to except some
3309        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3310        available (and never for static libraries). I have made a note to try to
3311        find a way round (a) in the future.
3314    Version 6.2 01-Aug-05
3315    ---------------------
3317     1. There was no test for integer overflow of quantifier values. A construction
3318        such as {1111111111111111} would give undefined results. What is worse, if
3319        a minimum quantifier for a parenthesized subpattern overflowed and became
3320        negative, the calculation of the memory size went wrong. This could have
3321        led to memory overwriting.
3323     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3325     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3326        operating environments where this matters.
3328     4. Applied Giuseppe Maxia's patch to add additional features for controlling
3329        PCRE options from within the C++ wrapper.
3331     5. Named capturing subpatterns were not being correctly counted when a pattern
3332        was compiled. This caused two problems: (a) If there were more than 100
3333        such subpatterns, the calculation of the memory needed for the whole
3334        compiled pattern went wrong, leading to an overflow error. (b) Numerical
3335        back references of the form \12, where the number was greater than 9, were
3336        not recognized as back references, even though there were sufficient
3337        previous subpatterns.
3339     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3340        versions of gcc, e.g. 2.95.4.
3343    Version 6.1 21-Jun-05
3344    ---------------------
3346     1. There was one reference to the variable "posix" in pcretest.c that was not
3347        surrounded by "#if !defined NOPOSIX".
3349     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3350        the cross-check on the old pcre_info() function, for the benefit of the
3351        cut-down version of PCRE that is currently imported into Exim.
3353     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3354        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3355        patterns that start (?i) but I don't think that matters. The use of (?i) is
3356        just an example; this all applies to the other options as well.
3358     4. Since libtool seems to echo the compile commands it is issuing, the output
3359        from "make" can be reduced a bit by putting "@" in front of each libtool
3360        compile command.
3362     5. Patch from the folks at Google for configure.in to be a bit more thorough
3363        in checking for a suitable C++ installation before trying to compile the
3364        C++ stuff. This should fix a reported problem when a compiler was present,
3365        but no suitable headers.
3367     6. The man pages all had just "PCRE" as their title. I have changed them to
3368        be the relevant file name. I have also arranged that these names are
3369        retained in the file doc/pcre.txt, which is a concatenation in text format
3370        of all the man pages except the little individual ones for each function.
3372     7. The NON-UNIX-USE file had not been updated for the different set of source
3373        files that come with release 6. I also added a few comments about the C++
3374        wrapper.
3377    Version 6.0 07-Jun-05
3378    ---------------------
3380     1. Some minor internal re-organization to help with my DFA experiments.
3382     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3383        didn't matter for the library itself when fully configured, but did matter
3384        when compiling without UCP support, or within Exim, where the ucp files are
3385        not imported.
3387     3. Refactoring of the library code to split up the various functions into
3388        different source modules. The addition of the new DFA matching code (see
3389        below) to a single monolithic source would have made it really too
3390        unwieldy, quite apart from causing all the code to be include in a
3391        statically linked application, when only some functions are used. This is
3392        relevant even without the DFA addition now that patterns can be compiled in
3393        one application and matched in another.
3395        The downside of splitting up is that there have to be some external
3396        functions and data tables that are used internally in different modules of
3397        the library but which are not part of the API. These have all had their
3398        names changed to start with "_pcre_" so that they are unlikely to clash
3399        with other external names.
3401     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3402        a different (DFA) algorithm. Although it is slower than the original
3403        function, it does have some advantages for certain types of matching
3404        problem.
3406     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3407        including restarting after a partial match.
3409     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3410        defined when compiling for Windows was sent to me. I have put it into the
3411        code, though I have no means of testing or verifying it.
3413     7. Added the pcre_refcount() auxiliary function.
3415     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3416        match before or at the first newline in the subject string. In pcretest,
3417        the /f option on a pattern can be used to set this.
3419     9. A repeated \w when used in UTF-8 mode with characters greater than 256
3420        would behave wrongly. This has been present in PCRE since release 4.0.
3422    10. A number of changes to the pcregrep command:
3424        (a) Refactored how -x works; insert ^(...)$ instead of setting
3425            PCRE_ANCHORED and checking the length, in preparation for adding
3426            something similar for -w.
3428        (b) Added the -w (match as a word) option.
3430        (c) Refactored the way lines are read and buffered so as to have more
3431            than one at a time available.
3433        (d) Implemented a pcregrep test script.
3435        (e) Added the -M (multiline match) option. This allows patterns to match
3436            over several lines of the subject. The buffering ensures that at least
3437            8K, or the rest of the document (whichever is the shorter) is available
3438            for matching (and similarly the previous 8K for lookbehind assertions).
3440        (f) Changed the --help output so that it now says
3442              -w, --word-regex(p)
3444            instead of two lines, one with "regex" and the other with "regexp"
3445            because that confused at least one person since the short forms are the
3446            same. (This required a bit of code, as the output is generated
3447            automatically from a table. It wasn't just a text change.)
3449        (g) -- can be used to terminate pcregrep options if the next thing isn't an
3450            option but starts with a hyphen. Could be a pattern or a path name
3451            starting with a hyphen, for instance.
3453        (h) "-" can be given as a file name to represent stdin.
3455        (i) When file names are being printed, "(standard input)" is used for
3456            the standard input, for compatibility with GNU grep. Previously
3457            "<stdin>" was used.
3459        (j) The option --label=xxx can be used to supply a name to be used for
3460            stdin when file names are being printed. There is no short form.
3462        (k) Re-factored the options decoding logic because we are going to add
3463            two more options that take data. Such options can now be given in four
3464            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3466        (l) Added the -A, -B, and -C options for requesting that lines of context
3467            around matches be printed.
3469        (m) Added the -L option to print the names of files that do not contain
3470            any matching lines, that is, the complement of -l.
3472        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3473            continue to scan other files.
3475        (o) The -s option was incorrectly implemented. For compatibility with other
3476            greps, it now suppresses the error message for a non-existent or non-
3477            accessible file (but not the return code). There is a new option called
3478            -q that suppresses the output of matching lines, which was what -s was
3479            previously doing.
3481        (p) Added --include and --exclude options to specify files for inclusion
3482            and exclusion when recursing.
3484    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3485        Hopefully, it now does.
3487    12. Missing cast in pcre_study().
3489    13. Added an "uninstall" target to the makefile.
3491    14. Replaced "extern" in the function prototypes in Makefile.in with
3492        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3493        world, but is set differently for Windows.
3495    15. Added a second compiling function called pcre_compile2(). The only
3496        difference is that it has an extra argument, which is a pointer to an
3497        integer error code. When there is a compile-time failure, this is set
3498        non-zero, in addition to the error test pointer being set to point to an
3499        error message. The new argument may be NULL if no error number is required
3500        (but then you may as well call pcre_compile(), which is now just a
3501        wrapper). This facility is provided because some applications need a
3502        numeric error indication, but it has also enabled me to tidy up the way
3503        compile-time errors are handled in the POSIX wrapper.
3505    16. Added VPATH=.libs to the makefile; this should help when building with one
3506        prefix path and installing with another. (Or so I'm told by someone who
3507        knows more about this stuff than I do.)
3509    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3510        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3511        match everything, including newlines. This is not POSIX-compatible, but
3512        somebody wanted the feature. From pcretest it can be activated by using
3513        both the P and the s flags.
3515    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3517    19. libpcre.pc was being incorrectly installed as executable.
3519    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3520        it now also looks for '\r' so that it will work unmodified on Windows.
3522    21. Added Google's contributed C++ wrapper to the distribution.
3524    22. Added some untidy missing memory free() calls in pcretest, to keep
3525        Electric Fence happy when testing.
3529    Version 5.0 13-Sep-04
3530    ---------------------
3532     1. Internal change: literal characters are no longer packed up into items
3533        containing multiple characters in a single byte-string. Each character
3534        is now matched using a separate opcode. However, there may be more than one
3535        byte in the character in UTF-8 mode.
3537     2. The pcre_callout_block structure has two new fields: pattern_position and
3538        next_item_length. These contain the offset in the pattern to the next match
3539        item, and its length, respectively.
3541     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3542        insertion of callouts before each pattern item. Added the /C option to
3543        pcretest to make use of this.
3545     4. On the advice of a Windows user, the lines
3547          #if defined(_WIN32) || defined(WIN32)
3548          _setmode( _fileno( stdout ), 0x8000 );
3549          #endif  /* defined(_WIN32) || defined(WIN32) */
3551        have been added to the source of pcretest. This apparently does useful
3552        magic in relation to line terminators.
3554     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3555        for the benefit of those environments where the "b" makes a difference.
3557     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3558        to know about it. I have put a hack into configure.in that adds in code
3559        to set GCC=yes if CC=icc. This seems to end up at a point in the
3560        generated configure script that is early enough to affect the setting of
3561        compiler options, which is what is needed, but I have no means of testing
3562        whether it really works. (The user who reported this had patched the
3563        generated configure script, which of course I cannot do.)
3565        LATER: After change 22 below (new libtool files), the configure script
3566        seems to know about icc (and also ecc). Therefore, I have commented out
3567        this hack in configure.in.
3569     7. Added support for pkg-config (2 patches were sent in).
3571     8. Negated POSIX character classes that used a combination of internal tables
3572        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3573        [[:^ascii]]. Typically, they would match almost any characters. The other
3574        POSIX classes were not broken in this way.
3576     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3577        to find the match, as PCRE was deluded into thinking that the match had to
3578        start at the start point or following a newline. The same bug applied to
3579        patterns with negative forward assertions or any backward assertions
3580        preceding ".*" at the start, unless the pattern required a fixed first
3581        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3583    10. In UTF-8 mode,