ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 167 by ph10, Wed May 9 15:53:54 2007 UTC revision 1415 by zherczeg, Sun Dec 22 20:47:08 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4  Version 7.2 01-May-07  Version 8.35-RC1 xx-xxxx-201x
5    -----------------------------
7    1.  A new flag is set, when property checks are present in an XCLASS.
8        When this flag is not set, PCRE can perform certain optimizations
9        such as studying these XCLASS-es.
11    2.  The auto-possessification of character sets were improved: a normal
12        and an extended character set can be compared now. Furthermore
13        the JIT compiler optimizes more character set checks.
16    Version 8.34 15-December-2013
17    -----------------------------
19    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
20        executable memory. Patch inspired by Carsten Klein.
22    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
23        this macro is never tested and has no effect, because the work to support
24        coverage involves only compiling and linking options and special targets in
25        the Makefile. The comment in config.h implied that defining the macro would
26        enable coverage support, which is totally false. There was also support for
27        setting this macro in the CMake files (my fault, I just copied it from
28        configure). SUPPORT_GCOV has now been removed.
30    3.  Make a small performance improvement in strlen16() and strlen32() in
31        pcretest.
33    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
34        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
35        these cases). There was also one in pcretest.c.
37    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
39    6.  In UTF mode, the code for checking whether a group could match an empty
40        string (which is used for indefinitely repeated groups to allow for
41        breaking an infinite loop) was broken when the group contained a repeated
42        negated single-character class with a character that occupied more than one
43        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
44        UTF-8 mode). The effect was undefined: the group might or might not be
45        deemed as matching an empty string, or the program might have crashed.
47    7.  The code for checking whether a group could match an empty string was not
48        recognizing that \h, \H, \v, \V, and \R must match a character.
50    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
51        an empty string. If it can, pcretest shows this in its information output.
53    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
54        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
55        matched by pcre_exec() without using JIT:
57        (a) If the rest of the pattern did not match after a maximal run of
58            grapheme clusters, the code for backing up to try with fewer of them
59            did not always back up over a full grapheme when characters that do not
60            have the modifier quality were involved, e.g. Hangul syllables.
62        (b) If the match point in a subject started with modifier character, and
63            there was no match, the code could incorrectly back up beyond the match
64            point, and potentially beyond the first character in the subject,
65            leading to a segfault or an incorrect match result.
67    10. A conditional group with an assertion condition could lead to PCRE
68        recording an incorrect first data item for a match if no other first data
69        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
70        first data item, and therefore matched "ca" after "c" instead of at the
71        start.
73    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
74        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
76    12. The source of pcregrep now includes z/OS-specific code so that it can be
77        compiled for z/OS as part of the special z/OS distribution.
79    13. Added the -T and -TM options to pcretest.
81    14. The code in pcre_compile.c for creating the table of named capturing groups
82        has been refactored. Instead of creating the table dynamically during the
83        actual compiling pass, the information is remembered during the pre-compile
84        pass (on the stack unless there are more than 20 named groups, in which
85        case malloc() is used) and the whole table is created before the actual
86        compile happens. This has simplified the code (it is now nearly 150 lines
87        shorter) and prepared the way for better handling of references to groups
88        with duplicate names.
90    15. A back reference to a named subpattern when there is more than one of the
91        same name now checks them in the order in which they appear in the pattern.
92        The first one that is set is used for the reference. Previously only the
93        first one was inspected. This change makes PCRE more compatible with Perl.
95    16. Unicode character properties were updated from Unicode 6.3.0.
97    17. The compile-time code for auto-possessification has been refactored, based
98        on a patch by Zoltan Herczeg. It now happens after instead of during
99        compilation. The code is cleaner, and more cases are handled. The option
100        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
101        options in pcretest are provided to set it. It can also be set by
102        (*NO_AUTO_POSSESS) at the start of a pattern.
104    18. The character VT has been added to the default ("C" locale) set of
105        characters that match \s and are generally treated as white space,
106        following this same change in Perl 5.18. There is now no difference between
107        "Perl space" and "POSIX space". Whether VT is treated as white space in
108        other locales depends on the locale.
110    19. The code for checking named groups as conditions, either for being set or
111        for being recursed, has been refactored (this is related to 14 and 15
112        above). Processing unduplicated named groups should now be as fast at
113        numerical groups, and processing duplicated groups should be faster than
114        before.
116    20. Two patches to the CMake build system, by Alexander Barkov:
118          (1) Replace the "source" command by "." in CMakeLists.txt because
119              "source" is a bash-ism.
121          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
122              without these the CMake build does not work on Solaris.
124    21. Perl has changed its handling of \8 and \9. If there is no previously
125        encountered capturing group of those numbers, they are treated as the
126        literal characters 8 and 9 instead of a binary zero followed by the
127        literals. PCRE now does the same.
129    22. Following Perl, added \o{} to specify codepoints in octal, making it
130        possible to specify values greater than 0777 and also making them
131        unambiguous.
133    23. Perl now gives an error for missing closing braces after \x{... instead of
134        treating the string as literal. PCRE now does the same.
136    24. RunTest used to grumble if an inappropriate test was selected explicitly,
137        but just skip it when running all tests. This make it awkward to run ranges
138        of tests when one of them was inappropriate. Now it just skips any
139        inappropriate tests, as it always did when running all tests.
141    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
142        character types such as \d or \w, too many callouts were inserted, and the
143        data that they returned was rubbish.
145    26. In UCP mode, \s was not matching two of the characters that Perl matches,
146        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
147        were matched by \h. The code has now been refactored so that the lists of
148        the horizontal and vertical whitespace characters used for \h and \v (which
149        are defined only in one place) are now also used for \s.
151    27. Add JIT support for the 64 bit TileGX architecture.
152        Patch by Jiong Wang (Tilera Corporation).
154    28. Possessive quantifiers for classes (both explicit and automatically
155        generated) now use special opcodes instead of wrapping in ONCE brackets.
157    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
158        (because it's meaningless), this was not happening when PCRE_CASELESS was
159        set. Not wrong, but inefficient.
161    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
162        properties for \w, \d, etc) is present in a test regex. Otherwise if the
163        test contains no characters greater than 255, Perl doesn't realise it
164        should be using Unicode semantics.
166    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
167        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
168        does in Unicode mode.
170    32. Added the "forbid" facility to pcretest so that putting tests into the
171        wrong test files can sometimes be quickly detected.
173    33. There is now a limit (default 250) on the depth of nesting of parentheses.
174        This limit is imposed to control the amount of system stack used at compile
175        time. It can be changed at build time by --with-parens-nest-limit=xxx or
176        the equivalent in CMake.
178    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
179        errors. Perl warns for these when in warning mode, but PCRE has no facility
180        for giving warnings.
182    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
183        However, this was not working for (?!) because it is optimized to (*FAIL),
184        for which PCRE does not allow quantifiers. The optimization is now disabled
185        when a quantifier follows (?!). I can't see any use for this, but it makes
186        things uniform.
188    36. Perl no longer allows group names to start with digits, so I have made this
189        change also in PCRE. It simplifies the code a bit.
191    37. In extended mode, Perl ignores spaces before a + that indicates a
192        possessive quantifier. PCRE allowed a space before the quantifier, but not
193        before the possessive +. It now does.
195    38. The use of \K (reset reported match start) within a repeated possessive
196        group such as (a\Kb)*+ was not working.
198    40. Document that the same character tables must be used at compile time and
199        run time, and that the facility to pass tables to pcre_exec() and
200        pcre_dfa_exec() is for use only with saved/restored patterns.
202    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
203        features for Builds with MSVC:
205        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
206           builds). The .rc files can be used to set FileDescription and many other
207           attributes.
209        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
210           This allows higher-level build scripts which want .pdb files to avoid
211           hard-coding the exact files needed."
213    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
214        mean "start of word" and "end of word", respectively, as a transition aid.
216    43. A minimizing repeat of a class containing codepoints greater than 255 in
217        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
218        compiled to use the heap for recursion.
220    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
221        is configured.
224    Version 8.33 28-May-2013
225    ------------------------
227    1.  Added 'U' to some constants that are compared to unsigned integers, to
228        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
229        variables that are added to signed variables, to ensure the result is
230        signed and can be negated.
232    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
233        CMake config files.
235    3.  Revise the creation of config.h.generic so that all boolean macros are
236        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
237        overriding via -D on the command line possible.
239    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
240        to unsigned int is reported to make a quite noticeable speed difference in
241        a specific Windows environment. Testing on Linux did also appear to show
242        some benefit (and it is clearly not harmful). Also fixed the definition of
243        Xop which should be unsigned.
245    5.  Related to (4), changing the definition of the intermediate variable cc
246        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
247        improvements.
249    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
250        unnecessary spaces.
252    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
253        and later.
255    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
257    9.  Optimizing fast_forward_start_bits in JIT.
259    10. Adding support for callouts in JIT, and fixing some issues revealed
260        during this work. Namely:
262        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
264        (b) Minimum length was not checked before the matching is started.
266    11. The value of capture_last that is passed to callouts was incorrect in some
267        cases when there was a capture on one path that was subsequently abandoned
268        after a backtrack. Also, the capture_last value is now reset after a
269        recursion, since all captures are also reset in this case.
271    12. The interpreter no longer returns the "too many substrings" error in the
272        case when an overflowing capture is in a branch that is subsequently
273        abandoned after a backtrack.
275    13. In the pathological case when an offset vector of size 2 is used, pcretest
276        now prints out the matched string after a yield of 0 or 1.
278    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
279        Only supported by the JIT compiler at the moment.
281    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
283    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
284        offset of the starting point of the matching process, provided the offsets
285        vector is large enough.
287    17. The \A escape now records a lookbehind value of 1, though its execution
288        does not actually inspect the previous character. This is to ensure that,
289        in partial multi-segment matching, at least one character from the old
290        segment is retained when a new segment is processed. Otherwise, if there
291        are no lookbehinds in the pattern, \A might match incorrectly at the start
292        of a new segment.
294    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
296    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
297        gives some modest performance improvement in 8-bit mode.
299    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
300        be expressed in certain programming languages using Universal Character
301        Names.
303    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
304        which points out that "non characters" are not "characters that may not
305        appear in Unicode strings" but rather "characters that are reserved for
306        internal use and have only local meaning".
308    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
309        there was a conditional group that depended on an assertion, if the
310        assertion was false, the callout that immediately followed the alternation
311        in the condition was skipped when pcre_exec() was used for matching.
313    23. Allow an explicit callout to be inserted before an assertion that is the
314        condition for a conditional group, for compatibility with automatic
315        callouts, which always insert a callout at this point.
317    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
318        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
320    25. (*PRUNE) is now supported by the JIT compiler.
322    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
324    27. Fix the case where there are two or more SKIPs with arguments that may be
325        ignored.
327    28. (*SKIP) is now supported by the JIT compiler.
329    29. (*THEN) is now supported by the JIT compiler.
331    30. Update RunTest with additional test selector options.
333    31. The way PCRE handles backtracking verbs has been changed in two ways.
335        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
336        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
337        backtracking. In some cases this makes it more Perl-compatible, but Perl's
338        rather obscure rules do not always do the same thing.
340        (2) Previously, backtracking verbs were confined within assertions. This is
341        no longer the case for positive assertions, except for (*ACCEPT). Again,
342        this sometimes improves Perl compatibility, and sometimes does not.
344    32. A number of tests that were in test 2 because Perl did things differently
345        have been moved to test 1, because either Perl or PCRE has changed, and
346        these tests are now compatible.
348    32. Backtracking control verbs are now handled in the same way in JIT and
349        interpreter.
351    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
352        contained a forward subroutine reference caused a compile error.
354    34. Auto-detect and optimize limited repetitions in JIT.
356    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
357        blocking (*UTF) etc.
359    36. In the interpreter, maximizing pattern repetitions for characters and
360        character types now use tail recursion, which reduces stack usage.
362    37. The value of the max lookbehind was not correctly preserved if a compiled
363        and saved regex was reloaded on a host of different endianness.
365    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
366        of the compiled pattern block, expand the flags field from 16 to 32 bits
367        because it was almost full.
369    39. Try madvise first before posix_madvise.
371    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
372        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
373        original change locked out all matches of empty strings. This has been
374        changed so that one match of an empty string per line is recognized.
375        Subsequent searches on the same line (for colouring or for --only-matching,
376        for example) do not recognize empty strings.
378    41. Applied a user patch to fix a number of spelling mistakes in comments.
380    42. Data lines longer than 65536 caused pcretest to crash.
382    43. Clarified the data type for length and startoffset arguments for pcre_exec
383        and pcre_dfa_exec in the function-specific man pages, where they were
384        explicitly stated to be in bytes, never having been updated. I also added
385        some clarification to the pcreapi man page.
387    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
388        a segmentation fault.
391    Version 8.32 30-November-2012
392    -----------------------------
394    1.  Improved JIT compiler optimizations for first character search and single
395        character iterators.
397    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
398        Patch by Daniel Richard G.
400    3.  Single character iterator optimizations in the JIT compiler.
402    4.  Improved JIT compiler optimizations for character ranges.
404    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
405        Reported by Giuseppe D'Angelo.
407    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
408        of a line, was being set incorrectly in cases where .* appeared inside
409        atomic brackets at the start of a pattern, or where there was a subsequent
410        *PRUNE or *SKIP.
412    7.  Improved instruction cache flush for POWER/PowerPC.
413        Patch by Daniel Richard G.
415    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
416        grep:
418        (a) There is now no limit to the number of patterns to be matched.
420        (b) An error is given if a pattern is too long.
422        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
423            are now supported.
425        (d) --exclude-from and --include-from (multiple use) have been added.
427        (e) Exclusions and inclusions now apply to all files and directories, not
428            just to those obtained from scanning a directory recursively.
430        (f) Multiple uses of -f and --file-list are now supported.
432        (g) In a Windows environment, the default for -d has been changed from
433            "read" (the GNU grep default) to "skip", because otherwise the presence
434            of a directory in the file list provokes an error.
436        (h) The documentation has been revised and clarified in places.
438    9.  Improve the matching speed of capturing brackets.
440    10. Changed the meaning of \X so that it now matches a Unicode extended
441        grapheme cluster.
443    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
444        out POSIX threads when JIT support is configured.
446    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
448    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
449        structure could go wrong in environments where size_t is not the same size
450        as int.
452    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
453        set.
455    15. The EBCDIC support had decayed; later updates to the code had included
456        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
457        general tidy up of EBCDIC-related issues, and the documentation was also
458        not quite right. There is now a test that can be run on ASCII systems to
459        check some of the EBCDIC-related things (but is it not a full test).
461    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
462        in a small tidy to the code.
464    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
466    18. If the --only-matching (-o) option in pcregrep is specified multiple
467        times, each one causes appropriate output. For example, -o1 -o2 outputs the
468        substrings matched by the 1st and 2nd capturing parentheses. A separating
469        string can be specified by --om-separator (default empty).
471    19. Improving the first n character searches.
473    20. Turn case lists for horizontal and vertical white space into macros so that
474        they are defined only once.
476    21. This set of changes together give more compatible Unicode case-folding
477        behaviour for characters that have more than one other case when UCP
478        support is available.
480        (a) The Unicode property table now has offsets into a new table of sets of
481            three or more characters that are case-equivalent. The MultiStage2.py
482            script that generates these tables (the pcre_ucd.c file) now scans
483            CaseFolding.txt instead of UnicodeData.txt for character case
484            information.
486        (b) The code for adding characters or ranges of characters to a character
487            class has been abstracted into a generalized function that also handles
488            case-independence. In UTF-mode with UCP support, this uses the new data
489            to handle characters with more than one other case.
491        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
492            whose other case is greater than 256 are now correctly matched
493            caselessly. Previously, the high codepoint matched the low one, but not
494            vice versa.
496        (d) The processing of \h, \H, \v, and \ in character classes now makes use
497            of the new class addition function, using character lists defined as
498            macros alongside the case definitions of 20 above.
500        (e) Caseless back references now work with characters that have more than
501            one other case.
503        (f) General caseless matching of characters with more than one other case
504            is supported.
506    22. Unicode character properties were updated from Unicode 6.2.0
508    23. Improved CMake support under Windows. Patch by Daniel Richard G.
510    24. Add support for 32-bit character strings, and UTF-32
512    25. Major JIT compiler update (code refactoring and bugfixing).
513        Experimental Sparc 32 support is added.
515    26. Applied a modified version of Daniel Richard G's patch to create
516        pcre.h.generic and config.h.generic by "make" instead of in the
517        PrepareRelease script.
519    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
520        pcre_compile.c when checking for a zero character.
522    28. Introducing a native interface for JIT. Through this interface, the compiled
523        machine code can be directly executed. The purpose of this interface is to
524        provide fast pattern matching, so several sanity checks are not performed.
525        However, feature tests are still performed. The new interface provides
526        1.4x speedup compared to the old one.
528    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
529        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
530        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
532    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
533        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
534        the "old" RFC 2279). Instead, it ended up passing a negative length to
535        pcre_exec().
537    31. Add support for GCC's visibility feature to hide internal functions.
539    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
540        "unknown -C option" after outputting 0 or 1.
542    33. There is now support for generating a code coverage report for the test
543        suite in environments where gcc is the compiler and lcov is installed. This
544        is mainly for the benefit of the developers.
546    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
547        unaddressable using valgrind annotations, allowing valgrind to detect
548        invalid memory accesses. This is mainly for the benefit of the developers.
550    25. (*UTF) can now be used to start a pattern in any of the three libraries.
552    26. Give configure error if --enable-cpp but no C++ compiler found.
555    Version 8.31 06-July-2012
556    -------------------------
558    1.  Fixing a wrong JIT test case and some compiler warnings.
560    2.  Removed a bashism from the RunTest script.
562    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
563        to unsigned type, result still unsigned" that was given by an MS compiler
564        on encountering the code "-sizeof(xxx)".
566    4.  Partial matching support is added to the JIT compiler.
568    5.  Fixed several bugs concerned with partial matching of items that consist
569        of more than one character:
571        (a) /^(..)\1/ did not partially match "aba" because checking references was
572            done on an "all or nothing" basis. This also applied to repeated
573            references.
575        (b) \R did not give a hard partial match if \r was found at the end of the
576            subject.
578        (c) \X did not give a hard partial match after matching one or more
579            characters at the end of the subject.
581        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
582            a partial match for the string "\r".
584        (e) When newline was set to CRLF, the metacharacter "." did not recognize
585            a partial match for a CR character at the end of the subject string.
587    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
588        running pcretest, the text "(JIT)" added to the output whenever JIT is
589        actually used to run the match.
591    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
592        or /S+[+] with a digit between 1 and 7.
594    8.  OP_NOT now supports any UTF character not just single-byte ones.
596    9.  (*MARK) control verb is now supported by the JIT compiler.
598    10. The command "./RunTest list" lists the available tests without actually
599        running any of them. (Because I keep forgetting what they all are.)
603    12. Applied a (slightly modified) user-supplied patch that improves performance
604        when the heap is used for recursion (compiled with --disable-stack-for-
605        recursion). Instead of malloc and free for each heap frame each time a
606        logical recursion happens, frames are retained on a chain and re-used where
607        possible. This sometimes gives as much as 30% improvement.
609    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
610        call.
612    14. As documented, (*COMMIT) is now confined to within a positive assertion.
614    15. It is now possible to link pcretest with libedit as an alternative to
615        libreadline.
617    16. (*COMMIT) control verb is now supported by the JIT compiler.
619    17. The Unicode data tables have been updated to Unicode 6.1.0.
621    18. Added --file-list option to pcregrep.
623    19. Added binary file support to pcregrep, including the -a, --binary-files,
624        -I, and --text options.
626    20. The madvise function is renamed for posix_madvise for QNX compatibility
627        reasons. Fixed by Giuseppe D'Angelo.
629    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
631    22. Changed the option for creating symbolic links for 16-bit man pages from
632        -s to -sf so that re-installing does not cause issues.
634    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
636    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
637        matches in certain environments (the workspace was not being correctly
638        retained). Also added to pcre_dfa_exec() a simple plausibility check on
639        some of the workspace data at the beginning of a restart.
641    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
642        was not doing so when it should - probably a typo introduced by SVN 528
643        (change 8.10/14).
645    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
646        \w+ when the character tables indicated that \x{c4} was a word character.
647        There were several related cases, all because the tests for doing a table
648        lookup were testing for characters less than 127 instead of 255.
650    27. If a pattern contains capturing parentheses that are not used in a match,
651        their slots in the ovector are set to -1. For those that are higher than
652        any matched groups, this happens at the end of processing. In the case when
653        there were back references that the ovector was too small to contain
654        (causing temporary malloc'd memory to be used during matching), and the
655        highest capturing number was not used, memory off the end of the ovector
656        was incorrectly being set to -1. (It was using the size of the temporary
657        memory instead of the true size.)
659    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
660        ovector size, it uses memory at the end of the block that it has got.
662    29. Check for an overlong MARK name and give an error at compile time. The
663        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
665    30. JIT compiler update.
667    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
668        Rill for the patch.
670    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
672    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
674    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
675        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
676        was enabled.
678    35. Improve JIT code generation for greedy plus quantifier.
680    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
681        1 to "aa" instead of to an empty string. The bug affected repeated groups
682        that could potentially match an empty string.
684    37. Optimizing single character iterators in JIT.
686    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
687        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
688        codepoints that are too big for the mode are faulted, and in a UTF mode,
689        disallowed codepoints are also faulted.
691    39. If PCRE was compiled with UTF support, in three places in the DFA
692        matcher there was code that should only have been obeyed in UTF mode, but
693        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
694        processing when bytes with values greater than 127 were present. In 16-bit
695        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
696        both cases the values are those that cannot be the first data item in a UTF
697        character. The three items that might have provoked this were recursions,
698        possessively repeated groups, and atomic groups.
700    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
701        and pcregrep, because some OS require shared objects to be explicitly
702        passed to ld, causing the link step to fail if they are not.
704    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
705        mode, patterns that started with \h* or \R* might be incorrectly matched.
708    Version 8.30 04-February-2012
709    -----------------------------
711    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
712        name is defined in ctype.h.
714    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
715        only in quite long subpatterns.
717    3.  Removed the function pcre_info(), which has been obsolete and deprecated
718        since it was replaced by pcre_fullinfo() in February 2000.
720    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
721        match a (*MARK), and the match failed at the start of the subject, a
722        reference to memory before the start of the subject could occur. This bug
723        was introduced by fix 17 of release 8.21.
725    5.  A reference to an unset group with zero minimum repetition was giving
726        totally wrong answers (in non-JavaScript-compatibility mode). For example,
727        /(another)?(\1?)test/ matched against "hello world test". This bug was
728        introduced in release 8.13.
730    6.  Add support for 16-bit character strings (a large amount of work involving
731        many changes and refactorings).
733    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
734        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
735        from a file.
737    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
738        rounding is not applied in this particular case).
740    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
741        if they appear, or are escaped, in patterns.
743    10. Get rid of a number of -Wunused-but-set-variable warnings.
745    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
746        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
747        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
748        also returns the mark "x". This bug applied to capturing parentheses,
749        non-capturing parentheses, and atomic parentheses. It also applied to some
750        assertions.
752    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
753        information out of configure.ac instead of relying on pcre.h.generic, which
754        is not stored in the repository.
756    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
757        -lreadline.
759    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
761    15. Applied Graycode's patch to put the top-level frame on the stack rather
762        than the heap when not using the stack for recursion. This gives a
763        performance improvement in many cases when recursion is not deep.
765    16. Experimental code added to "pcretest -C" to output the stack frame size.
768    Version 8.21 12-Dec-2011
769    ------------------------
771    1.  Updating the JIT compiler.
773    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
774        are added as well.
776    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
777        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
778        calling _pcre_jit_exec. Some extra comments are added.
780    4.  (*MARK) settings inside atomic groups that do not contain any capturing
781        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
782        was introduced by change 18 for 8.20.
784    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
785        ECMA-262 standard.
787    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
788        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
789        This bug was probably introduced by change 9 of 8.13.
791    7.  While fixing 6 above, I noticed that a number of other items were being
792        incorrectly rejected as "not fixed length". This arose partly because newer
793        opcodes had not been added to the fixed-length checking code. I have (a)
794        corrected the bug and added tests for these items, and (b) arranged for an
795        error to occur if an unknown opcode is encountered while checking for fixed
796        length instead of just assuming "not fixed length". The items that were
797        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
798        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
799        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
801    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
802        being incorrectly compiled and would have given unpredicatble results.
804    9.  A possessively repeated subpattern with minimum repeat count greater than
805        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
806        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
807        the first (A) could occur when it should not.
809    10. Add a cast and remove a redundant test from the code.
811    11. JIT should use pcre_malloc/pcre_free for allocation.
813    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
814        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
815        is anything other than /usr, -L is still shown).
817    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
819    14. Perl does not support \N without a following name in a [] class; PCRE now
820        also gives an error.
822    15. If a forward reference was repeated with an upper limit of around 2000,
823        it caused the error "internal error: overran compiling workspace". The
824        maximum number of forward references (including repeats) was limited by the
825        internal workspace, and dependent on the LINK_SIZE. The code has been
826        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
827        the default depends on LINK_SIZE. There is a new upper limit (for safety)
828        of around 200,000 forward references. While doing this, I also speeded up
829        the filling in of repeated forward references.
831    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
832        incorrectly expecting the subject to contain another "a" after the start.
834    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
835        in the match, the SKIP should be ignored. This was not happening; instead
836        the SKIP was being treated as NOMATCH. For patterns such as
837        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
838        tested.
840    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
841        now much more compatible with Perl, in particular in cases where the result
842        is a non-match for a non-anchored pattern. For example, if
843        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
844        "m", where previously it did not return a name. A side effect of this
845        change is that for partial matches, the last encountered mark name is
846        returned, as for non matches. A number of tests that were previously not
847        Perl-compatible have been moved into the Perl-compatible test files. The
848        refactoring has had the pleasing side effect of removing one argument from
849        the match() function, thus reducing its stack requirements.
851    19. If the /S+ option was used in pcretest to study a pattern using JIT,
852        subsequent uses of /S (without +) incorrectly behaved like /S+.
854    21. Retrieve executable code size support for the JIT compiler and fixing
855        some warnings.
857    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
858        not work when the shorter character appeared right at the end of the
859        subject string.
861    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
862        systems.
864    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
865        output it when the /M option is used in pcretest.
867    25. The CheckMan script was not being included in the distribution. Also, added
868        an explicit "perl" to run Perl scripts from the PrepareRelease script
869        because this is reportedly needed in Windows.
871    26. If study data was being save in a file and studying had not found a set of
872        "starts with" bytes for the pattern, the data written to the file (though
873        never used) was taken from uninitialized memory and so caused valgrind to
874        complain.
876    27. Updated RunTest.bat as provided by Sheri Pierce.
878    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
880    29. Computation of memory usage for the table of capturing group names was
881        giving an unnecessarily large value.
884    Version 8.20 21-Oct-2011
885    ------------------------
887    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
888        a POSIX class. After further experiments with Perl, which convinced me that
889        Perl has bugs and confusions, a closing square bracket is no longer allowed
890        in a POSIX name. This bug also affected patterns with classes that started
891        with full stops.
893    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
894        captured substring, but while checking the failing first alternative,
895        substring 1 is temporarily captured. If the output vector supplied to
896        pcre_exec() was not big enough for this capture, the yield of the function
897        was still zero ("insufficient space for captured substrings"). This cannot
898        be totally fixed without adding another stack variable, which seems a lot
899        of expense for a edge case. However, I have improved the situation in cases
900        such as /(a)(b)x|abc/ matched against "abc", where the return code
901        indicates that fewer than the maximum number of slots in the ovector have
902        been set.
904    3.  Related to (2) above: when there are more back references in a pattern than
905        slots in the output vector, pcre_exec() uses temporary memory during
906        matching, and copies in the captures as far as possible afterwards. It was
907        using the entire output vector, but this conflicts with the specification
908        that only 2/3 is used for passing back captured substrings. Now it uses
909        only the first 2/3, for compatibility. This is, of course, another edge
910        case.
912    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
913        main code base, and can be used by building with --enable-jit. When this is
914        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
915        runtime --no-jit option is given.
917    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
918        ovector, the return from the function was zero, implying that there were
919        other matches that did not fit. The correct "exactly full" value is now
920        returned.
922    6.  If a subpattern that was called recursively or as a subroutine contained
923        (*PRUNE) or any other control that caused it to give a non-standard return,
924        invalid errors such as "Error -26 (nested recursion at the same subject
925        position)" or even infinite loops could occur.
927    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
928        computing the minimum length on reaching *ACCEPT, and so ended up with the
929        wrong value of 1 rather than 0. Further investigation indicates that
930        computing a minimum subject length in the presence of *ACCEPT is difficult
931        (think back references, subroutine calls), and so I have changed the code
932        so that no minimum is registered for a pattern that contains *ACCEPT.
934    8.  If (*THEN) was present in the first (true) branch of a conditional group,
935        it was not handled as intended. [But see 16 below.]
937    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
938        Sheri Pierce.
940    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
941        the first byte in a match must be "a".
943    11. Change 17 for 8.13 increased the recursion depth for patterns like
944        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
945        pattern contains any instances of (*THEN). If it does not, the old
946        optimizations are restored. It would be nice to do this on a per-group
947        basis, but at the moment that is not feasible.
949    12. In some environments, the output of pcretest -C is CRLF terminated. This
950        broke RunTest's code that checks for the link size. A single white space
951        character after the value is now allowed for.
953    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
954        For "fr", it uses the Windows-specific input and output files.
956    14. If (*THEN) appeared in a group that was called recursively or as a
957        subroutine, it did not work as intended. [But see next item.]
959    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
960        pattern fragments (but not containing any | characters). If A and B are
961        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
962        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
963        D. In other words, Perl considers parentheses that do not contain any |
964        characters to be part of a surrounding alternative, whereas PCRE was
965        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
966        differently. PCRE now behaves in the same way as Perl, except in the case
967        of subroutine/recursion calls such as (?1) which have in any case always
968        been different (but PCRE had them first :-).
970    16. Related to 15 above: Perl does not treat the | in a conditional group as
971        creating alternatives. Such a group is treated in the same way as an
972        ordinary group without any | characters when processing (*THEN). PCRE has
973        been changed to match Perl's behaviour.
975    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
976        RunGrepTest script failed.
978    18. Change 22 for version 13 caused atomic groups to use more stack. This is
979        inevitable for groups that contain captures, but it can lead to a lot of
980        stack use in large patterns. The old behaviour has been restored for atomic
981        groups that do not contain any capturing parentheses.
983    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
984        suppress the check for a minimum subject length at run time. (If it was
985        given to pcre_exec() or pcre_dfa_exec() it did work.)
987    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
988        fail to work when decoding hex characters in data strings in EBCDIC
989        environments.
991    21. It appears that in at least one Mac OS environment, the isxdigit() function
992        is implemented as a macro that evaluates to its argument more than once,
993        contravening the C 90 Standard (I haven't checked a later standard). There
994        was an instance in pcretest which caused it to go wrong when processing
995        \x{...} escapes in subject strings. The has been rewritten to avoid using
996        things like p++ in the argument of isxdigit().
999    Version 8.13 16-Aug-2011
1000    ------------------------
1002    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1004    2.  Two minor typos in pcre_internal.h have been fixed.
1006    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1007        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1008        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1010    4.  There were a number of related bugs in the code for matching backrefences
1011        caselessly in UTF-8 mode when codes for the characters concerned were
1012        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1013        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1014        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1015        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1016        2-byte code at the end of the subject (it thought there wasn't enough data
1017        left).
1019    5.  Comprehensive information about what went wrong is now returned by
1020        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1021        as the output vector has at least 2 elements. The offset of the start of
1022        the failing character and a reason code are placed in the vector.
1024    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1025        now returned is for the first byte of the failing character, instead of the
1026        last byte inspected. This is an incompatible change, but I hope it is small
1027        enough not to be a problem. It makes the returned offset consistent with
1028        pcre_exec() and pcre_dfa_exec().
1030    7.  pcretest now gives a text phrase as well as the error number when
1031        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1032        failure, the offset and reason code are output.
1034    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1035        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1036        back over a single character (\n). This seems wrong (because it treated the
1037        two characters as a single entity when going forwards), conflicts with the
1038        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1039        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1040        has been changed.
1042    9.  Some internal refactoring has changed the processing so that the handling
1043        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1044        time (the PCRE_DOTALL option was changed this way some time ago: version
1045        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1046        which was always a bit of a fudge. It also means that there is one less
1047        argument for the match() function, which reduces its stack requirements
1048        slightly. This change also fixes an incompatibility with Perl: the pattern
1049        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1051    10. More internal refactoring has drastically reduced the number of recursive
1052        calls to match() for possessively repeated groups such as (abc)++ when
1053        using pcre_exec().
1055    11. While implementing 10, a number of bugs in the handling of groups were
1056        discovered and fixed:
1058        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1059        (a|)*(?1) gave a compile-time internal error.
1060        ((a|)+)+  did not notice that the outer group could match an empty string.
1061        (^a|^)+   was not marked as anchored.
1062        (.*a|.*)+ was not marked as matching at start or after a newline.
1064    12. Yet more internal refactoring has removed another argument from the match()
1065        function. Special calls to this function are now indicated by setting a
1066        value in a variable in the "match data" data block.
1068    13. Be more explicit in pcre_study() instead of relying on "default" for
1069        opcodes that mean there is no starting character; this means that when new
1070        ones are added and accidentally left out of pcre_study(), testing should
1071        pick them up.
1073    14. The -s option of pcretest has been documented for ages as being an old
1074        synonym of -m (show memory usage). I have changed it to mean "force study
1075        for every regex", that is, assume /S for every regex. This is similar to -i
1076        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1077        using it. It makes it easier to run collections of tests with and without
1078        study enabled, and thereby test pcre_study() more easily. All the standard
1079        tests are now run with and without -s (but some patterns can be marked as
1080        "never study" - see 20 below).
1082    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1083        restoration of the capturing data to the outer values was not happening
1084        correctly.
1086    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1087        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1088        pattern had matched an empty string, and so incorrectly returned a no
1089        match.
1091    17. There was optimizing code for the last branch of non-capturing parentheses,
1092        and also for the obeyed branch of a conditional subexpression, which used
1093        tail recursion to cut down on stack usage. Unfortunately, now that there is
1094        the possibility of (*THEN) occurring in these branches, tail recursion is
1095        no longer possible because the return has to be checked for (*THEN). These
1096        two optimizations have therefore been removed. [But see 8.20/11 above.]
1098    18. If a pattern containing \R was studied, it was assumed that \R always
1099        matched two bytes, thus causing the minimum subject length to be
1100        incorrectly computed because \R can also match just one byte.
1102    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1103        was incorrectly computed.
1105    20. If /S is present twice on a test pattern in pcretest input, it now
1106        *disables* studying, thereby overriding the use of -s on the command line
1107        (see 14 above). This is necessary for one or two tests to keep the output
1108        identical in both cases.
1110    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1111        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1113    22. When an atomic group that contained a capturing parenthesis was
1114        successfully matched, but the branch in which it appeared failed, the
1115        capturing was not being forgotten if a higher numbered group was later
1116        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1117        group 1 to "a", when in fact it should be unset. This applied to multi-
1118        branched capturing and non-capturing groups, repeated or not, and also to
1119        positive assertions (capturing in negative assertions does not happen
1120        in PCRE) and also to nested atomic groups.
1122    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1123        subject after a captured substring, to make it easier to tell which of a
1124        number of identical substrings has been captured.
1126    24. The way atomic groups are processed by pcre_exec() has been changed so that
1127        if they are repeated, backtracking one repetition now resets captured
1128        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1129        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1130        "aaa". Previously, it would have been "a". As part of this code
1131        refactoring, the way recursive calls are handled has also been changed.
1133    25. If an assertion condition captured any substrings, they were not passed
1134        back unless some other capturing happened later. For example, if
1135        (?(?=(a))a) was matched against "a", no capturing was returned.
1137    26. When studying a pattern that contained subroutine calls or assertions,
1138        the code for finding the minimum length of a possible match was handling
1139        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1140        group 1 called group 2 while simultaneously a separate group 2 called group
1141        1). A stack overflow occurred in this case. I have fixed this by limiting
1142        the recursion depth to 10.
1144    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1145        Fortmann. This supports explicit test numbers on the command line, and has
1146        argument validation and error reporting.
1148    28. An instance of \X with an unlimited repeat could fail if at any point the
1149        first character it looked at was a mark character.
1151    29. Some minor code refactoring concerning Unicode properties and scripts
1152        should reduce the stack requirement of match() slightly.
1154    30. Added the '=' option to pcretest to check the setting of unused capturing
1155        slots at the end of the pattern, which are documented as being -1, but are
1156        not included in the return count.
1158    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1159        compiled something random. Now it gives a compile-time error (as does
1160        Perl).
1162    32. A *MARK encountered during the processing of a positive assertion is now
1163        recorded and passed back (compatible with Perl).
1165    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1166        had alternative anchored branches, the search for a second match in a line
1167        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1168        matched the line "0102" twice. The same bug affected patterns that started
1169        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1170        twice.
1172    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1173        does, and because of capturing effects, quantifying parenthesized
1174        assertions may at times be useful. Quantifiers are now allowed for
1175        parenthesized assertions.
1177    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1179    36. \g was being checked for fancy things in a character class, when it should
1180        just be a literal "g".
1182    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1183        appearance of a nested POSIX class supersedes an apparent external class.
1184        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1185        unescaped square brackets may also appear as part of class names. For
1186        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1187        more like Perl. (But see 8.20/1 above.)
1189    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1190        was because it thought it was \N{name}, which is not supported).
1192    39. Add minix to OS list not supporting the -S option in pcretest.
1194    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1195        cannot analyze patterns in sufficient detail to catch mutual recursions
1196        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1197        subgroup is called recursively as a subpattern for a second time at the
1198        same position in the subject string. In previous releases this might have
1199        been caught by the recursion limit, or it might have run out of stack.
1201    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1202        happen only once. PCRE was, however incorrectly giving a compile time error
1203        "recursive call could loop indefinitely" because it cannot analyze the
1204        pattern in sufficient detail. The compile time test no longer happens when
1205        PCRE is compiling a conditional subpattern, but actual runaway loops are
1206        now caught at runtime (see 40 above).
1208    42. It seems that Perl allows any characters other than a closing parenthesis
1209        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1210        has been changed to be the same.
1212    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1213        as not to get warnings when autogen.sh is called. Also changed
1214        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1216    44. To help people who use pcregrep to scan files containing exceedingly long
1217        lines, the following changes have been made:
1219        (a) The default value of the buffer size parameter has been increased from
1220            8K to 20K. (The actual buffer used is three times this size.)
1222        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1223            PCRE is built.
1225        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1226            to be set at run time.
1228        (d) Numerical values in pcregrep options can be followed by K or M, for
1229            example --buffer-size=50K.
1231        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1232            given and the return code is set to 2.
1234    45. Add a pointer to the latest mark to the callout data block.
1236    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1237        partial match of an empty string instead of no match. This was specific to
1238        the use of ".".
1240    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1241        complete match instead of a partial match. This bug was dependent on both
1242        the PCRE_UTF8 and PCRE_DOTALL options being set.
1244    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1245        starting byte set, because \b was not being ignored.
1248    Version 8.12 15-Jan-2011
1249    ------------------------
1251    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1252        checks for such things as part of the documentation building process.
1254    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1255        --match-limit and --recursion-limit options (added for 8.11). In
1256        particular, this made one of the standard tests fail. (The integer value
1257        went into the wrong half of a long int.)
1259    3.  If the --colour option was given to pcregrep with -v (invert match), it
1260        did strange things, either producing crazy output, or crashing. It should,
1261        of course, ignore a request for colour when reporting lines that do not
1262        match.
1264    4.  Another pcregrep bug caused similar problems if --colour was specified with
1265        -M (multiline) and the pattern match finished with a line ending.
1267    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1268        matched in multiline mode, the following line was shown as part of the
1269        match. This seems wrong, so I have changed it.
1271    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1272        the check for further matches in the same line (so they could be coloured)
1273        to overrun the end of the current line. If another match was found, it was
1274        incorrectly shown (and then shown again when found in the next line).
1276    7.  If pcregrep was compiled under Windows, there was a reference to the
1277        function pcregrep_exit() before it was defined. I am assuming this was
1278        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1279        reported by a user. I've moved the definition above the reference.
1282    Version 8.11 10-Dec-2010
1283    ------------------------
1285    1.  (*THEN) was not working properly if there were untried alternatives prior
1286        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1287        backtracked to try for "b" instead of moving to the next alternative branch
1288        at the same level (in this case, to look for "c"). The Perl documentation
1289        is clear that when (*THEN) is backtracked onto, it goes to the "next
1290        alternative in the innermost enclosing group".
1292    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1293        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1294        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1295        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1296        (*THEN).
1298    3.  If \s appeared in a character class, it removed the VT character from
1299        the class, even if it had been included by some previous item, for example
1300        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1301        of \s, but is part of the POSIX "space" class.)
1303    4.  A partial match never returns an empty string (because you can always
1304        match an empty string at the end of the subject); however the checking for
1305        an empty string was starting at the "start of match" point. This has been
1306        changed to the "earliest inspected character" point, because the returned
1307        data for a partial match starts at this character. This means that, for
1308        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1309        (previously it gave "no match").
1311    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1312        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1313        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1314        has an implication that the given string is incomplete (because a partial
1315        match is preferred over a full match). For this reason, these items now
1316        give a partial match in this situation. [Aside: previously, the one case
1317        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1318        match rather than a full match, which was wrong by the old rules, but is
1319        now correct.]
1321    6.  There was a bug in the handling of #-introduced comments, recognized when
1322        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1323        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1324        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1325        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1326        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1327        places in pcre_compile().
1329    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1330        comments when looking ahead for named forward references to subpatterns,
1331        the only newline sequence it recognized was NL. It now handles newlines
1332        according to the set newline convention.
1334    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1335        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1336        cater for a lack of strerror(). These oversights have been fixed.
1338    9.  Added --match-limit and --recursion-limit to pcregrep.
1340    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1342    11. When the -o option was used, pcregrep was setting a return code of 1, even
1343        when matches were found, and --line-buffered was not being honoured.
1345    12. Added an optional parentheses number to the -o and --only-matching options
1346        of pcregrep.
1348    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1349        can match an empty string. The code to do it in pcretest and pcredemo
1350        needed fixing:
1352        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1353            only one byte after an empty string match just before CRLF (this case
1354            just got forgotten; "any" and "anycrlf" were OK).
1356        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1357            mode when an empty string match preceded an ASCII character followed by
1358            a non-ASCII character. (The code for advancing by one character rather
1359            than one byte was nonsense.)
1361        (c) The pcredemo.c sample program did not have any code at all to handle
1362            the cases when CRLF is a valid newline sequence.
1364    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1365        as a starting offset was within the subject string. There is now a new
1366        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1367        negative or greater than the length of the string. In order to test this,
1368        pcretest is extended to allow the setting of negative starting offsets.
1370    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1371        starting offset points to the beginning of a UTF-8 character was
1372        unnecessarily clumsy. I tidied it up.
1374    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1375        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1377    17. Nobody had reported that the --include_dir option, which was added in
1378        release 7.7 should have been called --include-dir (hyphen, not underscore)
1379        for compatibility with GNU grep. I have changed it to --include-dir, but
1380        left --include_dir as an undocumented synonym, and the same for
1381        --exclude-dir, though that is not available in GNU grep, at least as of
1382        release 2.5.4.
1384    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1385        characters from a string of bytes) have been redefined so as not to use
1386        loops, in order to improve performance in some environments. At the same
1387        time, I abstracted some of the common code into auxiliary macros to save
1388        repetition (this should not affect the compiled code).
1390    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1391        compile-time error is now given if \c is not followed by an ASCII
1392        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1393        different, and any byte value is allowed.)
1395    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1396        START_OPTIMIZE option, which is now allowed at compile time - but just
1397        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1398        to pcregrep and other applications that have no direct access to PCRE
1399        options. The new /Y option in pcretest sets this option when calling
1400        pcre_compile().
1402    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1403        back references. Groups containing recursive back references were forced to
1404        be atomic by that change, but in the case of named groups, the amount of
1405        memory required was incorrectly computed, leading to "Failed: internal
1406        error: code overflow". This has been fixed.
1408    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1409        pcretest.c, to avoid build problems in some Borland environments.
1412    Version 8.10 25-Jun-2010
1413    ------------------------
1415    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1416        THEN.
1418    2.  (*ACCEPT) was not working when inside an atomic group.
1420    3.  Inside a character class, \B is treated as a literal by default, but
1421        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1422        causes the error). The code is unchanged, but I tidied the documentation.
1424    4.  Inside a character class, PCRE always treated \R and \X as literals,
1425        whereas Perl faults them if its -w option is set. I have changed PCRE so
1426        that it faults them when PCRE_EXTRA is set.
1428    5.  Added support for \N, which always matches any character other than
1429        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1431    6.  When compiling pcregrep with newer versions of gcc which may have
1432        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1433        declared with attribute warn_unused_result" were given. Just casting the
1434        result to (void) does not stop the warnings; a more elaborate fudge is
1435        needed. I've used a macro to implement this.
1437    7.  Minor change to pcretest.c to avoid a compiler warning.
1439    8.  Added four artifical Unicode properties to help with an option to make
1440        \s etc use properties (see next item). The new properties are: Xan
1441        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1443    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1444        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1445        this option. Modified pcretest to add /W to test this facility. Added
1446        REG_UCP to make it available via the POSIX interface.
1448    10. Added --line-buffered to pcregrep.
1450    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1451        studied, and the match started with a letter with a code point greater than
1452        127 whose first byte was different to the first byte of the other case of
1453        the letter, the other case of this starting letter was not recognized
1454        (#976).
1456    12. If a pattern that was studied started with a repeated Unicode property
1457        test, for example, \p{Nd}+, there was the theoretical possibility of
1458        setting up an incorrect bitmap of starting bytes, but fortunately it could
1459        not have actually happened in practice until change 8 above was made (it
1460        added property types that matched character-matching opcodes).
1462    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1463        possible starting bytes for non-anchored patterns.
1465    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1466        \R, and also a number of cases that involve Unicode properties, both
1467        explicit and implicit when PCRE_UCP is set.
1469    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1470        input, it could crash or give wrong results if characters with values
1471        greater than 0xc0 were present in the subject string. (Detail: it assumed
1472        UTF-8 input when processing these items.)
1474    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1475        size_t is 64-bit (#991).
1477    17. Added a check for running out of memory when PCRE is compiled with
1478        --disable-stack-for-recursion (#990).
1480    18. If the last data line in a file for pcretest does not have a newline on
1481        the end, a newline was missing in the output.
1483    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1484        less than 128) in its various bitmaps. However, there is a facility for
1485        generating tables according to the current locale when PCRE is compiled. It
1486        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1487        characters, are recognized by isspace() and therefore were getting set in
1488        these tables, and indeed these tables seem to approximate to ISO 8859. This
1489        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1490        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1491        which of course cannot start UTF-8 characters. I have changed the code so
1492        that only real ASCII characters (less than 128) and the correct starting
1493        bytes for UTF-8 encodings are set for characters greater than 127 when in
1494        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1495        altogether.)
1497    20. Added the /T option to pcretest so as to be able to run tests with non-
1498        standard character tables, thus making it possible to include the tests
1499        used for 19 above in the standard set of tests.
1501    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1502        reference to a subpattern the other side of a comment that contains an
1503        opening parenthesis caused either an internal compiling error, or a
1504        reference to the wrong subpattern.
1507    Version 8.02 19-Mar-2010
1508    ------------------------
1510    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1512    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1513        configured.
1515    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1516        original author of that file, following a query about its status.
1518    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1519        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1521    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1522        quantifier applied to a forward-referencing subroutine call, could compile
1523        incorrect code or give the error "internal error: previously-checked
1524        referenced subpattern not found".
1526    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1527        variables to point to external functions. For these systems, therefore,
1528        pcre_malloc etc. are now initialized to local functions that call the
1529        relevant global functions.
1531    7.  There were two entries missing in the vectors called coptable and poptable
1532        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1533        I've fixed the data, and added a kludgy way of testing at compile time that
1534        the lengths are correct (equal to the number of opcodes).
1536    8.  Following on from 7, I added a similar kludge to check the length of the
1537        eint vector in pcreposix.c.
1539    9.  Error texts for pcre_compile() are held as one long string to avoid too
1540        much relocation at load time. To find a text, the string is searched,
1541        counting zeros. There was no check for running off the end of the string,
1542        which could happen if a new error number was added without updating the
1543        string.
1545    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1547    11. \K was not working if it appeared in an atomic group or in a group that
1548        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1549        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1550        the assertion is positive, but not if it is negative.
1552    12. Change 11 fortuitously reduced the size of the stack frame used in the
1553        "match()" function of pcre_exec.c by one pointer. Forthcoming
1554        implementation of support for (*MARK) will need an extra pointer on the
1555        stack; I have reserved it now, so that the stack frame size does not
1556        decrease.
1558    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1559        item in branch that calls a recursion is a subroutine call - as in the
1560        second branch in the above example - was incorrectly given the compile-
1561        time error "recursive call could loop indefinitely" because pcre_compile()
1562        was not correctly checking the subroutine for matching a non-empty string.
1564    14. The checks for overrunning compiling workspace could trigger after an
1565        overrun had occurred. This is a "should never occur" error, but it can be
1566        triggered by pathological patterns such as hundreds of nested parentheses.
1567        The checks now trigger 100 bytes before the end of the workspace.
1569    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1572    Version 8.01 19-Jan-2010
1573    ------------------------
1575    1.  If a pattern contained a conditional subpattern with only one branch (in
1576        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1577        computed the wrong minimum data length (which is of course zero for such
1578        subpatterns). This could cause incorrect "no match" results.
1580    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1581        the pattern is reset in the first branch, pcre_compile() failed with
1582        "internal error: code overflow at offset...". This happened only when
1583        the reset was to the original external option setting. (An optimization
1584        abstracts leading options settings into an external setting, which was the
1585        cause of this.)
1587    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1588        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1589        assertion pattern did not match (meaning that the assertion was true), it
1590        was incorrectly treated as false if the SKIP had been reached during the
1591        matching. This also applied to assertions used as conditions.
1593    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1594        assertion subpattern, including such a pattern used as a condition,
1595        unpredictable results occurred, instead of the error return
1598    5.  The C++ GlobalReplace function was not working like Perl for the special
1599        situation when an empty string is matched. It now does the fancy magic
1600        stuff that is necessary.
1602    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1603        removed. (These were left over from very, very early versions of PCRE.)
1605    7.  Some cosmetic changes to the code to make life easier when compiling it
1606        as part of something else:
1608        (a) Change DEBUG to PCRE_DEBUG.
1610        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1611            called "current" as "current_branch", to prevent a collision with the
1612            Linux macro when compiled as a kernel module.
1614        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1615            prevent a collision with the Linux macro when compiled as a kernel
1616            module.
1618    8.  In pcre_compile() there are some checks for integer overflows that used to
1619        cast potentially large values to (double). This has been changed to that
1620        when building, a check for int64_t is made, and if it is found, it is used
1621        instead, thus avoiding the use of floating point arithmetic. (There is no
1622        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1623        double.
1625    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1626        2005 (difference between two addresses compared to an unsigned value).
1628    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1629        custom one, because of the following reported problem in Windows:
1631          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1632              under Win32.
1633          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1634              therefore missing the function definition.
1635          - The compiler thus generates a "C" signature for the test function.
1636          - The linker fails to find the "C" function.
1637          - PCRE fails to configure if asked to do so against libbz2.
1639    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1640        messages were output:
1642          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1643          rerunning libtoolize, to keep the correct libtool macros in-tree.
1644          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1646        I have done both of these things.
1648    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1649        most of the time, it *can* run out if it is given a pattern that contains a
1650        runaway infinite recursion. I updated the discussion in the pcrestack man
1651        page.
1653    13. Now that we have gone to the x.xx style of version numbers, the minor
1654        version may start with zero. Using 08 or 09 is a bad idea because users
1655        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1656        interpreted as invalid octal numbers. I've updated the previous comment in
1657        configure.ac, and also added a check that gives an error if 08 or 09 are
1658        used.
1660    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1661        causing partial matching to fail when the end of the subject matched \W
1662        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1664    15. There were some discrepancies between the declarations in pcre_internal.h
1665        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1666        their definitions. The declarations used "const uschar *" and the
1667        definitions used USPTR. Even though USPTR is normally defined as "const
1668        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1669        reported that: "This difference in casting confuses some C++ compilers, for
1670        example, SunCC recognizes above declarations as different functions and
1671        generates broken code for hbpcre." I have changed the declarations to use
1672        USPTR.
1674    16. GNU libtool is named differently on some systems. The autogen.sh script now
1675        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1676        (FreeBSD).
1678    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1679        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1680        comment: "Figure out how to create a longlong from a string: strtoll and
1681        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1682        instance, but it only takes 2 args instead of 3!"
1684    18. A subtle bug concerned with back references has been fixed by a change of
1685        specification, with a corresponding code fix. A pattern such as
1686        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1687        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1688        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1689        same bug. Such groups have to be quantified to be useful, or contained
1690        inside another quantified group. (If there's no repetition, the reference
1691        can never match.) The problem arises because, having left the group and
1692        moved on to the rest of the pattern, a later failure that backtracks into
1693        the group uses the captured value from the final iteration of the group
1694        rather than the correct earlier one. I have fixed this in PCRE by forcing
1695        any group that contains a reference to itself to be an atomic group; that
1696        is, there cannot be any backtracking into it once it has completed. This is
1697        similar to recursive and subroutine calls.
1700    Version 8.00 19-Oct-09
1701    ----------------------
1703    1.  The table for translating pcre_compile() error codes into POSIX error codes
1704        was out-of-date, and there was no check on the pcre_compile() error code
1705        being within the table. This could lead to an OK return being given in
1706        error.
1708    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1709        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1710        in a Windows environment.
1712    3.  The pcregrep --count option prints the count for each file even when it is
1713        zero, as does GNU grep. However, pcregrep was also printing all files when
1714        --files-with-matches was added. Now, when both options are given, it prints
1715        counts only for those files that have at least one match. (GNU grep just
1716        prints the file name in this circumstance, but including the count seems
1717        more useful - otherwise, why use --count?) Also ensured that the
1718        combination -clh just lists non-zero counts, with no names.
1720    4.  The long form of the pcregrep -F option was incorrectly implemented as
1721        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1722        but it seems right to fix it, and I didn't think it was worth preserving
1723        the old behaviour.
1725    5.  The command line items --regex=pattern and --regexp=pattern were not
1726        recognized by pcregrep, which required --regex pattern or --regexp pattern
1727        (with a space rather than an '='). The man page documented the '=' forms,
1728        which are compatible with GNU grep; these now work.
1730    6.  No libpcreposix.pc file was created for pkg-config; there was just
1731        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1733    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1734        when UCP support is not needed, by modifying the Python script that
1735        generates it from Unicode data files. This should not matter if the module
1736        is correctly used as a library, but I received one complaint about 50K of
1737        unwanted data. My guess is that the person linked everything into his
1738        program rather than using a library. Anyway, it does no harm.
1740    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1741        was a minimum greater than 1 for a wide character in a possessive
1742        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1743        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1744        character. Chaos in the form of incorrect output or a compiling loop could
1745        result.
1747    9.  The restrictions on what a pattern can contain when partial matching is
1748        requested for pcre_exec() have been removed. All patterns can now be
1749        partially matched by this function. In addition, if there are at least two
1750        slots in the offset vector, the offset of the earliest inspected character
1751        for the match and the offset of the end of the subject are set in them when
1752        PCRE_ERROR_PARTIAL is returned.
1754    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1755        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1756        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1757        and may be more useful for multi-segment matching.
1759    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1760        used to be given if ever the end of the subject was reached; now it is
1761        given only if matching could not proceed because another character was
1762        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1763        string "Z", which now yields "no match" instead of "partial match". In the
1764        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1765        final character ended with (*FAIL).
1767    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1768        if the pattern had a "must contain" character that was already found in the
1769        earlier partial match, unless partial matching was again requested. For
1770        example, with the pattern /dog.(body)?/, the "must contain" character is
1771        "g". If the first part-match was for the string "dog", restarting with
1772        "sbody" failed. This bug has been fixed.
1774    13. The string returned by pcre_dfa_exec() after a partial match has been
1775        changed so that it starts at the first inspected character rather than the
1776        first character of the match. This makes a difference only if the pattern
1777        starts with a lookbehind assertion or \b or \B (\K is not supported by
1778        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1779        matching functions compatible, and I think it's the right thing to do.
1781    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1782        so that the demonstration program is easily available in environments where
1783        PCRE has not been installed from source.
1785    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1786        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1787        library.
1789    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1790        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1791        is not the first non-POSIX option to be added. Clearly some people find
1792        these options useful.
1794    17. If a caller to the POSIX matching function regexec() passes a non-zero
1795        value for nmatch with a NULL value for pmatch, the value of
1796        nmatch is forced to zero.
1798    18. RunGrepTest did not have a test for the availability of the -u option of
1799        the diff command, as RunTest does. It now checks in the same way as
1800        RunTest, and also checks for the -b option.
1802    19. If an odd number of negated classes containing just a single character
1803        interposed, within parentheses, between a forward reference to a named
1804        subpattern and the definition of the subpattern, compilation crashed with
1805        an internal error, complaining that it could not find the referenced
1806        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1807        [The bug was that it was starting one character too far in when skipping
1808        over the character class, thus treating the ] as data rather than
1809        terminating the class. This meant it could skip too much.]
1811    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1812        /g option in pcretest when the pattern contains \K, which makes it possible
1813        to have an empty string match not at the start, even when the pattern is
1814        anchored. Updated pcretest and pcredemo to use this option.
1816    21. If the maximum number of capturing subpatterns in a recursion was greater
1817        than the maximum at the outer level, the higher number was returned, but
1818        with unset values at the outer level. The correct (outer level) value is
1819        now given.
1821    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1822        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1823        make it do so. The string so far is captured, making this feature
1824        compatible with Perl.
1826    23. The tests have been re-organized, adding tests 11 and 12, to make it
1827        possible to check the Perl 5.10 features against Perl 5.10.
1829    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1830        pattern matches a fixed length string. PCRE did not allow this; now it
1831        does. Neither allows recursion.
1833    25. I finally figured out how to implement a request to provide the minimum
1834        length of subject string that was needed in order to match a given pattern.
1835        (It was back references and recursion that I had previously got hung up
1836        on.) This code has now been added to pcre_study(); it finds a lower bound
1837        to the length of subject needed. It is not necessarily the greatest lower
1838        bound, but using it to avoid searching strings that are too short does give
1839        some useful speed-ups. The value is available to calling programs via
1840        pcre_fullinfo().
1842    26. While implementing 25, I discovered to my embarrassment that pcretest had
1843        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1844        study optimizations had never been tested with that matching function.
1845        Oops. What is worse, even when it was passed study data, there was a bug in
1846        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1847        were also very few tests of studied patterns with pcre_dfa_exec().
1849    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1850        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1851        on the other side of the coin, they are no longer allowed to have different
1852        names, because these cannot be distinguished in PCRE, and this has caused
1853        confusion. (This is a difference from Perl.)
1855    28. When duplicate subpattern names are present (necessarily with different
1856        numbers, as required by 27 above), and a test is made by name in a
1857        conditional pattern, either for a subpattern having been matched, or for
1858        recursion in such a pattern, all the associated numbered subpatterns are
1859        tested, and the overall condition is true if the condition is true for any
1860        one of them. This is the way Perl works, and is also more like the way
1861        testing by number works.
1864    Version 7.9 11-Apr-09
1865    ---------------------
1867    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1868        (pcretest), all targets were linked against these libraries. This included
1869        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1870        libraries. This caused unwanted dependencies to be created. This problem
1871        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1872        pcretest is linked with readline.
1874    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1875        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1876        moved outside it again, because FALSE and TRUE are already defined in AIX,
1877        but BOOL is not.
1879    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1880        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1882    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1883        hyphens) following file names and line numbers when outputting matching
1884        lines. This is not true; no spaces are inserted. I have also clarified the
1885        wording for the --colour (or --color) option.
1887    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1888        was not coloured; this is different to GNU grep, so I have changed it to be
1889        the same.
1891    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1892        each matching line was coloured. Now it goes on to look for further matches
1893        of any of the test patterns, which is the same behaviour as GNU grep.
1895    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1896        doesn't make sense to accept an empty string match in pcregrep, so I have
1897        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1898        seems to be how GNU grep behaves. [But see later change 40 for release
1899        8.33.]
1901    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1902        start or after a newline", because the conditional assertion was not being
1903        correctly handled. The rule now is that both the assertion and what follows
1904        in the first alternative must satisfy the test.
1906    9.  If auto-callout was enabled in a pattern with a conditional group whose
1907        condition was an assertion, PCRE could crash during matching, both with
1908        pcre_exec() and pcre_dfa_exec().
1910    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1911        used for matching.
1913    11. Unicode property support in character classes was not working for
1914        characters (bytes) greater than 127 when not in UTF-8 mode.
1916    12. Added the -M command line option to pcretest.
1918    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1920    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1922    16. Added comments and documentation about mis-use of no_arg in the C++
1923        wrapper.
1925    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1926        from Martin Jerabek that uses macro names for all relevant character and
1927        string constants.
1929    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1930        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1931        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1932        these, but not everybody uses configure.
1934    19. A conditional group that had only one branch was not being correctly
1935        recognized as an item that could match an empty string. This meant that an
1936        enclosing group might also not be so recognized, causing infinite looping
1937        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1938        with the subject "ab", where knowledge that the repeated group can match
1939        nothing is needed in order to break the loop.
1941    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1942        exec(), but without supplying a callout function, matching went wrong.
1944    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1945        leak if the size of the offset vector was greater than 30. When the vector
1946        is smaller, the saved offsets during recursion go onto a local stack
1947        vector, but for larger vectors malloc() is used. It was failing to free
1948        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1949        error, in fact).
1951    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1952        heapframe that is used only when UTF-8 support is enabled. This caused no
1953        problem, but was untidy.
1955    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1956        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1957        included within another project.
1959    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1960        slightly modified by me:
1962          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1963              not building pcregrep.
1965          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1966              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1968    25. Forward references, both numeric and by name, in patterns that made use of
1969        duplicate group numbers, could behave incorrectly or give incorrect errors,
1970        because when scanning forward to find the reference group, PCRE was not
1971        taking into account the duplicate group numbers. A pattern such as
1972        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1974    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1975        the feature of a custom pointer more persuasive (as requested by a user).
1977    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1978        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1979        pre-defined.
1981    28. Added support for (*UTF8) at the start of a pattern.
1983    29. Arrange for flags added by the "release type" setting in CMake to be shown
1984        in the configuration summary.
1987    Version 7.8 05-Sep-08
1988    ---------------------
1990    1.  Replaced UCP searching code with optimized version as implemented for Ad
1991        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1992        stage table and inline lookup instead of a function, giving speed ups of 2
1993        to 5 times on some simple patterns that I tested. Permission was given to
1994        distribute the MultiStage2.py script that generates the tables (it's not in
1995        the tarball, but is in the Subversion repository).
1997    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1998        scripts.
2000    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2001        a group with a zero qualifier. The result of the study could be incorrect,
2002        or the function might crash, depending on the pattern.
2004    4.  Caseless matching was not working for non-ASCII characters in back
2005        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2006        It now works when Unicode Property Support is available.
2008    5.  In pcretest, an escape such as \x{de} in the data was always generating
2009        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2010        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2011        truncation.
2013    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2015    7.  Added two (int) casts to pcregrep when printing the difference of two
2016        pointers, in case they are 64-bit values.
2018    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2019        test 2 if it fails.
2021    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2022        and a #define of that name to empty if it is not externally set. This is to
2023        allow users of MSVC to set it if necessary.
2025    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2026        the convenience functions in the pcre_get.c source file.
2028    11. An option change at the start of a pattern that had top-level alternatives
2029        could cause overwriting and/or a crash. This command provoked a crash in
2030        some environments:
2032          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2034        This potential security problem was recorded as CVE-2008-2371.
2036    12. For a pattern where the match had to start at the beginning or immediately
2037        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2038        pcre_dfa_exec() could read past the end of the passed subject if there was
2039        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2040        pcretest so that it places the subject at the end of its malloc-ed buffer.
2042    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2043        exec() might read past the end of the data buffer in UTF-8 mode.
2045    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2046        the data contained the byte 0x85 as part of a UTF-8 character within its
2047        first line. This applied both to normal and DFA matching.
2049    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2050        /^[^d]*?$/8 failed to match "abc".
2052    16. Added a missing copyright notice to pcrecpp_internal.h.
2054    17. Make it more clear in the documentation that values returned from
2055        pcre_exec() in ovector are byte offsets, not character counts.
2057    18. Tidied a few places to stop certain compilers from issuing warnings.
2059    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2060        supplied by Stefan Weber. I made a further small update for 7.8 because
2061        there is a change of source arrangements: the pcre_searchfuncs.c module is
2062        replaced by pcre_ucd.c.
2065    Version 7.7 07-May-08
2066    ---------------------
2068    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2069        a string to a long long, pretend we don't even have a long long." This is
2070        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2072    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2073        pre-7.6 versions, which defined a global no_arg variable instead of putting
2074        it in the RE class. (See also #8 below.)
2076    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2077        Lopes.
2079    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2081        (1) The include/exclude patterns were being applied to the whole pathnames
2082            of files, instead of just to the final components.
2084        (2) If there was more than one level of directory, the subdirectories were
2085            skipped unless they satisfied the include/exclude conditions. This is
2086            inconsistent with GNU grep (and could even be seen as contrary to the
2087            pcregrep specification - which I improved to make it absolutely clear).
2088            The action now is always to scan all levels of directory, and just
2089            apply the include/exclude patterns to regular files.
2091    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2092        --exclude_dir in the tests to avoid scanning .svn directories.
2094    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2095        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2096        doesn't support NULs in patterns.
2098    7.  Added some missing "const"s to declarations of static tables in
2099        pcre_compile.c and pcre_dfa_exec.c.
2101    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2102        caused by fix #2  above. (Subsequently also a second patch to fix the
2103        first patch. And a third patch - this was a messy problem.)
2105    9.  Applied Craig's patch to remove the use of push_back().
2107    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2108        matching function regexec().
2110    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2111        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2112        references. PCRE supports relative numbers with this syntax (I don't think
2113        Oniguruma does).
2115    12. Previously, a group with a zero repeat such as (...){0} was completely
2116        omitted from the compiled regex. However, this means that if the group
2117        was called as a subroutine from elsewhere in the pattern, things went wrong
2118        (an internal error was given). Such groups are now left in the compiled
2119        pattern, with a new opcode that causes them to be skipped at execution
2120        time.
2122    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2123        to the way PCRE behaves:
2125        (a) A lone ] character is dis-allowed (Perl treats it as data).
2127        (b) A back reference to an unmatched subpattern matches an empty string
2128            (Perl fails the current match path).
2130        (c) A data ] in a character class must be notated as \] because if the
2131            first data character in a class is ], it defines an empty class. (In
2132            Perl it is not possible to have an empty class.) The empty class []
2133            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2134            The negative empty class [^] matches any one character, independently
2135            of the DOTALL setting.
2137    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2138        non-existent subpattern following a character class starting with ']' and
2139        containing () gave an internal compiling error instead of "reference to
2140        non-existent subpattern". Fortunately, when the pattern did exist, the
2141        compiled code was correct. (When scanning forwards to check for the
2142        existence of the subpattern, it was treating the data ']' as terminating
2143        the class, so got the count wrong. When actually compiling, the reference
2144        was subsequently set up correctly.)
2146    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2147        it was being rejected as not supported by pcre_dfa_exec(), even though
2148        other assertions are supported. I have made pcre_dfa_exec() support
2149        (*FAIL).
2151    16. The implementation of 13c above involved the invention of a new opcode,
2152        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2153        cannot be changed at match time, I realized I could make a small
2154        improvement to matching performance by compiling OP_ALLANY instead of
2155        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2156        on the OP_ANY path.
2158    17. Compiling pcretest on Windows with readline support failed without the
2159        following two fixes: (1) Make the unistd.h include conditional on
2160        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2162    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2163        ncurses library to be included for pcretest when ReadLine support is
2164        requested, but also to allow for it to be overridden. This patch came from
2165        Daniel Bergström.
2167    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2168        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2169        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2170        spotting this.
2173    Version 7.6 28-Jan-08
2174    ---------------------
2176    1.  A character class containing a very large number of characters with
2177        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2178        overflow.
2180    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2181        HAVE_LONG_LONG is not defined.
2183    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2184        bring them up to date and include new features. This patch includes:
2186        - Fixed PH's badly added libz and libbz2 support.
2187        - Fixed a problem with static linking.
2188        - Added pcredemo. [But later removed - see 7 below.]
2189        - Fixed dftables problem and added an option.
2190        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2191            HAVE_LONG_LONG.
2192        - Added readline support for pcretest.
2193        - Added an listing of the option settings after cmake has run.
2195    4.  A user submitted a patch to Makefile that makes it easy to create
2196        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2197        Makefile.am that cause it to include this special target, without
2198        affecting anything else. Note that the same mingw target plus all
2199        the other distribution libraries and programs are now supported
2200        when configuring with CMake (see 6 below) instead of with
2201        Configure/Make.
2203    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2204        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2205        exported in the Windows port". It has not yet been confirmed that the patch
2206        solves the problem, but it does no harm.
2208    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2209        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2210        with CMake, and also correct the comment about stack recursion.
2212    7.  Remove the automatic building of pcredemo from the ./configure system and
2213        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2214        of a program that users should build themselves after PCRE is installed, so
2215        building it automatically is not really right. What is more, it gave
2216        trouble in some build environments.
2218    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2221    Version 7.5 10-Jan-08
2222    ---------------------
2224    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2225        values in parens when parsing an RE using the C++ wrapper."
2227    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2228        Characters greater than 255 were excluded from the class instead of being
2229        included.
2231    3.  The same bug as (2) above applied to negated POSIX classes such as
2232        [:^space:].
2234    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2235        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2236        I have changed it.
2238    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2239        first named subpattern) and a construct such as (?&a) would reference the
2240        first named subpattern whose name started with "a" (in other words, the
2241        length check was missing). Both these problems are fixed. "Subpattern name
2242        expected" is now given for (?&) (a zero-length name), and this patch also
2243        makes it give the same error for \k'' (previously it complained that that
2244        was a reference to a non-existent subpattern).
2246    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2247        this is right because (?- can be followed by option settings as well as by
2248        digits. I have, however, made the messages clearer.
2250    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2251        than the number used in the conditional) now cause a compile-time error.
2252        This is actually not compatible with Perl, which accepts such patterns, but
2253        treats the conditional as always being FALSE (as PCRE used to), but it
2254        seems to me that giving a diagnostic is better.
2256    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2257        and messages.
2259    9.  Fix two occurrences of "backslash" in comments that should have been
2260        "backspace".
2262    10. Remove two redundant lines of code that can never be obeyed (their function
2263        was moved elsewhere).
2265    11. The program that makes PCRE's Unicode character property table had a bug
2266        which caused it to generate incorrect table entries for sequences of
2267        characters that have the same character type, but are in different scripts.
2268        It amalgamated them into a single range, with the script of the first of
2269        them. In other words, some characters were in the wrong script. There were
2270        thirteen such cases, affecting characters in the following ranges:
2272          U+002b0 - U+002c1
2273          U+0060c - U+0060d
2274          U+0061e - U+00612
2275          U+0064b - U+0065e
2276          U+0074d - U+0076d
2277          U+01800 - U+01805
2278          U+01d00 - U+01d77
2279          U+01d9b - U+01dbf
2280          U+0200b - U+0200f
2281          U+030fc - U+030fe
2282          U+03260 - U+0327f
2283          U+0fb46 - U+0fbb1
2284          U+10450 - U+1049d
2286    12. The -o option (show only the matching part of a line) for pcregrep was not
2287        compatible with GNU grep in that, if there was more than one match in a
2288        line, it showed only the first of them. It now behaves in the same way as
2289        GNU grep.
2291    13. If the -o and -v options were combined for pcregrep, it printed a blank
2292        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2293        does the same. The return code can be used to tell if there were any
2294        non-matching lines.
2296    14. Added --file-offsets and --line-offsets to pcregrep.
2298    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2299        infinitely looping recursion. The bug was that positive lookaheads were not
2300        being skipped when checking for a possible empty match (negative lookaheads
2301        and both kinds of lookbehind were skipped).
2303    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2304        inclusion of <windows.h> to before rather than after the definition of
2305        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2307    17. Specifying a possessive quantifier with a specific limit for a Unicode
2308        character property caused pcre_compile() to compile bad code, which led at
2309        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2310        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2311        caused the error; without that there was no problem.
2313    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2315    19. Added --enable-pcretest-libreadline.
2317    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2318        RE::GlobalReplace(). As a result, the number of replacements returned was
2319        double what it should be. I removed one of the increments, but Craig sent a
2320        later patch that removed the other one (the right fix) and added unit tests
2321        that check the return values (which was not done before).
2323    21. Several CMake things:
2325        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2326            the names libpcre and libpcreposix, not just pcre and pcreposix.
2328        (2) The above change means that pcretest and pcregrep are now correctly
2329            linked with the newly-built libraries, not previously installed ones.
2333    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2334        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2335        UTF-8 newline character). The key issue is that the pattern starts .*;
2336        this means that the match must be either at the beginning, or after a
2337        newline. The bug was in the code for advancing after a failed match and
2338        checking that the new position followed a newline. It was not taking
2339        account of UTF-8 characters correctly.
2341    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2342        character classes. PCRE was not treating the sequence [:...:] as a
2343        character class unless the ... were all letters. Perl, however, seems to
2344        allow any characters between [: and :], though of course it rejects as
2345        unknown any "names" that contain non-letters, because all the known class
2346        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2347        for example, whereas PCRE did not - it did not recognize a POSIX character
2348        class. This seemed a bit dangerous, so the code has been changed to be
2349        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2350        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2351        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2352        Perl does, and where it didn't before.
2354    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2355        Windows environments %n is disabled by default.
2358    Version 7.4 21-Sep-07
2359    ---------------------
2361    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2362        means that a class such as [\s] counted as "explicit reference to CR or
2363        LF". That isn't really right - the whole point of the change was to try to
2364        help when there was an actual mention of one of the two characters. So now
2365        the change happens only if \r or \n (or a literal CR or LF) character is
2366        encountered.
2368    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2369        of both had grown to the point where there were only 3 bits left.
2370        Fortunately, there was spare space in the data structure, and so I have
2371        moved the internal flags into a new 16-bit field to free up more option
2372        bits.
2374    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2375        but did not set the internal JCHANGED flag - either of these is enough to
2376        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2377        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2378        start sets both bits.
2380    4.  Added options (at build time, compile time, exec time) to change \R from
2381        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2383    5.  doc/pcresyntax.html was missing from the distribution.
2385    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2386        compatibility, even though it is no longer used.
2388    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2389        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2390        windows.h file is present (where different names are used). [This was
2391        reversed later after testing - see 16 below.]
2393    8.  Changed all #include <config.h> to #include "config.h". There were also
2394        some further <pcre.h> cases that I changed to "pcre.h".
2396    9.  When pcregrep was used with the --colour option, it missed the line ending
2397        sequence off the lines that it output.
2399    10. It was pointed out to me that arrays of string pointers cause lots of
2400        relocations when a shared library is dynamically loaded. A technique of
2401        using a single long string with a table of offsets can drastically reduce
2402        these. I have refactored PCRE in four places to do this. The result is
2403        dramatic:
2405          Originally:                          290
2406          After changing UCP table:            187
2407          After changing error message table:   43
2408          After changing table of "verbs"       36
2409          After changing table of Posix names   22
2411        Thanks to the folks working on Gregex for glib for this insight.
2413    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2414        unicode-properties was also set.
2416    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2418    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2419        checked only for CRLF.
2421    14. Added casts to pcretest.c to avoid compiler warnings.
2423    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2425    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2426        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2427        entirely. This removes changes made in 7 above.
2429    17. The CMake files have been updated, and there is now more information about
2430        building with CMake in the NON-UNIX-USE document.
2433    Version 7.3 28-Aug-07
2434    ---------------------
2436     1. In the rejigging of the build system that eventually resulted in 7.1, the
2437        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2438        brackets there is not right, since it causes compilers to look for an
2439        installed pcre.h, not the version that is in the source that is being
2440        compiled (which of course may be different). I have changed it back to:
2442          #include "pcre.h"
2444        I have a vague recollection that the change was concerned with compiling in
2445        different directories, but in the new build system, that is taken care of
2446        by the VPATH setting the Makefile.
2448     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2449        when the subject happened to end in the byte 0x85 (e.g. if the last
2450        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2451        characters but of course it shouldn't be taken as a newline when it is part
2452        of another character. The bug was that, for an unlimited repeat of . in
2453        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2454        characters when looking for a newline.
2456     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2458     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2459        in debug output.
2461     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2462        long printing in the pcrecpp unittest when running under MinGW.
2464     6. ESC_K was left out of the EBCDIC table.
2466     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2467        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2468        limit also applies to "virtual nesting" when a pattern is recursive, and in
2469        this case 1000 isn't so big. I have been able to remove this limit at the
2470        expense of backing off one optimization in certain circumstances. Normally,
2471        when pcre_exec() would call its internal match() function recursively and
2472        immediately return the result unconditionally, it uses a "tail recursion"
2473        feature to save stack. However, when a subpattern that can match an empty
2474        string has an unlimited repetition quantifier, it no longer makes this
2475        optimization. That gives it a stack frame in which to save the data for
2476        checking that an empty string has been matched. Previously this was taken
2477        from the 1000-entry workspace that had been reserved. So now there is no
2478        explicit limit, but more stack is used.
2480     8. Applied Daniel's patches to solve problems with the import/export magic
2481        syntax that is required for Windows, and which was going wrong for the
2482        pcreposix and pcrecpp parts of the library. These were overlooked when this
2483        problem was solved for the main library.
2485     9. There were some crude static tests to avoid integer overflow when computing
2486        the size of patterns that contain repeated groups with explicit upper
2487        limits. As the maximum quantifier is 65535, the maximum group length was
2488        set at 30,000 so that the product of these two numbers did not overflow a
2489        32-bit integer. However, it turns out that people want to use groups that
2490        are longer than 30,000 bytes (though not repeat them that many times).
2491        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2492        made it possible to implement the integer overflow checks in a much more
2493        dynamic way, which I have now done. The artificial limitation on group
2494        length has been removed - we now have only the limit on the total length of
2495        the compiled pattern, which depends on the LINK_SIZE setting.
2497    10. Fixed a bug in the documentation for get/copy named substring when
2498        duplicate names are permitted. If none of the named substrings are set, the
2499        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2500        empty string.
2502    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2503        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2504        because the ] is interpreted as the first data character and the
2505        terminating ] is not found. PCRE has been made compatible with Perl in this
2506        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2507        cause memory overwriting.
2509    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2510        string has been matched (to stop an infinite loop). It was not recognizing
2511        a conditional subpattern that could match an empty string if that
2512        subpattern was within another subpattern. For example, it looped when
2513        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2514        condition was not nested. This bug has been fixed.
2516    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2517        past the start of the subject in the presence of bytes with the top bit
2518        set, for example "\x8aBCD".
2520    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2521        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2523    14. Optimized (?!) to (*FAIL).
2525    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2526        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2527        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2528        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2529        does: it's just the validity check that is more restrictive.
2531    16. Inserted checks for integer overflows during escape sequence (backslash)
2532        processing, and also fixed erroneous offset values for syntax errors during
2533        backslash processing.
2535    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2536        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2538    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2539        caused an overrun.
2541    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2542        something other than just ASCII characters) inside a group that had an
2543        unlimited repeat caused a loop at compile time (while checking to see
2544        whether the group could match an empty string).
2546    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2547        [\P{Any}] did so. (Error in the code for printing property names.)
2549    21. An orphan \E inside a character class could cause a crash.
2551    22. A repeated capturing bracket such as (A)? could cause a wild memory
2552        reference during compilation.
2554    23. There are several functions in pcre_compile() that scan along a compiled
2555        expression for various reasons (e.g. to see if it's fixed length for look
2556        behind). There were bugs in these functions when a repeated \p or \P was
2557        present in the pattern. These operators have additional parameters compared
2558        with \d, etc, and these were not being taken into account when moving along
2559        the compiled data. Specifically:
2561        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2562            length.
2564        (b) An item such as \pL+ within a repeated group could cause crashes or
2565            loops.
2567        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2568            "reference to non-existent subpattern" error.
2570        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2572    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2573        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2575    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2577    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2578        character were causing crashes (broken optimization).
2580    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2581        \p or \P) caused a compile-time loop.
2583    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2584        break. For example, the unstudied pattern [\r\n]A does not match the string
2585        "\r\nA" because change 7.0/46 below moves the current point on by two
2586        characters after failing to match at the start. However, the pattern \nA
2587        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2588        the same is true. There doesn't seem any very clean way out of this, but
2589        what I have chosen to do makes the common cases work: PCRE now takes note
2590        of whether there can be an explicit match for \r or \n anywhere in the
2591        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2592        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2593        pattern has explicit CR or LF references.
2595    29. Added (*CR) etc for changing newline setting at start of pattern.
2598    Version 7.2 19-Jun-07
2599  ---------------------  ---------------------
2601   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
# Line 21  Version 7.2 01-May-07 Line 2615  Version 7.2 01-May-07
2615      stack recursion. This gives a massive performance boost under BSD, but just      stack recursion. This gives a massive performance boost under BSD, but just
2616      a small improvement under Linux. However, it saves one field in the frame      a small improvement under Linux. However, it saves one field in the frame
2617      in all cases.      in all cases.
2619   6. Added more features from the forthcoming Perl 5.10:   6. Added more features from the forthcoming Perl 5.10:
2621      (a) (?-n) (where n is a string of digits) is a relative subroutine or      (a) (?-n) (where n is a string of digits) is a relative subroutine or
2622          recursion call. It refers to the nth most recently opened parentheses.          recursion call. It refers to the nth most recently opened parentheses.
2624      (b) (?+n) is also a relative subroutine call; it refers to the nth next      (b) (?+n) is also a relative subroutine call; it refers to the nth next
2625          to be opened parentheses.          to be opened parentheses.
2627      (c) Conditions that refer to capturing parentheses can be specified      (c) Conditions that refer to capturing parentheses can be specified
2628          relatively, for example, (?(-2)... or (?(+3)...          relatively, for example, (?(-2)... or (?(+3)...
2630        (d) \K resets the start of the current match so that everything before
2631            is not part of it.
2633        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2635        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2636            reference syntax.
2638        (g) (?| introduces a group in which the numbering of parentheses in each
2639            alternative starts with the same number.
2641        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2643     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2646     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2647        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2648        for detecting groups that can match an empty string.
2650     9. A pattern with a very large number of alternatives (more than several
2651        hundred) was running out of internal workspace during the pre-compile
2652        phase, where pcre_compile() figures out how much memory will be needed. A
2653        bit of new cunning has reduced the workspace needed for groups with
2654        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2655        workspace instead of running out of the 4096 that are available.
2657    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2659    11. Applied patch from Google to remove an optimization that didn't quite work.
2660        The report of the bug said:
2662          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2663          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2664          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2666    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2667        it matched the wrong number of bytes.
2670  Version 7.1 24-Apr-07  Version 7.1 24-Apr-07
2671  ---------------------  ---------------------

Removed from v.167  
changed lines
  Added in v.1415

  ViewVC Help
Powered by ViewVC 1.1.5