ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 31 by nigel, Sat Feb 24 21:38:57 2007 UTC revision 1392 by ph10, Wed Nov 6 18:00:09 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.34 xx-xxxx-201x
5    --------------------------
7    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
8        executable memory. Patch inspired by Carsten Klein.
10    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
11        this macro is never tested and has no effect, because the work to support
12        coverage involves only compiling and linking options and special targets in
13        the Makefile. The comment in config.h implied that defining the macro would
14        enable coverage support, which is totally false. There was also support for
15        setting this macro in the CMake files (my fault, I just copied it from
16        configure). SUPPORT_GCOV has now been removed.
18    3.  Make a small performance improvement in strlen16() and strlen32() in
19        pcretest.
21    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
22        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
23        these cases). There was also one in pcretest.c.
25    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
27    6.  In UTF mode, the code for checking whether a group could match an empty
28        string (which is used for indefinitely repeated groups to allow for
29        breaking an infinite loop) was broken when the group contained a repeated
30        negated single-character class with a character that occupied more than one
31        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
32        UTF-8 mode). The effect was undefined: the group might or might not be
33        deemed as matching an empty string, or the program might have crashed.
35    7.  The code for checking whether a group could match an empty string was not
36        recognizing that \h, \H, \v, \V, and \R must match a character.
38    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
39        an empty string. If it can, pcretest shows this in its information output.
41    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
42        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
43        matched by pcre_exec() without using JIT:
45        (a) If the rest of the pattern did not match after a maximal run of
46            grapheme clusters, the code for backing up to try with fewer of them
47            did not always back up over a full grapheme when characters that do not
48            have the modifier quality were involved, e.g. Hangul syllables.
50        (b) If the match point in a subject started with modifier character, and
51            there was no match, the code could incorrectly back up beyond the match
52            point, and potentially beyond the first character in the subject,
53            leading to a segfault or an incorrect match result.
55    10. A conditional group with an assertion condition could lead to PCRE
56        recording an incorrect first data item for a match if no other first data
57        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
58        first data item, and therefore matched "ca" after "c" instead of at the
59        start.
61    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
62        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
64    12. The source of pcregrep now includes z/OS-specific code so that it can be
65        compiled for z/OS as part of the special z/OS distribution.
67    13. Added the -T and -TM options to pcretest.
69    14. The code in pcre_compile.c for creating the table of named capturing groups
70        has been refactored. Instead of creating the table dynamically during the
71        actual compiling pass, the information is remembered during the pre-compile
72        pass (on the stack unless there are more than 20 named groups, in which
73        case malloc() is used) and the whole table is created before the actual
74        compile happens. This has simplified the code (it is now nearly 150 lines
75        shorter) and prepared the way for better handling of references to groups
76        with duplicate names.
78    15. A back reference to a named subpattern when there is more than one of the
79        same name now checks them in the order in which they appear in the pattern.
80        The first one that is set is used for the reference. Previously only the
81        first one was inspected. This change makes PCRE more compatible with Perl.
83    16. Unicode character properties were updated from Unicode 6.3.0.
85    17. The compile-time code for auto-possessification has been refactored, based
86        on a patch by Zoltan Herczeg. It now happens after instead of during
87        compilation. The code is cleaner, and more cases are handled. The option
88        PCRE_NO_AUTO_POSSESSIFY is added for testing purposes, and the -O and /O
89        options in pcretest are provided to set it.
91    18. The character VT has been added to the set of characters that match \s and
92        are generally treated as white space, following this same change in Perl
93        5.18. There is now no difference between "Perl space" and "POSIX space".
95    19. The code for checking named groups as conditions, either for being set or
96        for being recursed, has been refactored (this is related to 14 and 15
97        above). Processing unduplicated named groups should now be as fast at
98        numerical groups, and processing duplicated groups should be faster than
99        before.
101    20. Two patches to the CMake build system, by Alexander Barkov:
103          (1) Replace the "source" command by "." in CMakeLists.txt because
104              "source" is a bash-ism.
106          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
107              without these the CMake build does not work on Solaris.
109    21. Perl has changed its handling of \8 and \9. If there is no previously
110        encountered capturing group of those numbers, they are treated as the
111        literal characters 8 and 9 instead of a binary zero followed by the
112        literals. PCRE now does the same.
114    22. Following Perl, added \o{} to specify codepoints in octal, making it
115        possible to specify values greater than 0777 and also making them
116        unambiguous.
118    23. Perl now gives an error for missing closing braces after \x{... instead of
119        treating the string as literal. PCRE now does the same.
121    24. RunTest used to grumble if an inappropriate test was selected explicitly,
122        but just skip it when running all tests. This make it awkward to run ranges
123        of tests when one of them was inappropriate. Now it just skips any
124        inappropriate tests, as it always did when running all tests.
126    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
127        character types such as \d or \w, too many callouts were inserted, and the
128        data that they returned was rubbish.
130    26. In UCP mode, \s was not matching two of the characters that Perl matches,
131        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
132        were matched by \h. The code has now been refactored so that the lists of
133        the horizontal and vertical whitespace characters used for \h and \v (which
134        are defined only in one place) are now also used for \s.
136    27. Add JIT support for the 64 bit TileGX architecture.
137        Patch by Jiong Wang (Tilera Corporation).
139    28. Possessive quantifiers for classes (both explicit and automatically
140        generated) now use special opcodes instead of wrapping in ONCE brackets.
142    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
143        (because it's meaningless), this was not happening when PCRE_CASELESS was
144        set. Not wrong, but inefficient.
146    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
147        properties for \w, \d, etc) is present in a test regex. Otherwise if the
148        test contains no characters greater than 255, Perl doesn't realise it
149        should be using Unicode semantics.
151    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
152        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
153        does in Unicode mode.
155    32. Added the "forbid" facility to pcretest so that putting tests into the
156        wrong test files can sometimes be quickly detected.
158    33. There is now a limit (default 250) on the depth of nesting of parentheses.
159        This limit is imposed to control the amount of system stack used at compile
160        time. It can be changed at build time by --with-parens-nest-limit=xxx or
161        the equivalent in CMake.
163    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
164        errors. Perl warns for these when in warning mode, but PCRE has no facility
165        for giving warnings.
168    Version 8.33 28-May-2013
169    ------------------------
171    1.  Added 'U' to some constants that are compared to unsigned integers, to
172        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
173        variables that are added to signed variables, to ensure the result is
174        signed and can be negated.
176    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
177        CMake config files.
179    3.  Revise the creation of config.h.generic so that all boolean macros are
180        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
181        overriding via -D on the command line possible.
183    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
184        to unsigned int is reported to make a quite noticeable speed difference in
185        a specific Windows environment. Testing on Linux did also appear to show
186        some benefit (and it is clearly not harmful). Also fixed the definition of
187        Xop which should be unsigned.
189    5.  Related to (4), changing the definition of the intermediate variable cc
190        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
191        improvements.
193    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
194        unnecessary spaces.
196    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
197        and later.
199    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
201    9.  Optimizing fast_forward_start_bits in JIT.
203    10. Adding support for callouts in JIT, and fixing some issues revealed
204        during this work. Namely:
206        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
208        (b) Minimum length was not checked before the matching is started.
210    11. The value of capture_last that is passed to callouts was incorrect in some
211        cases when there was a capture on one path that was subsequently abandoned
212        after a backtrack. Also, the capture_last value is now reset after a
213        recursion, since all captures are also reset in this case.
215    12. The interpreter no longer returns the "too many substrings" error in the
216        case when an overflowing capture is in a branch that is subsequently
217        abandoned after a backtrack.
219    13. In the pathological case when an offset vector of size 2 is used, pcretest
220        now prints out the matched string after a yield of 0 or 1.
222    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
223        Only supported by the JIT compiler at the moment.
225    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
227    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
228        offset of the starting point of the matching process, provided the offsets
229        vector is large enough.
231    17. The \A escape now records a lookbehind value of 1, though its execution
232        does not actually inspect the previous character. This is to ensure that,
233        in partial multi-segment matching, at least one character from the old
234        segment is retained when a new segment is processed. Otherwise, if there
235        are no lookbehinds in the pattern, \A might match incorrectly at the start
236        of a new segment.
238    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
240    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
241        gives some modest performance improvement in 8-bit mode.
243    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
244        be expressed in certain programming languages using Universal Character
245        Names.
247    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
248        which points out that "non characters" are not "characters that may not
249        appear in Unicode strings" but rather "characters that are reserved for
250        internal use and have only local meaning".
252    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
253        there was a conditional group that depended on an assertion, if the
254        assertion was false, the callout that immediately followed the alternation
255        in the condition was skipped when pcre_exec() was used for matching.
257    23. Allow an explicit callout to be inserted before an assertion that is the
258        condition for a conditional group, for compatibility with automatic
259        callouts, which always insert a callout at this point.
261    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
262        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
264    25. (*PRUNE) is now supported by the JIT compiler.
266    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
268    27. Fix the case where there are two or more SKIPs with arguments that may be
269        ignored.
271    28. (*SKIP) is now supported by the JIT compiler.
273    29. (*THEN) is now supported by the JIT compiler.
275    30. Update RunTest with additional test selector options.
277    31. The way PCRE handles backtracking verbs has been changed in two ways.
279        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
280        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
281        backtracking. In some cases this makes it more Perl-compatible, but Perl's
282        rather obscure rules do not always do the same thing.
284        (2) Previously, backtracking verbs were confined within assertions. This is
285        no longer the case for positive assertions, except for (*ACCEPT). Again,
286        this sometimes improves Perl compatibility, and sometimes does not.
288    32. A number of tests that were in test 2 because Perl did things differently
289        have been moved to test 1, because either Perl or PCRE has changed, and
290        these tests are now compatible.
292    32. Backtracking control verbs are now handled in the same way in JIT and
293        interpreter.
295    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
296        contained a forward subroutine reference caused a compile error.
298    34. Auto-detect and optimize limited repetitions in JIT.
300    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
301        blocking (*UTF) etc.
303    36. In the interpreter, maximizing pattern repetitions for characters and
304        character types now use tail recursion, which reduces stack usage.
306    37. The value of the max lookbehind was not correctly preserved if a compiled
307        and saved regex was reloaded on a host of different endianness.
309    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
310        of the compiled pattern block, expand the flags field from 16 to 32 bits
311        because it was almost full.
313    39. Try madvise first before posix_madvise.
315    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
316        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
317        original change locked out all matches of empty strings. This has been
318        changed so that one match of an empty string per line is recognized.
319        Subsequent searches on the same line (for colouring or for --only-matching,
320        for example) do not recognize empty strings.
322    41. Applied a user patch to fix a number of spelling mistakes in comments.
324    42. Data lines longer than 65536 caused pcretest to crash.
326    43. Clarified the data type for length and startoffset arguments for pcre_exec
327        and pcre_dfa_exec in the function-specific man pages, where they were
328        explicitly stated to be in bytes, never having been updated. I also added
329        some clarification to the pcreapi man page.
331    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
332        a segmentation fault.
335    Version 8.32 30-November-2012
336    -----------------------------
338    1.  Improved JIT compiler optimizations for first character search and single
339        character iterators.
341    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
342        Patch by Daniel Richard G.
344    3.  Single character iterator optimizations in the JIT compiler.
346    4.  Improved JIT compiler optimizations for character ranges.
348    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
349        Reported by Giuseppe D'Angelo.
351    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
352        of a line, was being set incorrectly in cases where .* appeared inside
353        atomic brackets at the start of a pattern, or where there was a subsequent
354        *PRUNE or *SKIP.
356    7.  Improved instruction cache flush for POWER/PowerPC.
357        Patch by Daniel Richard G.
359    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
360        grep:
362        (a) There is now no limit to the number of patterns to be matched.
364        (b) An error is given if a pattern is too long.
366        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
367            are now supported.
369        (d) --exclude-from and --include-from (multiple use) have been added.
371        (e) Exclusions and inclusions now apply to all files and directories, not
372            just to those obtained from scanning a directory recursively.
374        (f) Multiple uses of -f and --file-list are now supported.
376        (g) In a Windows environment, the default for -d has been changed from
377            "read" (the GNU grep default) to "skip", because otherwise the presence
378            of a directory in the file list provokes an error.
380        (h) The documentation has been revised and clarified in places.
382    9.  Improve the matching speed of capturing brackets.
384    10. Changed the meaning of \X so that it now matches a Unicode extended
385        grapheme cluster.
387    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
388        out POSIX threads when JIT support is configured.
390    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
392    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
393        structure could go wrong in environments where size_t is not the same size
394        as int.
396    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
397        set.
399    15. The EBCDIC support had decayed; later updates to the code had included
400        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
401        general tidy up of EBCDIC-related issues, and the documentation was also
402        not quite right. There is now a test that can be run on ASCII systems to
403        check some of the EBCDIC-related things (but is it not a full test).
405    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
406        in a small tidy to the code.
408    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
410    18. If the --only-matching (-o) option in pcregrep is specified multiple
411        times, each one causes appropriate output. For example, -o1 -o2 outputs the
412        substrings matched by the 1st and 2nd capturing parentheses. A separating
413        string can be specified by --om-separator (default empty).
415    19. Improving the first n character searches.
417    20. Turn case lists for horizontal and vertical white space into macros so that
418        they are defined only once.
420    21. This set of changes together give more compatible Unicode case-folding
421        behaviour for characters that have more than one other case when UCP
422        support is available.
424        (a) The Unicode property table now has offsets into a new table of sets of
425            three or more characters that are case-equivalent. The MultiStage2.py
426            script that generates these tables (the pcre_ucd.c file) now scans
427            CaseFolding.txt instead of UnicodeData.txt for character case
428            information.
430        (b) The code for adding characters or ranges of characters to a character
431            class has been abstracted into a generalized function that also handles
432            case-independence. In UTF-mode with UCP support, this uses the new data
433            to handle characters with more than one other case.
435        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
436            whose other case is greater than 256 are now correctly matched
437            caselessly. Previously, the high codepoint matched the low one, but not
438            vice versa.
440        (d) The processing of \h, \H, \v, and \ in character classes now makes use
441            of the new class addition function, using character lists defined as
442            macros alongside the case definitions of 20 above.
444        (e) Caseless back references now work with characters that have more than
445            one other case.
447        (f) General caseless matching of characters with more than one other case
448            is supported.
450    22. Unicode character properties were updated from Unicode 6.2.0
452    23. Improved CMake support under Windows. Patch by Daniel Richard G.
454    24. Add support for 32-bit character strings, and UTF-32
456    25. Major JIT compiler update (code refactoring and bugfixing).
457        Experimental Sparc 32 support is added.
459    26. Applied a modified version of Daniel Richard G's patch to create
460        pcre.h.generic and config.h.generic by "make" instead of in the
461        PrepareRelease script.
463    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
464        pcre_compile.c when checking for a zero character.
466    28. Introducing a native interface for JIT. Through this interface, the compiled
467        machine code can be directly executed. The purpose of this interface is to
468        provide fast pattern matching, so several sanity checks are not performed.
469        However, feature tests are still performed. The new interface provides
470        1.4x speedup compared to the old one.
472    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
473        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
474        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
476    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
477        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
478        the "old" RFC 2279). Instead, it ended up passing a negative length to
479        pcre_exec().
481    31. Add support for GCC's visibility feature to hide internal functions.
483    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
484        "unknown -C option" after outputting 0 or 1.
486    33. There is now support for generating a code coverage report for the test
487        suite in environments where gcc is the compiler and lcov is installed. This
488        is mainly for the benefit of the developers.
490    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
491        unaddressable using valgrind annotations, allowing valgrind to detect
492        invalid memory accesses. This is mainly for the benefit of the developers.
494    25. (*UTF) can now be used to start a pattern in any of the three libraries.
496    26. Give configure error if --enable-cpp but no C++ compiler found.
499    Version 8.31 06-July-2012
500    -------------------------
502    1.  Fixing a wrong JIT test case and some compiler warnings.
504    2.  Removed a bashism from the RunTest script.
506    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
507        to unsigned type, result still unsigned" that was given by an MS compiler
508        on encountering the code "-sizeof(xxx)".
510    4.  Partial matching support is added to the JIT compiler.
512    5.  Fixed several bugs concerned with partial matching of items that consist
513        of more than one character:
515        (a) /^(..)\1/ did not partially match "aba" because checking references was
516            done on an "all or nothing" basis. This also applied to repeated
517            references.
519        (b) \R did not give a hard partial match if \r was found at the end of the
520            subject.
522        (c) \X did not give a hard partial match after matching one or more
523            characters at the end of the subject.
525        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
526            a partial match for the string "\r".
528        (e) When newline was set to CRLF, the metacharacter "." did not recognize
529            a partial match for a CR character at the end of the subject string.
531    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
532        running pcretest, the text "(JIT)" added to the output whenever JIT is
533        actually used to run the match.
535    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
536        or /S+[+] with a digit between 1 and 7.
538    8.  OP_NOT now supports any UTF character not just single-byte ones.
540    9.  (*MARK) control verb is now supported by the JIT compiler.
542    10. The command "./RunTest list" lists the available tests without actually
543        running any of them. (Because I keep forgetting what they all are.)
547    12. Applied a (slightly modified) user-supplied patch that improves performance
548        when the heap is used for recursion (compiled with --disable-stack-for-
549        recursion). Instead of malloc and free for each heap frame each time a
550        logical recursion happens, frames are retained on a chain and re-used where
551        possible. This sometimes gives as much as 30% improvement.
553    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
554        call.
556    14. As documented, (*COMMIT) is now confined to within a positive assertion.
558    15. It is now possible to link pcretest with libedit as an alternative to
559        libreadline.
561    16. (*COMMIT) control verb is now supported by the JIT compiler.
563    17. The Unicode data tables have been updated to Unicode 6.1.0.
565    18. Added --file-list option to pcregrep.
567    19. Added binary file support to pcregrep, including the -a, --binary-files,
568        -I, and --text options.
570    20. The madvise function is renamed for posix_madvise for QNX compatibility
571        reasons. Fixed by Giuseppe D'Angelo.
573    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
575    22. Changed the option for creating symbolic links for 16-bit man pages from
576        -s to -sf so that re-installing does not cause issues.
578    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
580    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
581        matches in certain environments (the workspace was not being correctly
582        retained). Also added to pcre_dfa_exec() a simple plausibility check on
583        some of the workspace data at the beginning of a restart.
585    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
586        was not doing so when it should - probably a typo introduced by SVN 528
587        (change 8.10/14).
589    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
590        \w+ when the character tables indicated that \x{c4} was a word character.
591        There were several related cases, all because the tests for doing a table
592        lookup were testing for characters less than 127 instead of 255.
594    27. If a pattern contains capturing parentheses that are not used in a match,
595        their slots in the ovector are set to -1. For those that are higher than
596        any matched groups, this happens at the end of processing. In the case when
597        there were back references that the ovector was too small to contain
598        (causing temporary malloc'd memory to be used during matching), and the
599        highest capturing number was not used, memory off the end of the ovector
600        was incorrectly being set to -1. (It was using the size of the temporary
601        memory instead of the true size.)
603    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
604        ovector size, it uses memory at the end of the block that it has got.
606    29. Check for an overlong MARK name and give an error at compile time. The
607        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
609    30. JIT compiler update.
611    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
612        Rill for the patch.
614    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
616    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
618    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
619        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
620        was enabled.
622    35. Improve JIT code generation for greedy plus quantifier.
624    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
625        1 to "aa" instead of to an empty string. The bug affected repeated groups
626        that could potentially match an empty string.
628    37. Optimizing single character iterators in JIT.
630    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
631        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
632        codepoints that are too big for the mode are faulted, and in a UTF mode,
633        disallowed codepoints are also faulted.
635    39. If PCRE was compiled with UTF support, in three places in the DFA
636        matcher there was code that should only have been obeyed in UTF mode, but
637        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
638        processing when bytes with values greater than 127 were present. In 16-bit
639        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
640        both cases the values are those that cannot be the first data item in a UTF
641        character. The three items that might have provoked this were recursions,
642        possessively repeated groups, and atomic groups.
644    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
645        and pcregrep, because some OS require shared objects to be explicitly
646        passed to ld, causing the link step to fail if they are not.
648    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
649        mode, patterns that started with \h* or \R* might be incorrectly matched.
652    Version 8.30 04-February-2012
653    -----------------------------
655    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
656        name is defined in ctype.h.
658    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
659        only in quite long subpatterns.
661    3.  Removed the function pcre_info(), which has been obsolete and deprecated
662        since it was replaced by pcre_fullinfo() in February 2000.
664    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
665        match a (*MARK), and the match failed at the start of the subject, a
666        reference to memory before the start of the subject could occur. This bug
667        was introduced by fix 17 of release 8.21.
669    5.  A reference to an unset group with zero minimum repetition was giving
670        totally wrong answers (in non-JavaScript-compatibility mode). For example,
671        /(another)?(\1?)test/ matched against "hello world test". This bug was
672        introduced in release 8.13.
674    6.  Add support for 16-bit character strings (a large amount of work involving
675        many changes and refactorings).
677    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
678        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
679        from a file.
681    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
682        rounding is not applied in this particular case).
684    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
685        if they appear, or are escaped, in patterns.
687    10. Get rid of a number of -Wunused-but-set-variable warnings.
689    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
690        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
691        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
692        also returns the mark "x". This bug applied to capturing parentheses,
693        non-capturing parentheses, and atomic parentheses. It also applied to some
694        assertions.
696    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
697        information out of configure.ac instead of relying on pcre.h.generic, which
698        is not stored in the repository.
700    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
701        -lreadline.
703    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
705    15. Applied Graycode's patch to put the top-level frame on the stack rather
706        than the heap when not using the stack for recursion. This gives a
707        performance improvement in many cases when recursion is not deep.
709    16. Experimental code added to "pcretest -C" to output the stack frame size.
712    Version 8.21 12-Dec-2011
713    ------------------------
715    1.  Updating the JIT compiler.
717    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
718        are added as well.
720    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
721        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
722        calling _pcre_jit_exec. Some extra comments are added.
724    4.  (*MARK) settings inside atomic groups that do not contain any capturing
725        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
726        was introduced by change 18 for 8.20.
728    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
729        ECMA-262 standard.
731    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
732        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
733        This bug was probably introduced by change 9 of 8.13.
735    7.  While fixing 6 above, I noticed that a number of other items were being
736        incorrectly rejected as "not fixed length". This arose partly because newer
737        opcodes had not been added to the fixed-length checking code. I have (a)
738        corrected the bug and added tests for these items, and (b) arranged for an
739        error to occur if an unknown opcode is encountered while checking for fixed
740        length instead of just assuming "not fixed length". The items that were
741        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
742        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
743        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
745    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
746        being incorrectly compiled and would have given unpredicatble results.
748    9.  A possessively repeated subpattern with minimum repeat count greater than
749        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
750        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
751        the first (A) could occur when it should not.
753    10. Add a cast and remove a redundant test from the code.
755    11. JIT should use pcre_malloc/pcre_free for allocation.
757    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
758        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
759        is anything other than /usr, -L is still shown).
761    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
763    14. Perl does not support \N without a following name in a [] class; PCRE now
764        also gives an error.
766    15. If a forward reference was repeated with an upper limit of around 2000,
767        it caused the error "internal error: overran compiling workspace". The
768        maximum number of forward references (including repeats) was limited by the
769        internal workspace, and dependent on the LINK_SIZE. The code has been
770        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
771        the default depends on LINK_SIZE. There is a new upper limit (for safety)
772        of around 200,000 forward references. While doing this, I also speeded up
773        the filling in of repeated forward references.
775    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
776        incorrectly expecting the subject to contain another "a" after the start.
778    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
779        in the match, the SKIP should be ignored. This was not happening; instead
780        the SKIP was being treated as NOMATCH. For patterns such as
781        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
782        tested.
784    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
785        now much more compatible with Perl, in particular in cases where the result
786        is a non-match for a non-anchored pattern. For example, if
787        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
788        "m", where previously it did not return a name. A side effect of this
789        change is that for partial matches, the last encountered mark name is
790        returned, as for non matches. A number of tests that were previously not
791        Perl-compatible have been moved into the Perl-compatible test files. The
792        refactoring has had the pleasing side effect of removing one argument from
793        the match() function, thus reducing its stack requirements.
795    19. If the /S+ option was used in pcretest to study a pattern using JIT,
796        subsequent uses of /S (without +) incorrectly behaved like /S+.
798    21. Retrieve executable code size support for the JIT compiler and fixing
799        some warnings.
801    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
802        not work when the shorter character appeared right at the end of the
803        subject string.
805    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
806        systems.
808    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
809        output it when the /M option is used in pcretest.
811    25. The CheckMan script was not being included in the distribution. Also, added
812        an explicit "perl" to run Perl scripts from the PrepareRelease script
813        because this is reportedly needed in Windows.
815    26. If study data was being save in a file and studying had not found a set of
816        "starts with" bytes for the pattern, the data written to the file (though
817        never used) was taken from uninitialized memory and so caused valgrind to
818        complain.
820    27. Updated RunTest.bat as provided by Sheri Pierce.
822    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
824    29. Computation of memory usage for the table of capturing group names was
825        giving an unnecessarily large value.
828    Version 8.20 21-Oct-2011
829    ------------------------
831    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
832        a POSIX class. After further experiments with Perl, which convinced me that
833        Perl has bugs and confusions, a closing square bracket is no longer allowed
834        in a POSIX name. This bug also affected patterns with classes that started
835        with full stops.
837    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
838        captured substring, but while checking the failing first alternative,
839        substring 1 is temporarily captured. If the output vector supplied to
840        pcre_exec() was not big enough for this capture, the yield of the function
841        was still zero ("insufficient space for captured substrings"). This cannot
842        be totally fixed without adding another stack variable, which seems a lot
843        of expense for a edge case. However, I have improved the situation in cases
844        such as /(a)(b)x|abc/ matched against "abc", where the return code
845        indicates that fewer than the maximum number of slots in the ovector have
846        been set.
848    3.  Related to (2) above: when there are more back references in a pattern than
849        slots in the output vector, pcre_exec() uses temporary memory during
850        matching, and copies in the captures as far as possible afterwards. It was
851        using the entire output vector, but this conflicts with the specification
852        that only 2/3 is used for passing back captured substrings. Now it uses
853        only the first 2/3, for compatibility. This is, of course, another edge
854        case.
856    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
857        main code base, and can be used by building with --enable-jit. When this is
858        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
859        runtime --no-jit option is given.
861    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
862        ovector, the return from the function was zero, implying that there were
863        other matches that did not fit. The correct "exactly full" value is now
864        returned.
866    6.  If a subpattern that was called recursively or as a subroutine contained
867        (*PRUNE) or any other control that caused it to give a non-standard return,
868        invalid errors such as "Error -26 (nested recursion at the same subject
869        position)" or even infinite loops could occur.
871    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
872        computing the minimum length on reaching *ACCEPT, and so ended up with the
873        wrong value of 1 rather than 0. Further investigation indicates that
874        computing a minimum subject length in the presence of *ACCEPT is difficult
875        (think back references, subroutine calls), and so I have changed the code
876        so that no minimum is registered for a pattern that contains *ACCEPT.
878    8.  If (*THEN) was present in the first (true) branch of a conditional group,
879        it was not handled as intended. [But see 16 below.]
881    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
882        Sheri Pierce.
884    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
885        the first byte in a match must be "a".
887    11. Change 17 for 8.13 increased the recursion depth for patterns like
888        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
889        pattern contains any instances of (*THEN). If it does not, the old
890        optimizations are restored. It would be nice to do this on a per-group
891        basis, but at the moment that is not feasible.
893    12. In some environments, the output of pcretest -C is CRLF terminated. This
894        broke RunTest's code that checks for the link size. A single white space
895        character after the value is now allowed for.
897    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
898        For "fr", it uses the Windows-specific input and output files.
900    14. If (*THEN) appeared in a group that was called recursively or as a
901        subroutine, it did not work as intended. [But see next item.]
903    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
904        pattern fragments (but not containing any | characters). If A and B are
905        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
906        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
907        D. In other words, Perl considers parentheses that do not contain any |
908        characters to be part of a surrounding alternative, whereas PCRE was
909        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
910        differently. PCRE now behaves in the same way as Perl, except in the case
911        of subroutine/recursion calls such as (?1) which have in any case always
912        been different (but PCRE had them first :-).
914    16. Related to 15 above: Perl does not treat the | in a conditional group as
915        creating alternatives. Such a group is treated in the same way as an
916        ordinary group without any | characters when processing (*THEN). PCRE has
917        been changed to match Perl's behaviour.
919    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
920        RunGrepTest script failed.
922    18. Change 22 for version 13 caused atomic groups to use more stack. This is
923        inevitable for groups that contain captures, but it can lead to a lot of
924        stack use in large patterns. The old behaviour has been restored for atomic
925        groups that do not contain any capturing parentheses.
927    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
928        suppress the check for a minimum subject length at run time. (If it was
929        given to pcre_exec() or pcre_dfa_exec() it did work.)
931    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
932        fail to work when decoding hex characters in data strings in EBCDIC
933        environments.
935    21. It appears that in at least one Mac OS environment, the isxdigit() function
936        is implemented as a macro that evaluates to its argument more than once,
937        contravening the C 90 Standard (I haven't checked a later standard). There
938        was an instance in pcretest which caused it to go wrong when processing
939        \x{...} escapes in subject strings. The has been rewritten to avoid using
940        things like p++ in the argument of isxdigit().
943    Version 8.13 16-Aug-2011
944    ------------------------
946    1.  The Unicode data tables have been updated to Unicode 6.0.0.
948    2.  Two minor typos in pcre_internal.h have been fixed.
950    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
951        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
952        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
954    4.  There were a number of related bugs in the code for matching backrefences
955        caselessly in UTF-8 mode when codes for the characters concerned were
956        different numbers of bytes. For example, U+023A and U+2C65 are an upper
957        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
958        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
959        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
960        2-byte code at the end of the subject (it thought there wasn't enough data
961        left).
963    5.  Comprehensive information about what went wrong is now returned by
964        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
965        as the output vector has at least 2 elements. The offset of the start of
966        the failing character and a reason code are placed in the vector.
968    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
969        now returned is for the first byte of the failing character, instead of the
970        last byte inspected. This is an incompatible change, but I hope it is small
971        enough not to be a problem. It makes the returned offset consistent with
972        pcre_exec() and pcre_dfa_exec().
974    7.  pcretest now gives a text phrase as well as the error number when
975        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
976        failure, the offset and reason code are output.
978    8.  When \R was used with a maximizing quantifier it failed to skip backwards
979        over a \r\n pair if the subsequent match failed. Instead, it just skipped
980        back over a single character (\n). This seems wrong (because it treated the
981        two characters as a single entity when going forwards), conflicts with the
982        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
983        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
984        has been changed.
986    9.  Some internal refactoring has changed the processing so that the handling
987        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
988        time (the PCRE_DOTALL option was changed this way some time ago: version
989        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
990        which was always a bit of a fudge. It also means that there is one less
991        argument for the match() function, which reduces its stack requirements
992        slightly. This change also fixes an incompatibility with Perl: the pattern
993        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
995    10. More internal refactoring has drastically reduced the number of recursive
996        calls to match() for possessively repeated groups such as (abc)++ when
997        using pcre_exec().
999    11. While implementing 10, a number of bugs in the handling of groups were
1000        discovered and fixed:
1002        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1003        (a|)*(?1) gave a compile-time internal error.
1004        ((a|)+)+  did not notice that the outer group could match an empty string.
1005        (^a|^)+   was not marked as anchored.
1006        (.*a|.*)+ was not marked as matching at start or after a newline.
1008    12. Yet more internal refactoring has removed another argument from the match()
1009        function. Special calls to this function are now indicated by setting a
1010        value in a variable in the "match data" data block.
1012    13. Be more explicit in pcre_study() instead of relying on "default" for
1013        opcodes that mean there is no starting character; this means that when new
1014        ones are added and accidentally left out of pcre_study(), testing should
1015        pick them up.
1017    14. The -s option of pcretest has been documented for ages as being an old
1018        synonym of -m (show memory usage). I have changed it to mean "force study
1019        for every regex", that is, assume /S for every regex. This is similar to -i
1020        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1021        using it. It makes it easier to run collections of tests with and without
1022        study enabled, and thereby test pcre_study() more easily. All the standard
1023        tests are now run with and without -s (but some patterns can be marked as
1024        "never study" - see 20 below).
1026    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1027        restoration of the capturing data to the outer values was not happening
1028        correctly.
1030    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1031        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1032        pattern had matched an empty string, and so incorrectly returned a no
1033        match.
1035    17. There was optimizing code for the last branch of non-capturing parentheses,
1036        and also for the obeyed branch of a conditional subexpression, which used
1037        tail recursion to cut down on stack usage. Unfortunately, now that there is
1038        the possibility of (*THEN) occurring in these branches, tail recursion is
1039        no longer possible because the return has to be checked for (*THEN). These
1040        two optimizations have therefore been removed. [But see 8.20/11 above.]
1042    18. If a pattern containing \R was studied, it was assumed that \R always
1043        matched two bytes, thus causing the minimum subject length to be
1044        incorrectly computed because \R can also match just one byte.
1046    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1047        was incorrectly computed.
1049    20. If /S is present twice on a test pattern in pcretest input, it now
1050        *disables* studying, thereby overriding the use of -s on the command line
1051        (see 14 above). This is necessary for one or two tests to keep the output
1052        identical in both cases.
1054    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1055        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1057    22. When an atomic group that contained a capturing parenthesis was
1058        successfully matched, but the branch in which it appeared failed, the
1059        capturing was not being forgotten if a higher numbered group was later
1060        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1061        group 1 to "a", when in fact it should be unset. This applied to multi-
1062        branched capturing and non-capturing groups, repeated or not, and also to
1063        positive assertions (capturing in negative assertions does not happen
1064        in PCRE) and also to nested atomic groups.
1066    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1067        subject after a captured substring, to make it easier to tell which of a
1068        number of identical substrings has been captured.
1070    24. The way atomic groups are processed by pcre_exec() has been changed so that
1071        if they are repeated, backtracking one repetition now resets captured
1072        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1073        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1074        "aaa". Previously, it would have been "a". As part of this code
1075        refactoring, the way recursive calls are handled has also been changed.
1077    25. If an assertion condition captured any substrings, they were not passed
1078        back unless some other capturing happened later. For example, if
1079        (?(?=(a))a) was matched against "a", no capturing was returned.
1081    26. When studying a pattern that contained subroutine calls or assertions,
1082        the code for finding the minimum length of a possible match was handling
1083        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1084        group 1 called group 2 while simultaneously a separate group 2 called group
1085        1). A stack overflow occurred in this case. I have fixed this by limiting
1086        the recursion depth to 10.
1088    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1089        Fortmann. This supports explicit test numbers on the command line, and has
1090        argument validation and error reporting.
1092    28. An instance of \X with an unlimited repeat could fail if at any point the
1093        first character it looked at was a mark character.
1095    29. Some minor code refactoring concerning Unicode properties and scripts
1096        should reduce the stack requirement of match() slightly.
1098    30. Added the '=' option to pcretest to check the setting of unused capturing
1099        slots at the end of the pattern, which are documented as being -1, but are
1100        not included in the return count.
1102    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1103        compiled something random. Now it gives a compile-time error (as does
1104        Perl).
1106    32. A *MARK encountered during the processing of a positive assertion is now
1107        recorded and passed back (compatible with Perl).
1109    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1110        had alternative anchored branches, the search for a second match in a line
1111        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1112        matched the line "0102" twice. The same bug affected patterns that started
1113        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1114        twice.
1116    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1117        does, and because of capturing effects, quantifying parenthesized
1118        assertions may at times be useful. Quantifiers are now allowed for
1119        parenthesized assertions.
1121    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1123    36. \g was being checked for fancy things in a character class, when it should
1124        just be a literal "g".
1126    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1127        appearance of a nested POSIX class supersedes an apparent external class.
1128        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1129        unescaped square brackets may also appear as part of class names. For
1130        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1131        more like Perl. (But see 8.20/1 above.)
1133    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1134        was because it thought it was \N{name}, which is not supported).
1136    39. Add minix to OS list not supporting the -S option in pcretest.
1138    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1139        cannot analyze patterns in sufficient detail to catch mutual recursions
1140        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1141        subgroup is called recursively as a subpattern for a second time at the
1142        same position in the subject string. In previous releases this might have
1143        been caught by the recursion limit, or it might have run out of stack.
1145    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1146        happen only once. PCRE was, however incorrectly giving a compile time error
1147        "recursive call could loop indefinitely" because it cannot analyze the
1148        pattern in sufficient detail. The compile time test no longer happens when
1149        PCRE is compiling a conditional subpattern, but actual runaway loops are
1150        now caught at runtime (see 40 above).
1152    42. It seems that Perl allows any characters other than a closing parenthesis
1153        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1154        has been changed to be the same.
1156    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1157        as not to get warnings when autogen.sh is called. Also changed
1158        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1160    44. To help people who use pcregrep to scan files containing exceedingly long
1161        lines, the following changes have been made:
1163        (a) The default value of the buffer size parameter has been increased from
1164            8K to 20K. (The actual buffer used is three times this size.)
1166        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1167            PCRE is built.
1169        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1170            to be set at run time.
1172        (d) Numerical values in pcregrep options can be followed by K or M, for
1173            example --buffer-size=50K.
1175        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1176            given and the return code is set to 2.
1178    45. Add a pointer to the latest mark to the callout data block.
1180    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1181        partial match of an empty string instead of no match. This was specific to
1182        the use of ".".
1184    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1185        complete match instead of a partial match. This bug was dependent on both
1186        the PCRE_UTF8 and PCRE_DOTALL options being set.
1188    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1189        starting byte set, because \b was not being ignored.
1192    Version 8.12 15-Jan-2011
1193    ------------------------
1195    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1196        checks for such things as part of the documentation building process.
1198    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1199        --match-limit and --recursion-limit options (added for 8.11). In
1200        particular, this made one of the standard tests fail. (The integer value
1201        went into the wrong half of a long int.)
1203    3.  If the --colour option was given to pcregrep with -v (invert match), it
1204        did strange things, either producing crazy output, or crashing. It should,
1205        of course, ignore a request for colour when reporting lines that do not
1206        match.
1208    4.  Another pcregrep bug caused similar problems if --colour was specified with
1209        -M (multiline) and the pattern match finished with a line ending.
1211    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1212        matched in multiline mode, the following line was shown as part of the
1213        match. This seems wrong, so I have changed it.
1215    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1216        the check for further matches in the same line (so they could be coloured)
1217        to overrun the end of the current line. If another match was found, it was
1218        incorrectly shown (and then shown again when found in the next line).
1220    7.  If pcregrep was compiled under Windows, there was a reference to the
1221        function pcregrep_exit() before it was defined. I am assuming this was
1222        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1223        reported by a user. I've moved the definition above the reference.
1226    Version 8.11 10-Dec-2010
1227    ------------------------
1229    1.  (*THEN) was not working properly if there were untried alternatives prior
1230        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1231        backtracked to try for "b" instead of moving to the next alternative branch
1232        at the same level (in this case, to look for "c"). The Perl documentation
1233        is clear that when (*THEN) is backtracked onto, it goes to the "next
1234        alternative in the innermost enclosing group".
1236    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1237        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1238        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1239        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1240        (*THEN).
1242    3.  If \s appeared in a character class, it removed the VT character from
1243        the class, even if it had been included by some previous item, for example
1244        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1245        of \s, but is part of the POSIX "space" class.)
1247    4.  A partial match never returns an empty string (because you can always
1248        match an empty string at the end of the subject); however the checking for
1249        an empty string was starting at the "start of match" point. This has been
1250        changed to the "earliest inspected character" point, because the returned
1251        data for a partial match starts at this character. This means that, for
1252        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1253        (previously it gave "no match").
1255    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1256        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1257        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1258        has an implication that the given string is incomplete (because a partial
1259        match is preferred over a full match). For this reason, these items now
1260        give a partial match in this situation. [Aside: previously, the one case
1261        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1262        match rather than a full match, which was wrong by the old rules, but is
1263        now correct.]
1265    6.  There was a bug in the handling of #-introduced comments, recognized when
1266        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1267        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1268        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1269        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1270        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1271        places in pcre_compile().
1273    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1274        comments when looking ahead for named forward references to subpatterns,
1275        the only newline sequence it recognized was NL. It now handles newlines
1276        according to the set newline convention.
1278    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1279        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1280        cater for a lack of strerror(). These oversights have been fixed.
1282    9.  Added --match-limit and --recursion-limit to pcregrep.
1284    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1286    11. When the -o option was used, pcregrep was setting a return code of 1, even
1287        when matches were found, and --line-buffered was not being honoured.
1289    12. Added an optional parentheses number to the -o and --only-matching options
1290        of pcregrep.
1292    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1293        can match an empty string. The code to do it in pcretest and pcredemo
1294        needed fixing:
1296        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1297            only one byte after an empty string match just before CRLF (this case
1298            just got forgotten; "any" and "anycrlf" were OK).
1300        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1301            mode when an empty string match preceded an ASCII character followed by
1302            a non-ASCII character. (The code for advancing by one character rather
1303            than one byte was nonsense.)
1305        (c) The pcredemo.c sample program did not have any code at all to handle
1306            the cases when CRLF is a valid newline sequence.
1308    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1309        as a starting offset was within the subject string. There is now a new
1310        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1311        negative or greater than the length of the string. In order to test this,
1312        pcretest is extended to allow the setting of negative starting offsets.
1314    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1315        starting offset points to the beginning of a UTF-8 character was
1316        unnecessarily clumsy. I tidied it up.
1318    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1319        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1321    17. Nobody had reported that the --include_dir option, which was added in
1322        release 7.7 should have been called --include-dir (hyphen, not underscore)
1323        for compatibility with GNU grep. I have changed it to --include-dir, but
1324        left --include_dir as an undocumented synonym, and the same for
1325        --exclude-dir, though that is not available in GNU grep, at least as of
1326        release 2.5.4.
1328    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1329        characters from a string of bytes) have been redefined so as not to use
1330        loops, in order to improve performance in some environments. At the same
1331        time, I abstracted some of the common code into auxiliary macros to save
1332        repetition (this should not affect the compiled code).
1334    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1335        compile-time error is now given if \c is not followed by an ASCII
1336        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1337        different, and any byte value is allowed.)
1339    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1340        START_OPTIMIZE option, which is now allowed at compile time - but just
1341        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1342        to pcregrep and other applications that have no direct access to PCRE
1343        options. The new /Y option in pcretest sets this option when calling
1344        pcre_compile().
1346    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1347        back references. Groups containing recursive back references were forced to
1348        be atomic by that change, but in the case of named groups, the amount of
1349        memory required was incorrectly computed, leading to "Failed: internal
1350        error: code overflow". This has been fixed.
1352    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1353        pcretest.c, to avoid build problems in some Borland environments.
1356    Version 8.10 25-Jun-2010
1357    ------------------------
1359    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1360        THEN.
1362    2.  (*ACCEPT) was not working when inside an atomic group.
1364    3.  Inside a character class, \B is treated as a literal by default, but
1365        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1366        causes the error). The code is unchanged, but I tidied the documentation.
1368    4.  Inside a character class, PCRE always treated \R and \X as literals,
1369        whereas Perl faults them if its -w option is set. I have changed PCRE so
1370        that it faults them when PCRE_EXTRA is set.
1372    5.  Added support for \N, which always matches any character other than
1373        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1375    6.  When compiling pcregrep with newer versions of gcc which may have
1376        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1377        declared with attribute warn_unused_result" were given. Just casting the
1378        result to (void) does not stop the warnings; a more elaborate fudge is
1379        needed. I've used a macro to implement this.
1381    7.  Minor change to pcretest.c to avoid a compiler warning.
1383    8.  Added four artifical Unicode properties to help with an option to make
1384        \s etc use properties (see next item). The new properties are: Xan
1385        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1387    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1388        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1389        this option. Modified pcretest to add /W to test this facility. Added
1390        REG_UCP to make it available via the POSIX interface.
1392    10. Added --line-buffered to pcregrep.
1394    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1395        studied, and the match started with a letter with a code point greater than
1396        127 whose first byte was different to the first byte of the other case of
1397        the letter, the other case of this starting letter was not recognized
1398        (#976).
1400    12. If a pattern that was studied started with a repeated Unicode property
1401        test, for example, \p{Nd}+, there was the theoretical possibility of
1402        setting up an incorrect bitmap of starting bytes, but fortunately it could
1403        not have actually happened in practice until change 8 above was made (it
1404        added property types that matched character-matching opcodes).
1406    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1407        possible starting bytes for non-anchored patterns.
1409    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1410        \R, and also a number of cases that involve Unicode properties, both
1411        explicit and implicit when PCRE_UCP is set.
1413    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1414        input, it could crash or give wrong results if characters with values
1415        greater than 0xc0 were present in the subject string. (Detail: it assumed
1416        UTF-8 input when processing these items.)
1418    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1419        size_t is 64-bit (#991).
1421    17. Added a check for running out of memory when PCRE is compiled with
1422        --disable-stack-for-recursion (#990).
1424    18. If the last data line in a file for pcretest does not have a newline on
1425        the end, a newline was missing in the output.
1427    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1428        less than 128) in its various bitmaps. However, there is a facility for
1429        generating tables according to the current locale when PCRE is compiled. It
1430        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1431        characters, are recognized by isspace() and therefore were getting set in
1432        these tables, and indeed these tables seem to approximate to ISO 8859. This
1433        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1434        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1435        which of course cannot start UTF-8 characters. I have changed the code so
1436        that only real ASCII characters (less than 128) and the correct starting
1437        bytes for UTF-8 encodings are set for characters greater than 127 when in
1438        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1439        altogether.)
1441    20. Added the /T option to pcretest so as to be able to run tests with non-
1442        standard character tables, thus making it possible to include the tests
1443        used for 19 above in the standard set of tests.
1445    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1446        reference to a subpattern the other side of a comment that contains an
1447        opening parenthesis caused either an internal compiling error, or a
1448        reference to the wrong subpattern.
1451    Version 8.02 19-Mar-2010
1452    ------------------------
1454    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1456    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1457        configured.
1459    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1460        original author of that file, following a query about its status.
1462    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1463        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1465    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1466        quantifier applied to a forward-referencing subroutine call, could compile
1467        incorrect code or give the error "internal error: previously-checked
1468        referenced subpattern not found".
1470    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1471        variables to point to external functions. For these systems, therefore,
1472        pcre_malloc etc. are now initialized to local functions that call the
1473        relevant global functions.
1475    7.  There were two entries missing in the vectors called coptable and poptable
1476        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1477        I've fixed the data, and added a kludgy way of testing at compile time that
1478        the lengths are correct (equal to the number of opcodes).
1480    8.  Following on from 7, I added a similar kludge to check the length of the
1481        eint vector in pcreposix.c.
1483    9.  Error texts for pcre_compile() are held as one long string to avoid too
1484        much relocation at load time. To find a text, the string is searched,
1485        counting zeros. There was no check for running off the end of the string,
1486        which could happen if a new error number was added without updating the
1487        string.
1489    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1491    11. \K was not working if it appeared in an atomic group or in a group that
1492        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1493        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1494        the assertion is positive, but not if it is negative.
1496    12. Change 11 fortuitously reduced the size of the stack frame used in the
1497        "match()" function of pcre_exec.c by one pointer. Forthcoming
1498        implementation of support for (*MARK) will need an extra pointer on the
1499        stack; I have reserved it now, so that the stack frame size does not
1500        decrease.
1502    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1503        item in branch that calls a recursion is a subroutine call - as in the
1504        second branch in the above example - was incorrectly given the compile-
1505        time error "recursive call could loop indefinitely" because pcre_compile()
1506        was not correctly checking the subroutine for matching a non-empty string.
1508    14. The checks for overrunning compiling workspace could trigger after an
1509        overrun had occurred. This is a "should never occur" error, but it can be
1510        triggered by pathological patterns such as hundreds of nested parentheses.
1511        The checks now trigger 100 bytes before the end of the workspace.
1513    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1516    Version 8.01 19-Jan-2010
1517    ------------------------
1519    1.  If a pattern contained a conditional subpattern with only one branch (in
1520        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1521        computed the wrong minimum data length (which is of course zero for such
1522        subpatterns). This could cause incorrect "no match" results.
1524    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1525        the pattern is reset in the first branch, pcre_compile() failed with
1526        "internal error: code overflow at offset...". This happened only when
1527        the reset was to the original external option setting. (An optimization
1528        abstracts leading options settings into an external setting, which was the
1529        cause of this.)
1531    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1532        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1533        assertion pattern did not match (meaning that the assertion was true), it
1534        was incorrectly treated as false if the SKIP had been reached during the
1535        matching. This also applied to assertions used as conditions.
1537    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1538        assertion subpattern, including such a pattern used as a condition,
1539        unpredictable results occurred, instead of the error return
1542    5.  The C++ GlobalReplace function was not working like Perl for the special
1543        situation when an empty string is matched. It now does the fancy magic
1544        stuff that is necessary.
1546    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1547        removed. (These were left over from very, very early versions of PCRE.)
1549    7.  Some cosmetic changes to the code to make life easier when compiling it
1550        as part of something else:
1552        (a) Change DEBUG to PCRE_DEBUG.
1554        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1555            called "current" as "current_branch", to prevent a collision with the
1556            Linux macro when compiled as a kernel module.
1558        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1559            prevent a collision with the Linux macro when compiled as a kernel
1560            module.
1562    8.  In pcre_compile() there are some checks for integer overflows that used to
1563        cast potentially large values to (double). This has been changed to that
1564        when building, a check for int64_t is made, and if it is found, it is used
1565        instead, thus avoiding the use of floating point arithmetic. (There is no
1566        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1567        double.
1569    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1570        2005 (difference between two addresses compared to an unsigned value).
1572    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1573        custom one, because of the following reported problem in Windows:
1575          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1576              under Win32.
1577          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1578              therefore missing the function definition.
1579          - The compiler thus generates a "C" signature for the test function.
1580          - The linker fails to find the "C" function.
1581          - PCRE fails to configure if asked to do so against libbz2.
1583    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1584        messages were output:
1586          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1587          rerunning libtoolize, to keep the correct libtool macros in-tree.
1588          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1590        I have done both of these things.
1592    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1593        most of the time, it *can* run out if it is given a pattern that contains a
1594        runaway infinite recursion. I updated the discussion in the pcrestack man
1595        page.
1597    13. Now that we have gone to the x.xx style of version numbers, the minor
1598        version may start with zero. Using 08 or 09 is a bad idea because users
1599        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1600        interpreted as invalid octal numbers. I've updated the previous comment in
1601        configure.ac, and also added a check that gives an error if 08 or 09 are
1602        used.
1604    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1605        causing partial matching to fail when the end of the subject matched \W
1606        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1608    15. There were some discrepancies between the declarations in pcre_internal.h
1609        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1610        their definitions. The declarations used "const uschar *" and the
1611        definitions used USPTR. Even though USPTR is normally defined as "const
1612        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1613        reported that: "This difference in casting confuses some C++ compilers, for
1614        example, SunCC recognizes above declarations as different functions and
1615        generates broken code for hbpcre." I have changed the declarations to use
1616        USPTR.
1618    16. GNU libtool is named differently on some systems. The autogen.sh script now
1619        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1620        (FreeBSD).
1622    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1623        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1624        comment: "Figure out how to create a longlong from a string: strtoll and
1625        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1626        instance, but it only takes 2 args instead of 3!"
1628    18. A subtle bug concerned with back references has been fixed by a change of
1629        specification, with a corresponding code fix. A pattern such as
1630        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1631        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1632        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1633        same bug. Such groups have to be quantified to be useful, or contained
1634        inside another quantified group. (If there's no repetition, the reference
1635        can never match.) The problem arises because, having left the group and
1636        moved on to the rest of the pattern, a later failure that backtracks into
1637        the group uses the captured value from the final iteration of the group
1638        rather than the correct earlier one. I have fixed this in PCRE by forcing
1639        any group that contains a reference to itself to be an atomic group; that
1640        is, there cannot be any backtracking into it once it has completed. This is
1641        similar to recursive and subroutine calls.
1644    Version 8.00 19-Oct-09
1645    ----------------------
1647    1.  The table for translating pcre_compile() error codes into POSIX error codes
1648        was out-of-date, and there was no check on the pcre_compile() error code
1649        being within the table. This could lead to an OK return being given in
1650        error.
1652    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1653        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1654        in a Windows environment.
1656    3.  The pcregrep --count option prints the count for each file even when it is
1657        zero, as does GNU grep. However, pcregrep was also printing all files when
1658        --files-with-matches was added. Now, when both options are given, it prints
1659        counts only for those files that have at least one match. (GNU grep just
1660        prints the file name in this circumstance, but including the count seems
1661        more useful - otherwise, why use --count?) Also ensured that the
1662        combination -clh just lists non-zero counts, with no names.
1664    4.  The long form of the pcregrep -F option was incorrectly implemented as
1665        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1666        but it seems right to fix it, and I didn't think it was worth preserving
1667        the old behaviour.
1669    5.  The command line items --regex=pattern and --regexp=pattern were not
1670        recognized by pcregrep, which required --regex pattern or --regexp pattern
1671        (with a space rather than an '='). The man page documented the '=' forms,
1672        which are compatible with GNU grep; these now work.
1674    6.  No libpcreposix.pc file was created for pkg-config; there was just
1675        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1677    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1678        when UCP support is not needed, by modifying the Python script that
1679        generates it from Unicode data files. This should not matter if the module
1680        is correctly used as a library, but I received one complaint about 50K of
1681        unwanted data. My guess is that the person linked everything into his
1682        program rather than using a library. Anyway, it does no harm.
1684    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1685        was a minimum greater than 1 for a wide character in a possessive
1686        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1687        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1688        character. Chaos in the form of incorrect output or a compiling loop could
1689        result.
1691    9.  The restrictions on what a pattern can contain when partial matching is
1692        requested for pcre_exec() have been removed. All patterns can now be
1693        partially matched by this function. In addition, if there are at least two
1694        slots in the offset vector, the offset of the earliest inspected character
1695        for the match and the offset of the end of the subject are set in them when
1696        PCRE_ERROR_PARTIAL is returned.
1698    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1699        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1700        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1701        and may be more useful for multi-segment matching.
1703    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1704        used to be given if ever the end of the subject was reached; now it is
1705        given only if matching could not proceed because another character was
1706        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1707        string "Z", which now yields "no match" instead of "partial match". In the
1708        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1709        final character ended with (*FAIL).
1711    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1712        if the pattern had a "must contain" character that was already found in the
1713        earlier partial match, unless partial matching was again requested. For
1714        example, with the pattern /dog.(body)?/, the "must contain" character is
1715        "g". If the first part-match was for the string "dog", restarting with
1716        "sbody" failed. This bug has been fixed.
1718    13. The string returned by pcre_dfa_exec() after a partial match has been
1719        changed so that it starts at the first inspected character rather than the
1720        first character of the match. This makes a difference only if the pattern
1721        starts with a lookbehind assertion or \b or \B (\K is not supported by
1722        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1723        matching functions compatible, and I think it's the right thing to do.
1725    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1726        so that the demonstration program is easily available in environments where
1727        PCRE has not been installed from source.
1729    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1730        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1731        library.
1733    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1734        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1735        is not the first non-POSIX option to be added. Clearly some people find
1736        these options useful.
1738    17. If a caller to the POSIX matching function regexec() passes a non-zero
1739        value for nmatch with a NULL value for pmatch, the value of
1740        nmatch is forced to zero.
1742    18. RunGrepTest did not have a test for the availability of the -u option of
1743        the diff command, as RunTest does. It now checks in the same way as
1744        RunTest, and also checks for the -b option.
1746    19. If an odd number of negated classes containing just a single character
1747        interposed, within parentheses, between a forward reference to a named
1748        subpattern and the definition of the subpattern, compilation crashed with
1749        an internal error, complaining that it could not find the referenced
1750        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1751        [The bug was that it was starting one character too far in when skipping
1752        over the character class, thus treating the ] as data rather than
1753        terminating the class. This meant it could skip too much.]
1755    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1756        /g option in pcretest when the pattern contains \K, which makes it possible
1757        to have an empty string match not at the start, even when the pattern is
1758        anchored. Updated pcretest and pcredemo to use this option.
1760    21. If the maximum number of capturing subpatterns in a recursion was greater
1761        than the maximum at the outer level, the higher number was returned, but
1762        with unset values at the outer level. The correct (outer level) value is
1763        now given.
1765    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1766        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1767        make it do so. The string so far is captured, making this feature
1768        compatible with Perl.
1770    23. The tests have been re-organized, adding tests 11 and 12, to make it
1771        possible to check the Perl 5.10 features against Perl 5.10.
1773    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1774        pattern matches a fixed length string. PCRE did not allow this; now it
1775        does. Neither allows recursion.
1777    25. I finally figured out how to implement a request to provide the minimum
1778        length of subject string that was needed in order to match a given pattern.
1779        (It was back references and recursion that I had previously got hung up
1780        on.) This code has now been added to pcre_study(); it finds a lower bound
1781        to the length of subject needed. It is not necessarily the greatest lower
1782        bound, but using it to avoid searching strings that are too short does give
1783        some useful speed-ups. The value is available to calling programs via
1784        pcre_fullinfo().
1786    26. While implementing 25, I discovered to my embarrassment that pcretest had
1787        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1788        study optimizations had never been tested with that matching function.
1789        Oops. What is worse, even when it was passed study data, there was a bug in
1790        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1791        were also very few tests of studied patterns with pcre_dfa_exec().
1793    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1794        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1795        on the other side of the coin, they are no longer allowed to have different
1796        names, because these cannot be distinguished in PCRE, and this has caused
1797        confusion. (This is a difference from Perl.)
1799    28. When duplicate subpattern names are present (necessarily with different
1800        numbers, as required by 27 above), and a test is made by name in a
1801        conditional pattern, either for a subpattern having been matched, or for
1802        recursion in such a pattern, all the associated numbered subpatterns are
1803        tested, and the overall condition is true if the condition is true for any
1804        one of them. This is the way Perl works, and is also more like the way
1805        testing by number works.
1808    Version 7.9 11-Apr-09
1809    ---------------------
1811    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1812        (pcretest), all targets were linked against these libraries. This included
1813        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1814        libraries. This caused unwanted dependencies to be created. This problem
1815        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1816        pcretest is linked with readline.
1818    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1819        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1820        moved outside it again, because FALSE and TRUE are already defined in AIX,
1821        but BOOL is not.
1823    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1824        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1826    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1827        hyphens) following file names and line numbers when outputting matching
1828        lines. This is not true; no spaces are inserted. I have also clarified the
1829        wording for the --colour (or --color) option.
1831    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1832        was not coloured; this is different to GNU grep, so I have changed it to be
1833        the same.
1835    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1836        each matching line was coloured. Now it goes on to look for further matches
1837        of any of the test patterns, which is the same behaviour as GNU grep.
1839    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1840        doesn't make sense to accept an empty string match in pcregrep, so I have
1841        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1842        seems to be how GNU grep behaves. [But see later change 40 for release
1843        8.33.]
1845    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1846        start or after a newline", because the conditional assertion was not being
1847        correctly handled. The rule now is that both the assertion and what follows
1848        in the first alternative must satisfy the test.
1850    9.  If auto-callout was enabled in a pattern with a conditional group whose
1851        condition was an assertion, PCRE could crash during matching, both with
1852        pcre_exec() and pcre_dfa_exec().
1854    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1855        used for matching.
1857    11. Unicode property support in character classes was not working for
1858        characters (bytes) greater than 127 when not in UTF-8 mode.
1860    12. Added the -M command line option to pcretest.
1862    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1864    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1866    16. Added comments and documentation about mis-use of no_arg in the C++
1867        wrapper.
1869    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1870        from Martin Jerabek that uses macro names for all relevant character and
1871        string constants.
1873    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1874        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1875        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1876        these, but not everybody uses configure.
1878    19. A conditional group that had only one branch was not being correctly
1879        recognized as an item that could match an empty string. This meant that an
1880        enclosing group might also not be so recognized, causing infinite looping
1881        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1882        with the subject "ab", where knowledge that the repeated group can match
1883        nothing is needed in order to break the loop.
1885    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1886        exec(), but without supplying a callout function, matching went wrong.
1888    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1889        leak if the size of the offset vector was greater than 30. When the vector
1890        is smaller, the saved offsets during recursion go onto a local stack
1891        vector, but for larger vectors malloc() is used. It was failing to free
1892        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1893        error, in fact).
1895    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1896        heapframe that is used only when UTF-8 support is enabled. This caused no
1897        problem, but was untidy.
1899    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1900        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1901        included within another project.
1903    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1904        slightly modified by me:
1906          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1907              not building pcregrep.
1909          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1910              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1912    25. Forward references, both numeric and by name, in patterns that made use of
1913        duplicate group numbers, could behave incorrectly or give incorrect errors,
1914        because when scanning forward to find the reference group, PCRE was not
1915        taking into account the duplicate group numbers. A pattern such as
1916        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1918    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1919        the feature of a custom pointer more persuasive (as requested by a user).
1921    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1922        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1923        pre-defined.
1925    28. Added support for (*UTF8) at the start of a pattern.
1927    29. Arrange for flags added by the "release type" setting in CMake to be shown
1928        in the configuration summary.
1931    Version 7.8 05-Sep-08
1932    ---------------------
1934    1.  Replaced UCP searching code with optimized version as implemented for Ad
1935        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1936        stage table and inline lookup instead of a function, giving speed ups of 2
1937        to 5 times on some simple patterns that I tested. Permission was given to
1938        distribute the MultiStage2.py script that generates the tables (it's not in
1939        the tarball, but is in the Subversion repository).
1941    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1942        scripts.
1944    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1945        a group with a zero qualifier. The result of the study could be incorrect,
1946        or the function might crash, depending on the pattern.
1948    4.  Caseless matching was not working for non-ASCII characters in back
1949        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1950        It now works when Unicode Property Support is available.
1952    5.  In pcretest, an escape such as \x{de} in the data was always generating
1953        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1954        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1955        truncation.
1957    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1959    7.  Added two (int) casts to pcregrep when printing the difference of two
1960        pointers, in case they are 64-bit values.
1962    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1963        test 2 if it fails.
1965    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1966        and a #define of that name to empty if it is not externally set. This is to
1967        allow users of MSVC to set it if necessary.
1969    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1970        the convenience functions in the pcre_get.c source file.
1972    11. An option change at the start of a pattern that had top-level alternatives
1973        could cause overwriting and/or a crash. This command provoked a crash in
1974        some environments:
1976          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1978        This potential security problem was recorded as CVE-2008-2371.
1980    12. For a pattern where the match had to start at the beginning or immediately
1981        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1982        pcre_dfa_exec() could read past the end of the passed subject if there was
1983        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1984        pcretest so that it places the subject at the end of its malloc-ed buffer.
1986    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1987        exec() might read past the end of the data buffer in UTF-8 mode.
1989    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1990        the data contained the byte 0x85 as part of a UTF-8 character within its
1991        first line. This applied both to normal and DFA matching.
1993    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1994        /^[^d]*?$/8 failed to match "abc".
1996    16. Added a missing copyright notice to pcrecpp_internal.h.
1998    17. Make it more clear in the documentation that values returned from
1999        pcre_exec() in ovector are byte offsets, not character counts.
2001    18. Tidied a few places to stop certain compilers from issuing warnings.
2003    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2004        supplied by Stefan Weber. I made a further small update for 7.8 because
2005        there is a change of source arrangements: the pcre_searchfuncs.c module is
2006        replaced by pcre_ucd.c.
2009    Version 7.7 07-May-08
2010    ---------------------
2012    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2013        a string to a long long, pretend we don't even have a long long." This is
2014        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2016    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2017        pre-7.6 versions, which defined a global no_arg variable instead of putting
2018        it in the RE class. (See also #8 below.)
2020    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2021        Lopes.
2023    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2025        (1) The include/exclude patterns were being applied to the whole pathnames
2026            of files, instead of just to the final components.
2028        (2) If there was more than one level of directory, the subdirectories were
2029            skipped unless they satisfied the include/exclude conditions. This is
2030            inconsistent with GNU grep (and could even be seen as contrary to the
2031            pcregrep specification - which I improved to make it absolutely clear).
2032            The action now is always to scan all levels of directory, and just
2033            apply the include/exclude patterns to regular files.
2035    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2036        --exclude_dir in the tests to avoid scanning .svn directories.
2038    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2039        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2040        doesn't support NULs in patterns.
2042    7.  Added some missing "const"s to declarations of static tables in
2043        pcre_compile.c and pcre_dfa_exec.c.
2045    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2046        caused by fix #2  above. (Subsequently also a second patch to fix the
2047        first patch. And a third patch - this was a messy problem.)
2049    9.  Applied Craig's patch to remove the use of push_back().
2051    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2052        matching function regexec().
2054    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2055        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2056        references. PCRE supports relative numbers with this syntax (I don't think
2057        Oniguruma does).
2059    12. Previously, a group with a zero repeat such as (...){0} was completely
2060        omitted from the compiled regex. However, this means that if the group
2061        was called as a subroutine from elsewhere in the pattern, things went wrong
2062        (an internal error was given). Such groups are now left in the compiled
2063        pattern, with a new opcode that causes them to be skipped at execution
2064        time.
2066    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2067        to the way PCRE behaves:
2069        (a) A lone ] character is dis-allowed (Perl treats it as data).
2071        (b) A back reference to an unmatched subpattern matches an empty string
2072            (Perl fails the current match path).
2074        (c) A data ] in a character class must be notated as \] because if the
2075            first data character in a class is ], it defines an empty class. (In
2076            Perl it is not possible to have an empty class.) The empty class []
2077            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2078            The negative empty class [^] matches any one character, independently
2079            of the DOTALL setting.
2081    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2082        non-existent subpattern following a character class starting with ']' and
2083        containing () gave an internal compiling error instead of "reference to
2084        non-existent subpattern". Fortunately, when the pattern did exist, the
2085        compiled code was correct. (When scanning forwards to check for the
2086        existence of the subpattern, it was treating the data ']' as terminating
2087        the class, so got the count wrong. When actually compiling, the reference
2088        was subsequently set up correctly.)
2090    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2091        it was being rejected as not supported by pcre_dfa_exec(), even though
2092        other assertions are supported. I have made pcre_dfa_exec() support
2093        (*FAIL).
2095    16. The implementation of 13c above involved the invention of a new opcode,
2096        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2097        cannot be changed at match time, I realized I could make a small
2098        improvement to matching performance by compiling OP_ALLANY instead of
2099        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2100        on the OP_ANY path.
2102    17. Compiling pcretest on Windows with readline support failed without the
2103        following two fixes: (1) Make the unistd.h include conditional on
2104        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2106    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2107        ncurses library to be included for pcretest when ReadLine support is
2108        requested, but also to allow for it to be overridden. This patch came from
2109        Daniel Bergström.
2111    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2112        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2113        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2114        spotting this.
2117    Version 7.6 28-Jan-08
2118    ---------------------
2120    1.  A character class containing a very large number of characters with
2121        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2122        overflow.
2124    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2125        HAVE_LONG_LONG is not defined.
2127    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2128        bring them up to date and include new features. This patch includes:
2130        - Fixed PH's badly added libz and libbz2 support.
2131        - Fixed a problem with static linking.
2132        - Added pcredemo. [But later removed - see 7 below.]
2133        - Fixed dftables problem and added an option.
2134        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2135            HAVE_LONG_LONG.
2136        - Added readline support for pcretest.
2137        - Added an listing of the option settings after cmake has run.
2139    4.  A user submitted a patch to Makefile that makes it easy to create
2140        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2141        Makefile.am that cause it to include this special target, without
2142        affecting anything else. Note that the same mingw target plus all
2143        the other distribution libraries and programs are now supported
2144        when configuring with CMake (see 6 below) instead of with
2145        Configure/Make.
2147    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2148        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2149        exported in the Windows port". It has not yet been confirmed that the patch
2150        solves the problem, but it does no harm.
2152    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2153        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2154        with CMake, and also correct the comment about stack recursion.
2156    7.  Remove the automatic building of pcredemo from the ./configure system and
2157        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2158        of a program that users should build themselves after PCRE is installed, so
2159        building it automatically is not really right. What is more, it gave
2160        trouble in some build environments.
2162    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2165    Version 7.5 10-Jan-08
2166    ---------------------
2168    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2169        values in parens when parsing an RE using the C++ wrapper."
2171    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2172        Characters greater than 255 were excluded from the class instead of being
2173        included.
2175    3.  The same bug as (2) above applied to negated POSIX classes such as
2176        [:^space:].
2178    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2179        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2180        I have changed it.
2182    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2183        first named subpattern) and a construct such as (?&a) would reference the
2184        first named subpattern whose name started with "a" (in other words, the
2185        length check was missing). Both these problems are fixed. "Subpattern name
2186        expected" is now given for (?&) (a zero-length name), and this patch also
2187        makes it give the same error for \k'' (previously it complained that that
2188        was a reference to a non-existent subpattern).
2190    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2191        this is right because (?- can be followed by option settings as well as by
2192        digits. I have, however, made the messages clearer.
2194    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2195        than the number used in the conditional) now cause a compile-time error.
2196        This is actually not compatible with Perl, which accepts such patterns, but
2197        treats the conditional as always being FALSE (as PCRE used to), but it
2198        seems to me that giving a diagnostic is better.
2200    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2201        and messages.
2203    9.  Fix two occurrences of "backslash" in comments that should have been
2204        "backspace".
2206    10. Remove two redundant lines of code that can never be obeyed (their function
2207        was moved elsewhere).
2209    11. The program that makes PCRE's Unicode character property table had a bug
2210        which caused it to generate incorrect table entries for sequences of
2211        characters that have the same character type, but are in different scripts.
2212        It amalgamated them into a single range, with the script of the first of
2213        them. In other words, some characters were in the wrong script. There were
2214        thirteen such cases, affecting characters in the following ranges:
2216          U+002b0 - U+002c1
2217          U+0060c - U+0060d
2218          U+0061e - U+00612
2219          U+0064b - U+0065e
2220          U+0074d - U+0076d
2221          U+01800 - U+01805
2222          U+01d00 - U+01d77
2223          U+01d9b - U+01dbf
2224          U+0200b - U+0200f
2225          U+030fc - U+030fe
2226          U+03260 - U+0327f
2227          U+0fb46 - U+0fbb1
2228          U+10450 - U+1049d
2230    12. The -o option (show only the matching part of a line) for pcregrep was not
2231        compatible with GNU grep in that, if there was more than one match in a
2232        line, it showed only the first of them. It now behaves in the same way as
2233        GNU grep.
2235    13. If the -o and -v options were combined for pcregrep, it printed a blank
2236        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2237        does the same. The return code can be used to tell if there were any
2238        non-matching lines.
2240    14. Added --file-offsets and --line-offsets to pcregrep.
2242    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2243        infinitely looping recursion. The bug was that positive lookaheads were not
2244        being skipped when checking for a possible empty match (negative lookaheads
2245        and both kinds of lookbehind were skipped).
2247    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2248        inclusion of <windows.h> to before rather than after the definition of
2249        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2251    17. Specifying a possessive quantifier with a specific limit for a Unicode
2252        character property caused pcre_compile() to compile bad code, which led at
2253        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2254        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2255        caused the error; without that there was no problem.
2257    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2259    19. Added --enable-pcretest-libreadline.
2261    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2262        RE::GlobalReplace(). As a result, the number of replacements returned was
2263        double what it should be. I removed one of the increments, but Craig sent a
2264        later patch that removed the other one (the right fix) and added unit tests
2265        that check the return values (which was not done before).
2267    21. Several CMake things:
2269        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2270            the names libpcre and libpcreposix, not just pcre and pcreposix.
2272        (2) The above change means that pcretest and pcregrep are now correctly
2273            linked with the newly-built libraries, not previously installed ones.
2277    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2278        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2279        UTF-8 newline character). The key issue is that the pattern starts .*;
2280        this means that the match must be either at the beginning, or after a
2281        newline. The bug was in the code for advancing after a failed match and
2282        checking that the new position followed a newline. It was not taking
2283        account of UTF-8 characters correctly.
2285    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2286        character classes. PCRE was not treating the sequence [:...:] as a
2287        character class unless the ... were all letters. Perl, however, seems to
2288        allow any characters between [: and :], though of course it rejects as
2289        unknown any "names" that contain non-letters, because all the known class
2290        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2291        for example, whereas PCRE did not - it did not recognize a POSIX character
2292        class. This seemed a bit dangerous, so the code has been changed to be
2293        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2294        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2295        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2296        Perl does, and where it didn't before.
2298    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2299        Windows environments %n is disabled by default.
2302    Version 7.4 21-Sep-07
2303    ---------------------
2305    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2306        means that a class such as [\s] counted as "explicit reference to CR or
2307        LF". That isn't really right - the whole point of the change was to try to
2308        help when there was an actual mention of one of the two characters. So now
2309        the change happens only if \r or \n (or a literal CR or LF) character is
2310        encountered.
2312    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2313        of both had grown to the point where there were only 3 bits left.
2314        Fortunately, there was spare space in the data structure, and so I have
2315        moved the internal flags into a new 16-bit field to free up more option
2316        bits.
2318    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2319        but did not set the internal JCHANGED flag - either of these is enough to
2320        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2321        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2322        start sets both bits.
2324    4.  Added options (at build time, compile time, exec time) to change \R from
2325        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2327    5.  doc/pcresyntax.html was missing from the distribution.
2329    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2330        compatibility, even though it is no longer used.
2332    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2333        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2334        windows.h file is present (where different names are used). [This was
2335        reversed later after testing - see 16 below.]
2337    8.  Changed all #include <config.h> to #include "config.h". There were also
2338        some further <pcre.h> cases that I changed to "pcre.h".
2340    9.  When pcregrep was used with the --colour option, it missed the line ending
2341        sequence off the lines that it output.
2343    10. It was pointed out to me that arrays of string pointers cause lots of
2344        relocations when a shared library is dynamically loaded. A technique of
2345        using a single long string with a table of offsets can drastically reduce
2346        these. I have refactored PCRE in four places to do this. The result is
2347        dramatic:
2349          Originally:                          290
2350          After changing UCP table:            187
2351          After changing error message table:   43
2352          After changing table of "verbs"       36
2353          After changing table of Posix names   22
2355        Thanks to the folks working on Gregex for glib for this insight.
2357    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2358        unicode-properties was also set.
2360    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2362    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2363        checked only for CRLF.
2365    14. Added casts to pcretest.c to avoid compiler warnings.
2367    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2369    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2370        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2371        entirely. This removes changes made in 7 above.
2373    17. The CMake files have been updated, and there is now more information about
2374        building with CMake in the NON-UNIX-USE document.
2377    Version 7.3 28-Aug-07
2378    ---------------------
2380     1. In the rejigging of the build system that eventually resulted in 7.1, the
2381        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2382        brackets there is not right, since it causes compilers to look for an
2383        installed pcre.h, not the version that is in the source that is being
2384        compiled (which of course may be different). I have changed it back to:
2386          #include "pcre.h"
2388        I have a vague recollection that the change was concerned with compiling in
2389        different directories, but in the new build system, that is taken care of
2390        by the VPATH setting the Makefile.
2392     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2393        when the subject happened to end in the byte 0x85 (e.g. if the last
2394        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2395        characters but of course it shouldn't be taken as a newline when it is part
2396        of another character. The bug was that, for an unlimited repeat of . in
2397        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2398        characters when looking for a newline.
2400     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2402     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2403        in debug output.
2405     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2406        long printing in the pcrecpp unittest when running under MinGW.
2408     6. ESC_K was left out of the EBCDIC table.
2410     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2411        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2412        limit also applies to "virtual nesting" when a pattern is recursive, and in
2413        this case 1000 isn't so big. I have been able to remove this limit at the
2414        expense of backing off one optimization in certain circumstances. Normally,
2415        when pcre_exec() would call its internal match() function recursively and
2416        immediately return the result unconditionally, it uses a "tail recursion"
2417        feature to save stack. However, when a subpattern that can match an empty
2418        string has an unlimited repetition quantifier, it no longer makes this
2419        optimization. That gives it a stack frame in which to save the data for
2420        checking that an empty string has been matched. Previously this was taken
2421        from the 1000-entry workspace that had been reserved. So now there is no
2422        explicit limit, but more stack is used.
2424     8. Applied Daniel's patches to solve problems with the import/export magic
2425        syntax that is required for Windows, and which was going wrong for the
2426        pcreposix and pcrecpp parts of the library. These were overlooked when this
2427        problem was solved for the main library.
2429     9. There were some crude static tests to avoid integer overflow when computing
2430        the size of patterns that contain repeated groups with explicit upper
2431        limits. As the maximum quantifier is 65535, the maximum group length was
2432        set at 30,000 so that the product of these two numbers did not overflow a
2433        32-bit integer. However, it turns out that people want to use groups that
2434        are longer than 30,000 bytes (though not repeat them that many times).
2435        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2436        made it possible to implement the integer overflow checks in a much more
2437        dynamic way, which I have now done. The artificial limitation on group
2438        length has been removed - we now have only the limit on the total length of
2439        the compiled pattern, which depends on the LINK_SIZE setting.
2441    10. Fixed a bug in the documentation for get/copy named substring when
2442        duplicate names are permitted. If none of the named substrings are set, the
2443        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2444        empty string.
2446    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2447        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2448        because the ] is interpreted as the first data character and the
2449        terminating ] is not found. PCRE has been made compatible with Perl in this
2450        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2451        cause memory overwriting.
2453    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2454        string has been matched (to stop an infinite loop). It was not recognizing
2455        a conditional subpattern that could match an empty string if that
2456        subpattern was within another subpattern. For example, it looped when
2457        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2458        condition was not nested. This bug has been fixed.
2460    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2461        past the start of the subject in the presence of bytes with the top bit
2462        set, for example "\x8aBCD".
2464    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2465        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2467    14. Optimized (?!) to (*FAIL).
2469    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2470        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2471        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2472        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2473        does: it's just the validity check that is more restrictive.
2475    16. Inserted checks for integer overflows during escape sequence (backslash)
2476        processing, and also fixed erroneous offset values for syntax errors during
2477        backslash processing.
2479    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2480        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2482    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2483        caused an overrun.
2485    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2486        something other than just ASCII characters) inside a group that had an
2487        unlimited repeat caused a loop at compile time (while checking to see
2488        whether the group could match an empty string).
2490    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2491        [\P{Any}] did so. (Error in the code for printing property names.)
2493    21. An orphan \E inside a character class could cause a crash.
2495    22. A repeated capturing bracket such as (A)? could cause a wild memory
2496        reference during compilation.
2498    23. There are several functions in pcre_compile() that scan along a compiled
2499        expression for various reasons (e.g. to see if it's fixed length for look
2500        behind). There were bugs in these functions when a repeated \p or \P was
2501        present in the pattern. These operators have additional parameters compared
2502        with \d, etc, and these were not being taken into account when moving along
2503        the compiled data. Specifically:
2505        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2506            length.
2508        (b) An item such as \pL+ within a repeated group could cause crashes or
2509            loops.
2511        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2512            "reference to non-existent subpattern" error.
2514        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2516    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2517        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2519    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2521    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2522        character were causing crashes (broken optimization).
2524    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2525        \p or \P) caused a compile-time loop.
2527    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2528        break. For example, the unstudied pattern [\r\n]A does not match the string
2529        "\r\nA" because change 7.0/46 below moves the current point on by two
2530        characters after failing to match at the start. However, the pattern \nA
2531        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2532        the same is true. There doesn't seem any very clean way out of this, but
2533        what I have chosen to do makes the common cases work: PCRE now takes note
2534        of whether there can be an explicit match for \r or \n anywhere in the
2535        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2536        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2537        pattern has explicit CR or LF references.
2539    29. Added (*CR) etc for changing newline setting at start of pattern.
2542    Version 7.2 19-Jun-07
2543    ---------------------
2545     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2546        which is apparently normally available under Windows.
2548     2. Re-jig the pcregrep tests with different newline settings in an attempt
2549        to make them independent of the local environment's newline setting.
2551     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2553     4. Some of the "internals" tests were previously cut out when the link size
2554        was not 2, because the output contained actual offsets. The recent new
2555        "Z" feature of pcretest means that these can be cut out, making the tests
2556        usable with all link sizes.
2558     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2559        stack recursion. This gives a massive performance boost under BSD, but just
2560        a small improvement under Linux. However, it saves one field in the frame
2561        in all cases.
2563     6. Added more features from the forthcoming Perl 5.10:
2565        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2566            recursion call. It refers to the nth most recently opened parentheses.
2568        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2569            to be opened parentheses.
2571        (c) Conditions that refer to capturing parentheses can be specified
2572            relatively, for example, (?(-2)... or (?(+3)...
2574        (d) \K resets the start of the current match so that everything before
2575            is not part of it.
2577        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2579        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2580            reference syntax.
2582        (g) (?| introduces a group in which the numbering of parentheses in each
2583            alternative starts with the same number.
2585        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2587     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2590     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2591        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2592        for detecting groups that can match an empty string.
2594     9. A pattern with a very large number of alternatives (more than several
2595        hundred) was running out of internal workspace during the pre-compile
2596        phase, where pcre_compile() figures out how much memory will be needed. A
2597        bit of new cunning has reduced the workspace needed for groups with
2598        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2599        workspace instead of running out of the 4096 that are available.
2601    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2603    11. Applied patch from Google to remove an optimization that didn't quite work.
2604        The report of the bug said:
2606          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2607          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2608          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2610    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2611        it matched the wrong number of bytes.
2614    Version 7.1 24-Apr-07
2615    ---------------------
2617     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2618        that is more "standard", making use of automake and other Autotools. There
2619        is some re-arrangement of the files and adjustment of comments consequent
2620        on this.
2622     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2623        for recursive directory scanning broke on some systems because the files
2624        are not scanned in any specific order and on different systems the order
2625        was different. A call to "sort" has been inserted into RunGrepTest for the
2626        approprate test as a short-term fix. In the longer term there may be an
2627        alternative.
2629     3. I had an email from Eric Raymond about problems translating some of PCRE's
2630        man pages to HTML (despite the fact that I distribute HTML pages, some
2631        people do their own conversions for various reasons). The problems
2632        concerned the use of low-level troff macros .br and .in. I have therefore
2633        removed all such uses from the man pages (some were redundant, some could
2634        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2635        HTML has been updated to handle .nf/.fi and to complain if it encounters
2636        .br or .in.
2638     4. Updated comments in configure.ac that get placed in config.h.in and also
2639        arranged for config.h to be included in the distribution, with the name
2640        config.h.generic, for the benefit of those who have to compile without
2641        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2643     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2644        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2645        makevp.bat for the current PCRE, using the additional files
2646        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2648     6. A Windows user reported a minor discrepancy with test 2, which turned out
2649        to be caused by a trailing space on an input line that had got lost in his
2650        copy. The trailing space was an accident, so I've just removed it.
2652     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2653        that is needed.
2655     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2656        as "const" (a) because they are and (b) because it helps the PHP
2657        maintainers who have recently made a script to detect big data structures
2658        in the php code that should be moved to the .rodata section. I remembered
2659        to update Builducptable as well, so it won't revert if ucptable.h is ever
2660        re-created.
2662     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2663        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2664        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2665        support is not required. This saves 1.5-2K of code, which is important in
2666        some applications.
2668        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2669        so as not to refer to the tables, even though these functions will never be
2670        called when UTF-8 support is disabled. Otherwise there are problems with a
2671        shared library.
2673    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2675        (a) It was defining its arguments as char * instead of void *.
2677        (b) It was assuming that all moves were upwards in memory; this was true
2678            a long time ago when I wrote it, but is no longer the case.
2680        The emulated memove() is provided for those environments that have neither
2681        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2682        is clearly not the case, as these two bugs were recently reported.
2684    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2685        and Detrail to create the HTML documentation, the .txt form of the man
2686        pages, and it removes trailing spaces from listed files. It also creates
2687        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2688        case, it wraps all the #defines with #ifndefs. This script should be run
2689        before "make dist".
2691    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2692        with Unicode property support.
2694        (a) For a maximizing quantifier, if the two different cases of the
2695            character were of different lengths in their UTF-8 codings (there are
2696            some cases like this - I found 11), and the matching function had to
2697            back up over a mixture of the two cases, it incorrectly assumed they
2698            were both the same length.
2700        (b) When PCRE was configured to use the heap rather than the stack for
2701            recursion during matching, it was not correctly preserving the data for
2702            the other case of a UTF-8 character when checking ahead for a match
2703            while processing a minimizing repeat. If the check also involved
2704            matching a wide character, but failed, corruption could cause an
2705            erroneous result when trying to check for a repeat of the original
2706            character.
2708    13. Some tidying changes to the testing mechanism:
2710        (a) The RunTest script now detects the internal link size and whether there
2711            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2712            values substituted by "configure". (The RunGrepTest script already did
2713            this for UTF-8.) The configure.ac script no longer substitutes the
2714            relevant variables.
2716        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2717            with length and offset values. This means that the output is different
2718            for different internal link sizes. Test 2 is skipped for link sizes
2719            other than 2 because of this, bypassing the problem. Unfortunately,
2720            there was also a test in test 3 (the locale tests) that used /B and
2721            failed for link sizes other than 2. Rather than cut the whole test out,
2722            I have added a new /Z option to pcretest that replaces the length and
2723            offset values with spaces. This is now used to make test 3 independent
2724            of link size. (Test 2 will be tidied up later.)
2726    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2727        segmentation fault instead of returning the appropriate error message.
2729    15. In multiline mode when the newline sequence was set to "any", the pattern
2730        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2731        This doesn't seem right; it now treats the CRLF combination as the line
2732        ending, and so does not match in that case. It's only a pattern such as ^$
2733        that would hit this one: something like ^ABC$ would have failed after \r
2734        and then tried again after \r\n.
2736    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2737        in an attempt to make files that differ only in their line terminators
2738        compare equal. This works on Linux.
2740    17. Under certain error circumstances pcregrep might try to free random memory
2741        as it exited. This is now fixed, thanks to valgrind.
2743    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2744        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2745        was because its rules for how to advance for /g after matching an empty
2746        string at the end of a line did not allow for this case. They now check for
2747        it specially.
2749    20. pcretest is supposed to handle patterns and data of any length, by
2750        extending its buffers when necessary. It was getting this wrong when the
2751        buffer for a data line had to be extended.
2753    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2754        CRLF as a newline sequence.
2756    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2757        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2758        I have nevertheless tidied it up.
2760    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2762    24. Added a man page for pcre-config.
2765    Version 7.0 19-Dec-06
2766    ---------------------
2768     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2769        moving to gcc 4.1.1.
2771     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2772        sys/time.h, which is documented as needed for this function. It doesn't
2773        seem to matter on Linux, but it showed up on some releases of OS X.
2775     3. It seems that there are systems where bytes whose values are greater than
2776        127 match isprint() in the "C" locale. The "C" locale should be the
2777        default when a C program starts up. In most systems, only ASCII printing
2778        characters match isprint(). This difference caused the output from pcretest
2779        to vary, making some of the tests fail. I have changed pcretest so that:
2781        (a) When it is outputting text in the compiled version of a pattern, bytes
2782            other than 32-126 are always shown as hex escapes.
2784        (b) When it is outputting text that is a matched part of a subject string,
2785            it does the same, unless a different locale has been set for the match
2786            (using the /L modifier). In this case, it uses isprint() to decide.
2788     4. Fixed a major bug that caused incorrect computation of the amount of memory
2789        required for a compiled pattern when options that changed within the
2790        pattern affected the logic of the preliminary scan that determines the
2791        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2792        that the computed length was too small. The symptoms of this bug were
2793        either the PCRE error "internal error: code overflow" from pcre_compile(),
2794        or a glibc crash with a message such as "pcretest: free(): invalid next
2795        size (fast)". Examples of patterns that provoked this bug (shown in
2796        pcretest format) are:
2798          /(?-x: )/x
2799          /(?x)(?-x: \s*#\s*)/
2800          /((?i)[\x{c0}])/8
2801          /(?i:[\x{c0}])/8
2803        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2804        is now done differently.
2806     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2807        wrapper classes; (b) implement a new function in the C++ scanner that is
2808        more efficient than the old way of doing things because it avoids levels of
2809        recursion in the regex matching; (c) add a paragraph to the documentation
2810        for the FullMatch() function.
2812     6. The escape sequence \n was being treated as whatever was defined as
2813        "newline". Not only was this contrary to the documentation, which states
2814        that \n is character 10 (hex 0A), but it also went horribly wrong when
2815        "newline" was defined as CRLF. This has been fixed.
2817     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2818        was being set to -1 for the "end of line" case (supposedly a value that no
2819        character can have). Though this value is never used (the check for end of
2820        line is "zero bytes in current character"), it caused compiler complaints.
2821        I've changed it to 0xffffffff.
2823     8. In pcre_version.c, the version string was being built by a sequence of
2824        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2825        string (as it is for production releases) called a macro with an empty
2826        argument. The C standard says the result of this is undefined. The gcc
2827        compiler treats it as an empty string (which was what was wanted) but it is
2828        reported that Visual C gives an error. The source has been hacked around to
2829        avoid this problem.
2831     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2832        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2833        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2834        of them did).
2836    10. Originally, pcretest opened its input and output without "b"; then I was
2837        told that "b" was needed in some environments, so it was added for release
2838        5.0 to both the input and output. (It makes no difference on Unix-like
2839        systems.) Later I was told that it is wrong for the input on Windows. I've
2840        now abstracted the modes into two macros, to make it easier to fiddle with
2841        them, and removed "b" from the input mode under Windows.
2843    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2845    12. Added -help and --help to pcretest as an official way of being reminded
2846        of the options.
2848    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2849        and pcrecpp.cc because they annoy compilers at high warning levels.
2851    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2853    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2854        scripts are not C programs :-) and which was not noticed because it works
2855        on Linux.
2857    16. pcretest is supposed to handle any length of pattern and data line (as one
2858        line or as a continued sequence of lines) by extending its input buffer if
2859        necessary. This feature was broken for very long pattern lines, leading to
2860        a string of junk being passed to pcre_compile() if the pattern was longer
2861        than about 50K.
2863    17. I have done a major re-factoring of the way pcre_compile() computes the
2864        amount of memory needed for a compiled pattern. Previously, there was code
2865        that made a preliminary scan of the pattern in order to do this. That was
2866        OK when PCRE was new, but as the facilities have expanded, it has become
2867        harder and harder to keep it in step with the real compile phase, and there
2868        have been a number of bugs (see for example, 4 above). I have now found a
2869        cunning way of running the real compile function in a "fake" mode that
2870        enables it to compute how much memory it would need, while actually only
2871        ever using a few hundred bytes of working memory and without too many
2872        tests of the mode. This should make future maintenance and development
2873        easier. A side effect of this work is that the limit of 200 on the nesting
2874        depth of parentheses has been removed (though this was never a serious
2875        limitation, I suspect). However, there is a downside: pcre_compile() now
2876        runs more slowly than before (30% or more, depending on the pattern). I
2877        hope this isn't a big issue. There is no effect on runtime performance.
2879    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2880        newline (only possible for the last line of a file) and it was a
2881        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2883    19. Added additional timing features to pcretest. (1) The -tm option now times
2884        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2885        separate command line item, by a number that specifies the number of
2886        repeats to use when timing. The default is 50000; this gives better
2887        precision, but takes uncomfortably long for very large patterns.
2889    20. Extended pcre_study() to be more clever in cases where a branch of a
2890        subpattern has no definite first character. For example, (a*|b*)[cd] would
2891        previously give no result from pcre_study(). Now it recognizes that the
2892        first character must be a, b, c, or d.
2894    21. There was an incorrect error "recursive call could loop indefinitely" if
2895        a subpattern (or the entire pattern) that was being tested for matching an
2896        empty string contained only one non-empty item after a nested subpattern.
2897        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2898        incorrectly, because the \d was being skipped in the check.
2900    22. The pcretest program now has a new pattern option /B and a command line
2901        option -b, which is equivalent to adding /B to every pattern. This causes
2902        it to show the compiled bytecode, without the additional information that
2903        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2904        is the same as /B/I).
2906    23. A new optimization is now able automatically to treat some sequences such
2907        as a*b as a*+b. More specifically, if something simple (such as a character
2908        or a simple class like \d) has an unlimited quantifier, and is followed by
2909        something that cannot possibly match the quantified thing, the quantifier
2910        is automatically "possessified".
2912    24. A recursive reference to a subpattern whose number was greater than 39
2913        went wrong under certain circumstances in UTF-8 mode. This bug could also
2914        have affected the operation of pcre_study().
2916    25. Realized that a little bit of performance could be had by replacing
2917        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2919    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2921    27. Possessive quantifiers such as a++ were previously implemented by turning
2922        them into atomic groups such as ($>a+). Now they have their own opcodes,
2923        which improves performance. This includes the automatically created ones
2924        from 23 above.
2926    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2927        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2928        the first matched character to be a colon. This applied both to named and
2929        numbered groups.
2931    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2933    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2934        building PCRE on Windows easier, so I have included it in the distribution.
2936    31. There is now a check in pcretest against a ridiculously large number being
2937        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2938        loop, the loop is abandoned.
2940    32. Forward references to subpatterns in conditions such as (?(2)...) where
2941        subpattern 2 is defined later cause pcre_compile() to search forwards in
2942        the pattern for the relevant set of parentheses. This search went wrong
2943        when there were unescaped parentheses in a character class, parentheses
2944        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2946    33. "Subroutine" calls and backreferences were previously restricted to
2947        referencing subpatterns earlier in the regex. This restriction has now
2948        been removed.
2950    34. Added a number of extra features that are going to be in Perl 5.10. On the
2951        whole, these are just syntactic alternatives for features that PCRE had
2952        previously implemented using the Python syntax or my own invention. The
2953        other formats are all retained for compatibility.
2955        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2956            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2957            also .NET compatible.
2959        (b) A recursion or subroutine call to a named group can now be defined as
2960            (?&name) as well as (?P>name).
2962        (c) A backreference to a named group can now be defined as \k<name> or
2963            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2964            5.10, are also .NET compatible.
2966        (d) A conditional reference to a named group can now use the syntax
2967            (?(<name>) or (?('name') as well as (?(name).
2969        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2970            groups (named and numbered) that are never evaluated inline, but can be
2971            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2972            is always false. There may be only one alternative in such a group.
2974        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2975            as the simple (?(R). The condition is true only if the most recent
2976            recursion is that of the given number or name. It does not search out
2977            through the entire recursion stack.
2979        (g) The escape \gN or \g{N} has been added, where N is a positive or
2980            negative number, specifying an absolute or relative reference.
2982    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2983        some "unreachable code" warnings.
2985    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2986        things, this adds five new scripts.
2988    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2989        There were also incompatibilities regarding the handling of \Q..\E inside
2990        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2991        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2993    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2994        matches an empty string, and forcibly breaks the loop. There were bugs in
2995        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2996        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2997        separate and independent bugs (that affected different cases) have been
2998        fixed.
3000    39. Refactored the code to abolish the use of different opcodes for small
3001        capturing bracket numbers. This is a tidy that I avoided doing when I
3002        removed the limit on the number of capturing brackets for 3.5 back in 2001.
3003        The new approach is not only tidier, it makes it possible to reduce the
3004        memory needed to fix the previous bug (38).
3006    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
3007        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
3008        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
3009        mode.
3011    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
3012        report.
3014    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
3015        copy construction and assignment in the C++ wrapper.
3017    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
3018        couple of bugs that could have given wrong results in the "--newline=crlf"
3019        case.
3021    44. Added a number of casts and did some reorganization of signed/unsigned int
3022        variables following suggestions from Dair Grant. Also renamed the variable
3023        "this" as "item" because it is a C++ keyword.
3025    45. Arranged for dftables to add
3027          #include "pcre_internal.h"
3029        to pcre_chartables.c because without it, gcc 4.x may remove the array
3030        definition from the final binary if PCRE is built into a static library and
3031        dead code stripping is activated.
3033    46. For an unanchored pattern, if a match attempt fails at the start of a
3034        newline sequence, and the newline setting is CRLF or ANY, and the next two
3035        characters are CRLF, advance by two characters instead of one.
3038    Version 6.7 04-Jul-06
3039    ---------------------
3041     1. In order to handle tests when input lines are enormously long, pcretest has
3042        been re-factored so that it automatically extends its buffers when
3043        necessary. The code is crude, but this _is_ just a test program. The
3044        default size has been increased from 32K to 50K.
3046     2. The code in pcre_study() was using the value of the re argument before
3047        testing it for NULL. (Of course, in any sensible call of the function, it
3048        won't be NULL.)
3050     3. The memmove() emulation function in pcre_internal.h, which is used on
3051        systems that lack both memmove() and bcopy() - that is, hardly ever -
3052        was missing a "static" storage class specifier.
3054     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
3055        containing an extended class (one that cannot be represented by a bitmap
3056        because it contains high-valued characters or Unicode property items, e.g.
3057        [\pZ]). Almost always one would set UTF-8 mode when processing such a
3058        pattern, but PCRE should not loop if you do not (it no longer does).
3059        [Detail: two cases were found: (a) a repeated subpattern containing an
3060        extended class; (b) a recursive reference to a subpattern that followed a
3061        previous extended class. It wasn't skipping over the extended class
3062        correctly when UTF-8 mode was not set.]
3064     5. A negated single-character class was not being recognized as fixed-length
3065        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
3066        compile error "lookbehind assertion is not fixed length".
3068     6. The RunPerlTest auxiliary script was showing an unexpected difference
3069        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
3070        write a Perl script that can interpret lines of an input file either as
3071        byte characters or as UTF-8, which is what "perltest" was being required to
3072        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
3073        can't do is switch easily at run time between having the "use utf8;" pragma
3074        or not. In the end, I fudged it by using the RunPerlTest script to insert
3075        "use utf8;" explicitly for the UTF-8 tests.
3077     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
3078        the end of the subject string, contrary to the documentation and to what
3079        Perl does. This was true of both matching functions. Now it matches only at
3080        the start of the subject and immediately after *internal* newlines.
3082     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
3083        a pointer to an int instead of a pointer to an unsigned long int. This
3084        caused problems on 64-bit systems.
3086     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
3087        instance of the 'standard' template library not being so standard".
3089    10. There was no check on the number of named subpatterns nor the maximum
3090        length of a subpattern name. The product of these values is used to compute
3091        the size of the memory block for a compiled pattern. By supplying a very
3092        long subpattern name and a large number of named subpatterns, the size
3093        computation could be caused to overflow. This is now prevented by limiting
3094        the length of names to 32 characters, and the number of named subpatterns
3095        to 10,000.
3097    11. Subpatterns that are repeated with specific counts have to be replicated in
3098        the compiled pattern. The size of memory for this was computed from the
3099        length of the subpattern and the repeat count. The latter is limited to
3100        65535, but there was no limit on the former, meaning that integer overflow
3101        could in principle occur. The compiled length of a repeated subpattern is
3102        now limited to 30,000 bytes in order to prevent this.
3104    12. Added the optional facility to have named substrings with the same name.
3106    13. Added the ability to use a named substring as a condition, using the
3107        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
3108        are numbers (not recommended). Forward references are permitted.
3110    14. Added forward references in named backreferences (if you see what I mean).
3112    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
3113        pattern could run off the end of the subject. For example, the pattern
3114        "(?s)(.{1,5})"8 did this with the subject "ab".
3116    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
3117        PCRE_CASELESS was set when matching characters that were quantified with ?
3118        or *.
3120    17. A character class other than a single negated character that had a minimum
3121        but no maximum quantifier - for example [ab]{6,} - was not handled
3122        correctly by pce_dfa_exec(). It would match only one character.
3124    18. A valid (though odd) pattern that looked like a POSIX character
3125        class but used an invalid character after [ (for example [[,abc,]]) caused
3126        pcre_compile() to give the error "Failed: internal error: code overflow" or
3127        in some cases to crash with a glibc free() error. This could even happen if
3128        the pattern terminated after [[ but there just happened to be a sequence of
3129        letters, a binary zero, and a closing ] in the memory that followed.
3131    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
3132        over the years. Originally (before any Unicode support), just the bottom 8
3133        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
3134        output from "man perlunicode" includes this:
3136          The regular expression compiler produces polymorphic opcodes.  That
3137          is, the pattern adapts to the data and automatically switches to
3138          the Unicode character scheme when presented with Unicode data--or
3139          instead uses a traditional byte scheme when presented with byte
3140          data.
3142        Sadly, a wide octal escape does not cause a switch, and in a string with
3143        no other multibyte characters, these octal escapes are treated as before.
3144        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
3145        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
3146        Unicode string.
3148        I have not perpetrated such confusion in PCRE. Up till now, it took just
3149        the bottom 8 bits, as in old Perl. I have now made octal escapes with
3150        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
3151        translate to the appropriate multibyte character.
3153    29. Applied some refactoring to reduce the number of warnings from Microsoft
3154        and Borland compilers. This has included removing the fudge introduced
3155        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
3156        a warning about an unused variable.
3158    21. PCRE has not included VT (character 0x0b) in the set of whitespace
3159        characters since release 4.0, because Perl (from release 5.004) does not.
3160        [Or at least, is documented not to: some releases seem to be in conflict
3161        with the documentation.] However, when a pattern was studied with
3162        pcre_study() and all its branches started with \s, PCRE still included VT
3163        as a possible starting character. Of course, this did no harm; it just
3164        caused an unnecessary match attempt.
3166    22. Removed a now-redundant internal flag bit that recorded the fact that case
3167        dependency changed within the pattern. This was once needed for "required
3168        byte" processing, but is no longer used. This recovers a now-scarce options
3169        bit. Also moved the least significant internal flag bit to the most-
3170        significant bit of the word, which was not previously used (hangover from
3171        the days when it was an int rather than a uint) to free up another bit for
3172        the future.
3174    23. Added support for CRLF line endings as well as CR and LF. As well as the
3175        default being selectable at build time, it can now be changed at runtime
3176        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3177        specify that it is scanning data with non-default line endings.
3179    24. Changed the definition of CXXLINK to make it agree with the definition of
3180        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3182    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3183        recursions. This makes a big different to stack usage for some patterns.
3185    26. If a subpattern containing a named recursion or subroutine reference such
3186        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3187        the space required for the compiled pattern went wrong and gave too small a
3188        value. Depending on the environment, this could lead to "Failed: internal
3189        error: code overflow at offset 49" or "glibc detected double free or
3190        corruption" errors.
3192    27. Applied patches from Google (a) to support the new newline modes and (b) to
3193        advance over multibyte UTF-8 characters in GlobalReplace.
3195    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3196        difference for some implementation of PCRE in some Windows version.
3198    29. Added some extra testing facilities to pcretest:
3200        \q<number>   in a data line sets the "match limit" value
3201        \Q<number>   in a data line sets the "match recursion limt" value
3202        -S <number>  sets the stack size, where <number> is in megabytes
3204        The -S option isn't available for Windows.
3207    Version 6.6 06-Feb-06
3208    ---------------------
3210     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3211        in pcreposix.h. I have copied the definition from pcre.h.
3213     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3214        because pcre.h is no longer a built file.
3216     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3217        not normally included in the compiled code.
3220    Version 6.5 01-Feb-06
3221    ---------------------
3223     1. When using the partial match feature with pcre_dfa_exec(), it was not
3224        anchoring the second and subsequent partial matches at the new starting
3225        point. This could lead to incorrect results. For example, with the pattern
3226        /1234/, partially matching against "123" and then "a4" gave a match.
3228     2. Changes to pcregrep:
3230        (a) All non-match returns from pcre_exec() were being treated as failures
3231            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3232            error message is output. Some extra information is given for the
3233            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
3234            probably the only errors that are likely to be caused by users (by
3235            specifying a regex that has nested indefinite repeats, for instance).
3236            If there are more than 20 of these errors, pcregrep is abandoned.
3238        (b) A binary zero was treated as data while matching, but terminated the
3239            output line if it was written out. This has been fixed: binary zeroes
3240            are now no different to any other data bytes.
3242        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3243            used to set a locale for matching. The --locale=xxxx long option has
3244            been added (no short equivalent) to specify a locale explicitly on the
3245            pcregrep command, overriding the environment variables.
3247        (d) When -B was used with -n, some line numbers in the output were one less
3248            than they should have been.
3250        (e) Added the -o (--only-matching) option.
3252        (f) If -A or -C was used with -c (count only), some lines of context were
3253            accidentally printed for the final match.
3255        (g) Added the -H (--with-filename) option.
3257        (h) The combination of options -rh failed to suppress file names for files
3258            that were found from directory arguments.
3260        (i) Added the -D (--devices) and -d (--directories) options.
3262        (j) Added the -F (--fixed-strings) option.
3264        (k) Allow "-" to be used as a file name for -f as well as for a data file.
3266        (l) Added the --colo(u)r option.
3268        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3269            is not present by default.
3271     3. A nasty bug was discovered in the handling of recursive patterns, that is,
3272        items such as (?R) or (?1), when the recursion could match a number of
3273        alternatives. If it matched one of the alternatives, but subsequently,
3274        outside the recursion, there was a failure, the code tried to back up into
3275        the recursion. However, because of the way PCRE is implemented, this is not
3276        possible, and the result was an incorrect result from the match.
3278        In order to prevent this happening, the specification of recursion has
3279        been changed so that all such subpatterns are automatically treated as
3280        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3282     4. I had overlooked the fact that, in some locales, there are characters for
3283        which isalpha() is true but neither isupper() nor islower() are true. In
3284        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3285        and ordfeminine) are like this. This affected the treatment of \w and \W
3286        when they appeared in character classes, but not when they appeared outside
3287        a character class. The bit map for "word" characters is now created
3288        separately from the results of isalnum() instead of just taking it from the
3289        upper, lower, and digit maps. (Plus the underscore character, of course.)
3291     5. The above bug also affected the handling of POSIX character classes such as
3292        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3293        permanent tables. Instead, the bit maps for such a class were previously
3294        created as the appropriate unions of the upper, lower, and digit bitmaps.
3295        Now they are created by subtraction from the [[:word:]] class, which has
3296        its own bitmap.
3298     6. The [[:blank:]] character class matches horizontal, but not vertical space.
3299        It is created by subtracting the vertical space characters (\x09, \x0a,
3300        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3301        subtraction was done in the overall bitmap for a character class, meaning
3302        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3303        be recognized. This bug has been fixed.
3305     7. Patches from the folks at Google:
3307          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3308          real life, but is still worth protecting against".
3310          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3311          regular expressions".
3313          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3314          have it.
3316          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3317          "configure" and the latter not, in order to fix a problem somebody had
3318          with compiling the Arg class on HP-UX.
3320          (e) Improve the error-handling of the C++ wrapper a little bit.
3322          (f) New tests for checking recursion limiting.
3324     8. The pcre_memmove() function, which is used only if the environment does not
3325        have a standard memmove() function (and is therefore rarely compiled),
3326        contained two bugs: (a) use of int instead of size_t, and (b) it was not
3327        returning a result (though PCRE never actually uses the result).
3329     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3330        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3331        returned instead of calling malloc() with an overflowing number that would
3332        most likely cause subsequent chaos.
3334    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3336    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3337        with this option is matched, the nmatch and pmatch options of regexec() are
3338        ignored.
3340    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3341        provided in case anyone wants to the the POSIX interface with UTF-8
3342        strings.
3344    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3345        C++ linking (needed for some HP-UX environments).
3347    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3348        (unused parameter) and in the pcre_printint() function (omitted "default"
3349        switch label when the default is to do nothing).
3351    15. Added some code to make it possible, when PCRE is compiled as a C++
3352        library, to replace subject pointers for pcre_exec() with a smart pointer
3353        class, thus making it possible to process discontinuous strings.
3355    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3356        much the same function. They were added by different people who were trying
3357        to make PCRE easy to compile on non-Unix systems. It has been suggested
3358        that PCRE_EXPORT be abolished now that there is more automatic apparatus
3359        for compiling on Windows systems. I have therefore replaced it with
3360        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3361        defaults to "extern" for C or "extern C" for C++, which works fine on
3362        Unix-like systems. It is now possible to override the value of PCRE_DATA_
3363        SCOPE with something explicit in config.h. In addition:
3365        (a) pcreposix.h still had just "extern" instead of either of these macros;
3366            I have replaced it with PCRE_DATA_SCOPE.
3368        (b) Functions such as _pcre_xclass(), which are internal to the library,
3369            but external in the C sense, all had PCRE_EXPORT in their definitions.
3370            This is apparently wrong for the Windows case, so I have removed it.
3371            (It makes no difference on Unix-like systems.)
3373    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3374        of recursive calls to match(). This is different to MATCH_LIMIT because
3375        that limits the total number of calls to match(), not all of which increase
3376        the depth of recursion. Limiting the recursion depth limits the amount of
3377        stack (or heap if NO_RECURSE is set) that is used. The default can be set
3378        when PCRE is compiled, and changed at run time. A patch from Google adds
3379        this functionality to the C++ interface.
3381    18. Changes to the handling of Unicode character properties:
3383        (a) Updated the table to Unicode 4.1.0.
3385        (b) Recognize characters that are not in the table as "Cn" (undefined).
3387        (c) I revised the way the table is implemented to a much improved format
3388            which includes recognition of ranges. It now supports the ranges that
3389            are defined in UnicodeData.txt, and it also amalgamates other
3390            characters into ranges. This has reduced the number of entries in the
3391            table from around 16,000 to around 3,000, thus reducing its size
3392            considerably. I realized I did not need to use a tree structure after
3393            all - a binary chop search is just as efficient. Having reduced the
3394            number of entries, I extended their size from 6 bytes to 8 bytes to
3395            allow for more data.
3397        (d) Added support for Unicode script names via properties such as \p{Han}.
3399    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3400        matching that character.
3402    20. When matching a repeated Unicode property with a minimum greater than zero,
3403        (for example \pL{2,}), PCRE could look past the end of the subject if it
3404        reached it while seeking the minimum number of characters. This could
3405        happen only if some of the characters were more than one byte long, because
3406        there is a check for at least the minimum number of bytes.
3408    21. Refactored the implementation of \p and \P so as to be more general, to
3409        allow for more different types of property in future. This has changed the
3410        compiled form incompatibly. Anybody with saved compiled patterns that use
3411        \p or \P will have to recompile them.
3413    22. Added "Any" and "L&" to the supported property types.
3415    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3416        but give a compile time error if the value is greater than 0xff.
3418    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3419        accidentally not being installed or uninstalled.
3421    25. The pcre.h file was built from pcre.h.in, but the only changes that were
3422        made were to insert the current release number. This seemed silly, because
3423        it made things harder for people building PCRE on systems that don't run
3424        "configure". I have turned pcre.h into a distributed file, no longer built
3425        by "configure", with the version identification directly included. There is
3426        no longer a pcre.h.in file.
3428        However, this change necessitated a change to the pcre-config script as
3429        well. It is built from pcre-config.in, and one of the substitutions was the
3430        release number. I have updated configure.ac so that ./configure now finds
3431        the release number by grepping pcre.h.
3433    26. Added the ability to run the tests under valgrind.
3436    Version 6.4 05-Sep-05
3437    ---------------------
3439     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3440        "--" to be printed when multiple files were scanned, even when none of the
3441        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3442        consider it to be a bug, and have restored the previous behaviour.
3444     2. A couple of code tidies to get rid of compiler warnings.
3446     3. The pcretest program used to cheat by referring to symbols in the library
3447        whose names begin with _pcre_. These are internal symbols that are not
3448        really supposed to be visible externally, and in some environments it is
3449        possible to suppress them. The cheating is now confined to including
3450        certain files from the library's source, which is a bit cleaner.
3452     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3453        file's purpose clearer.
3455     5. Reorganized pcre_ucp_findchar().
3458    Version 6.3 15-Aug-05
3459    ---------------------
3461     1. The file libpcre.pc.in did not have general read permission in the tarball.
3463     2. There were some problems when building without C++ support:
3465        (a) If C++ support was not built, "make install" and "make test" still
3466            tried to test it.
3468        (b) There were problems when the value of CXX was explicitly set. Some
3469            changes have been made to try to fix these, and ...
3471        (c) --disable-cpp can now be used to explicitly disable C++ support.
3473        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3474            backslash in a target when C++ was disabled. This confuses some
3475            versions of "make", apparently. Using an intermediate variable solves
3476            this. (Same for CPP_LOBJ.)
3478     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3479        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3480        necessary on certain architectures.
3482     4. Added a setting of -export-symbols-regex to the link command to remove
3483        those symbols that are exported in the C sense, but actually are local
3484        within the library, and not documented. Their names all begin with
3485        "_pcre_". This is not a perfect job, because (a) we have to except some
3486        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3487        available (and never for static libraries). I have made a note to try to
3488        find a way round (a) in the future.
3491    Version 6.2 01-Aug-05
3492    ---------------------
3494     1. There was no test for integer overflow of quantifier values. A construction
3495        such as {1111111111111111} would give undefined results. What is worse, if
3496        a minimum quantifier for a parenthesized subpattern overflowed and became
3497        negative, the calculation of the memory size went wrong. This could have
3498        led to memory overwriting.
3500     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3502     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3503        operating environments where this matters.
3505     4. Applied Giuseppe Maxia's patch to add additional features for controlling
3506        PCRE options from within the C++ wrapper.
3508     5. Named capturing subpatterns were not being correctly counted when a pattern
3509        was compiled. This caused two problems: (a) If there were more than 100
3510        such subpatterns, the calculation of the memory needed for the whole
3511        compiled pattern went wrong, leading to an overflow error. (b) Numerical
3512        back references of the form \12, where the number was greater than 9, were
3513        not recognized as back references, even though there were sufficient
3514        previous subpatterns.
3516     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3517        versions of gcc, e.g. 2.95.4.
3520    Version 6.1 21-Jun-05
3521    ---------------------
3523     1. There was one reference to the variable "posix" in pcretest.c that was not
3524        surrounded by "#if !defined NOPOSIX".
3526     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3527        the cross-check on the old pcre_info() function, for the benefit of the
3528        cut-down version of PCRE that is currently imported into Exim.
3530     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3531        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3532        patterns that start (?i) but I don't think that matters. The use of (?i) is
3533        just an example; this all applies to the other options as well.
3535     4. Since libtool seems to echo the compile commands it is issuing, the output
3536        from "make" can be reduced a bit by putting "@" in front of each libtool
3537        compile command.
3539     5. Patch from the folks at Google for configure.in to be a bit more thorough
3540        in checking for a suitable C++ installation before trying to compile the
3541        C++ stuff. This should fix a reported problem when a compiler was present,
3542        but no suitable headers.
3544     6. The man pages all had just "PCRE" as their title. I have changed them to
3545        be the relevant file name. I have also arranged that these names are
3546        retained in the file doc/pcre.txt, which is a concatenation in text format
3547        of all the man pages except the little individual ones for each function.
3549     7. The NON-UNIX-USE file had not been updated for the different set of source
3550        files that come with release 6. I also added a few comments about the C++
3551        wrapper.
3554    Version 6.0 07-Jun-05
3555    ---------------------
3557     1. Some minor internal re-organization to help with my DFA experiments.
3559     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3560        didn't matter for the library itself when fully configured, but did matter
3561        when compiling without UCP support, or within Exim, where the ucp files are
3562        not imported.
3564     3. Refactoring of the library code to split up the various functions into
3565        different source modules. The addition of the new DFA matching code (see
3566        below) to a single monolithic source would have made it really too
3567        unwieldy, quite apart from causing all the code to be include in a
3568        statically linked application, when only some functions are used. This is
3569        relevant even without the DFA addition now that patterns can be compiled in
3570        one application and matched in another.
3572        The downside of splitting up is that there have to be some external
3573        functions and data tables that are used internally in different modules of
3574        the library but which are not part of the API. These have all had their
3575        names changed to start with "_pcre_" so that they are unlikely to clash
3576        with other external names.
3578     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3579        a different (DFA) algorithm. Although it is slower than the original
3580        function, it does have some advantages for certain types of matching
3581        problem.
3583     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3584        including restarting after a partial match.
3586     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3587        defined when compiling for Windows was sent to me. I have put it into the
3588        code, though I have no means of testing or verifying it.
3590     7. Added the pcre_refcount() auxiliary function.
3592     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3593        match before or at the first newline in the subject string. In pcretest,
3594        the /f opti