ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 270 by ph10, Fri Nov 16 19:55:16 2007 UTC revision 1518 by ph10, Fri Feb 6 16:10:27 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4  Version 7.5 12-Nov-07  Version 8.37 xx-xxx-2015
5    ------------------------
7    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
8        for those parentheses to be closed with whatever has been captured so far.
9        However, it was failing to mark any other groups between the hightest
10        capture so far and the currrent group as "unset". Thus, the ovector for
11        those groups contained whatever was previously there. An example is the
12        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
14    2.  If an assertion condition was quantified with a minimum of zero (an odd
15        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
17    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
18        unrecognized modifier, a crash could occur.
20    4.  An attempt to do global matching in pcretest with a zero-length ovector
21        caused a crash.
23    5.  Fixed a memory leak during matching that could occur for a subpattern
24        subroutine call (recursive or otherwise) if the number of captured groups
25        that had to be saved was greater than ten.
27    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
28        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
29        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
32    Version 8.36 26-September-2014
33    ------------------------------
35    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
36        -Wmissing-field-initializers and -Wunused-parameter.
38    2.  The tests for quantifiers being too big (greater than 65535) were being
39        applied after reading the number, and stupidly assuming that integer
40        overflow would give a negative number. The tests are now applied as the
41        numbers are read.
43    3.  Tidy code in pcre_exec.c where two branches that used to be different are
44        now the same.
46    4.  The JIT compiler did not generate match limit checks for certain
47        bracketed expressions with quantifiers. This may lead to exponential
48        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
49        issue should be resolved now.
51    5.  Fixed an issue, which occures when nested alternatives are optimized
52        with table jumps.
54    6.  Inserted two casts and changed some ints to size_t in the light of some
55        reported 64-bit compiler warnings (Bugzilla 1477).
57    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
58        an empty string, which sometimes were behaving incorrectly in the
59        interpreter (though correctly in the JIT matcher). This pcretest input is
60        an example:
62          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
65        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
66        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
67        for an empty string was breaking the inner loop and carrying on at a lower
68        level, when possessive repeated groups should always return to a higher
69        level as they have no backtrack points in them. The empty string test now
70        occurs at the outer level.
72    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
73        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
75    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
76        doesn't).
78    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
79        to hold a pattern to be incorrectly computed (too small) when there were
80        named back references to duplicated names. This could cause "internal
81        error: code overflow" or "double free or corruption" or other memory
82        handling errors.
84    11. When named subpatterns had the same prefixes, back references could be
85        confused. For example, in this pattern:
87          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
89        the reference to 'Name' was incorrectly treated as a reference to a
90        duplicate name.
92    12. A pattern such as /^s?c/mi8 where the optional character has more than
93        one "other case" was incorrectly compiled such that it would only try to
94        match starting at "c".
96    13. When a pattern starting with \s was studied, VT was not included in the
97        list of possible starting characters; this should have been part of the
98        8.34/18 patch.
100    14. If a character class started [\Qx]... where x is any character, the class
101        was incorrectly terminated at the ].
103    15. If a pattern that started with a caseless match for a character with more
104        than one "other case" was studied, PCRE did not set up the starting code
105        unit bit map for the list of possible characters. Now it does. This is an
106        optimization improvement, not a bug fix.
108    16. The Unicode data tables have been updated to Unicode 7.0.0.
110    17. Fixed a number of memory leaks in pcregrep.
112    18. Avoid a compiler warning (from some compilers) for a function call with
113        a cast that removes "const" from an lvalue by using an intermediate
114        variable (to which the compiler does not object).
116    19. Incorrect code was compiled if a group that contained an internal recursive
117        back reference was optional (had quantifier with a minimum of zero). This
118        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
119        caused segmentation faults because of stack overflows at compile time.
121    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
122        group that is quantified with an indefinite repeat, caused a compile-time
123        loop which used up all the system stack and provoked a segmentation fault.
124        This was not the same bug as 19 above.
126    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
127        Patch by Mike Frysinger.
130    Version 8.35 04-April-2014
131    --------------------------
133    1.  A new flag is set, when property checks are present in an XCLASS.
134        When this flag is not set, PCRE can perform certain optimizations
135        such as studying these XCLASS-es.
137    2.  The auto-possessification of character sets were improved: a normal
138        and an extended character set can be compared now. Furthermore
139        the JIT compiler optimizes more character set checks.
141    3.  Got rid of some compiler warnings for potentially uninitialized variables
142        that show up only when compiled with -O2.
144    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
145        of a match later then the end of the match. The pcretest program was not
146        handling the case sensibly - it was outputting from the start to the next
147        binary zero. It now reports this situation in a message, and outputs the
148        text from the end to the start.
150    5.  Fast forward search is improved in JIT. Instead of the first three
151        characters, any three characters with fixed position can be searched.
152        Search order: first, last, middle.
154    6.  Improve character range checks in JIT. Characters are read by an inprecise
155        function now, which returns with an unknown value if the character code is
156        above a certain threshold (e.g: 256). The only limitation is that the value
157        must be bigger than the threshold as well. This function is useful when
158        the characters above the threshold are handled in the same way.
160    7.  The macros whose names start with RAWUCHAR are placeholders for a future
161        mode in which only the bottom 21 bits of 32-bit data items are used. To
162        make this more memorable for those maintaining the code, the names have
163        been changed to start with UCHAR21, and an extensive comment has been added
164        to their definition.
166    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
167        to the export list in Makefile.am (they were accidentally omitted from the
168        8.34 tarball).
170    9.  The informational output from pcretest used the phrase "starting byte set"
171        which is inappropriate for the 16-bit and 32-bit libraries. As the output
172        for "first char" and "need char" really means "non-UTF-char", I've changed
173        "byte" to "char", and slightly reworded the output. The documentation about
174        these values has also been (I hope) clarified.
176    10. Another JIT related optimization: use table jumps for selecting the correct
177        backtracking path, when more than four alternatives are present inside a
178        bracket.
180    11. Empty match is not possible, when the minimum length is greater than zero,
181        and there is no \K in the pattern. JIT should avoid empty match checks in
182        such cases.
184    12. In a caseless character class with UCP support, when a character with more
185        than one alternative case was not the first character of a range, not all
186        the alternative cases were added to the class. For example, s and \x{17f}
187        are both alternative cases for S: the class [RST] was handled correctly,
188        but [R-T] was not.
190    13. The configure.ac file always checked for pthread support when JIT was
191        enabled. This is not used in Windows, so I have put this test inside a
192        check for the presence of windows.h (which was already tested for).
194    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
195        The algorithm provides a way to skip certain starting offsets, and usually
196        faster than linear prefix searches.
198    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
199        as for 'fr_FR' and 'french'. For some reason, however, it then used the
200        Windows-specific input and output files, which have 'french' screwed in.
201        So this could never have worked. One of the problems with locales is that
202        they aren't always the same. I have now updated RunTest so that it checks
203        the output of the locale test (test 3) against three different output
204        files, and it allows the test to pass if any one of them matches. With luck
205        this should make the test pass on some versions of Solaris where it was
206        failing. Because of the uncertainty, the script did not used to stop if
207        test 3 failed; it now does. If further versions of a French locale ever
208        come to light, they can now easily be added.
210    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
211        there was a message during ./configure, but it did not stop. This now
212        provokes an error. The invalid example in README has been corrected.
213        If a value less than the minimum is given, the minimum value has always
214        been used, but now a warning is given.
216    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
217        was a bug in the test system, which is now fixed. Also, the list of various
218        configurations that are tested for each release did not have one with both
219        16/32 bits and --enable-bar-anycrlf. It now does.
221    18. pcretest was missing "-C bsr" for displaying the \R default setting.
223    19. Little endian PowerPC systems are supported now by the JIT compiler.
225    20. The fast forward newline mechanism could enter to an infinite loop on
226        certain invalid UTF-8 input. Although we don't support these cases
227        this issue can be fixed by a performance optimization.
229    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
230        not take account if existing stack usage. There is now a new global
231        variable called pcre_stack_guard that can be set to point to an external
232        function to check stack availability. It is called at the start of
233        processing every parenthesized group.
235    22. A typo in the code meant that in ungreedy mode the max/min qualifier
236        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
237        not match "ab".
239    23. When UTF was disabled, the JIT program reported some incorrect compile
240        errors. These messages are silenced now.
242    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
243        compiler.
245    25. Change all the temporary files used in RunGrepTest to be different to those
246        used by RunTest so that the tests can be run simultaneously, for example by
247        "make -j check".
250    Version 8.34 15-December-2013
251    -----------------------------
253    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
254        executable memory. Patch inspired by Carsten Klein.
256    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
257        this macro is never tested and has no effect, because the work to support
258        coverage involves only compiling and linking options and special targets in
259        the Makefile. The comment in config.h implied that defining the macro would
260        enable coverage support, which is totally false. There was also support for
261        setting this macro in the CMake files (my fault, I just copied it from
262        configure). SUPPORT_GCOV has now been removed.
264    3.  Make a small performance improvement in strlen16() and strlen32() in
265        pcretest.
267    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
268        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
269        these cases). There was also one in pcretest.c.
271    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
273    6.  In UTF mode, the code for checking whether a group could match an empty
274        string (which is used for indefinitely repeated groups to allow for
275        breaking an infinite loop) was broken when the group contained a repeated
276        negated single-character class with a character that occupied more than one
277        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
278        UTF-8 mode). The effect was undefined: the group might or might not be
279        deemed as matching an empty string, or the program might have crashed.
281    7.  The code for checking whether a group could match an empty string was not
282        recognizing that \h, \H, \v, \V, and \R must match a character.
284    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
285        an empty string. If it can, pcretest shows this in its information output.
287    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
288        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
289        matched by pcre_exec() without using JIT:
291        (a) If the rest of the pattern did not match after a maximal run of
292            grapheme clusters, the code for backing up to try with fewer of them
293            did not always back up over a full grapheme when characters that do not
294            have the modifier quality were involved, e.g. Hangul syllables.
296        (b) If the match point in a subject started with modifier character, and
297            there was no match, the code could incorrectly back up beyond the match
298            point, and potentially beyond the first character in the subject,
299            leading to a segfault or an incorrect match result.
301    10. A conditional group with an assertion condition could lead to PCRE
302        recording an incorrect first data item for a match if no other first data
303        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
304        first data item, and therefore matched "ca" after "c" instead of at the
305        start.
307    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
308        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
310    12. The source of pcregrep now includes z/OS-specific code so that it can be
311        compiled for z/OS as part of the special z/OS distribution.
313    13. Added the -T and -TM options to pcretest.
315    14. The code in pcre_compile.c for creating the table of named capturing groups
316        has been refactored. Instead of creating the table dynamically during the
317        actual compiling pass, the information is remembered during the pre-compile
318        pass (on the stack unless there are more than 20 named groups, in which
319        case malloc() is used) and the whole table is created before the actual
320        compile happens. This has simplified the code (it is now nearly 150 lines
321        shorter) and prepared the way for better handling of references to groups
322        with duplicate names.
324    15. A back reference to a named subpattern when there is more than one of the
325        same name now checks them in the order in which they appear in the pattern.
326        The first one that is set is used for the reference. Previously only the
327        first one was inspected. This change makes PCRE more compatible with Perl.
329    16. Unicode character properties were updated from Unicode 6.3.0.
331    17. The compile-time code for auto-possessification has been refactored, based
332        on a patch by Zoltan Herczeg. It now happens after instead of during
333        compilation. The code is cleaner, and more cases are handled. The option
334        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
335        options in pcretest are provided to set it. It can also be set by
336        (*NO_AUTO_POSSESS) at the start of a pattern.
338    18. The character VT has been added to the default ("C" locale) set of
339        characters that match \s and are generally treated as white space,
340        following this same change in Perl 5.18. There is now no difference between
341        "Perl space" and "POSIX space". Whether VT is treated as white space in
342        other locales depends on the locale.
344    19. The code for checking named groups as conditions, either for being set or
345        for being recursed, has been refactored (this is related to 14 and 15
346        above). Processing unduplicated named groups should now be as fast at
347        numerical groups, and processing duplicated groups should be faster than
348        before.
350    20. Two patches to the CMake build system, by Alexander Barkov:
352          (1) Replace the "source" command by "." in CMakeLists.txt because
353              "source" is a bash-ism.
355          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
356              without these the CMake build does not work on Solaris.
358    21. Perl has changed its handling of \8 and \9. If there is no previously
359        encountered capturing group of those numbers, they are treated as the
360        literal characters 8 and 9 instead of a binary zero followed by the
361        literals. PCRE now does the same.
363    22. Following Perl, added \o{} to specify codepoints in octal, making it
364        possible to specify values greater than 0777 and also making them
365        unambiguous.
367    23. Perl now gives an error for missing closing braces after \x{... instead of
368        treating the string as literal. PCRE now does the same.
370    24. RunTest used to grumble if an inappropriate test was selected explicitly,
371        but just skip it when running all tests. This make it awkward to run ranges
372        of tests when one of them was inappropriate. Now it just skips any
373        inappropriate tests, as it always did when running all tests.
375    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
376        character types such as \d or \w, too many callouts were inserted, and the
377        data that they returned was rubbish.
379    26. In UCP mode, \s was not matching two of the characters that Perl matches,
380        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
381        were matched by \h. The code has now been refactored so that the lists of
382        the horizontal and vertical whitespace characters used for \h and \v (which
383        are defined only in one place) are now also used for \s.
385    27. Add JIT support for the 64 bit TileGX architecture.
386        Patch by Jiong Wang (Tilera Corporation).
388    28. Possessive quantifiers for classes (both explicit and automatically
389        generated) now use special opcodes instead of wrapping in ONCE brackets.
391    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
392        (because it's meaningless), this was not happening when PCRE_CASELESS was
393        set. Not wrong, but inefficient.
395    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
396        properties for \w, \d, etc) is present in a test regex. Otherwise if the
397        test contains no characters greater than 255, Perl doesn't realise it
398        should be using Unicode semantics.
400    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
401        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
402        does in Unicode mode.
404    32. Added the "forbid" facility to pcretest so that putting tests into the
405        wrong test files can sometimes be quickly detected.
407    33. There is now a limit (default 250) on the depth of nesting of parentheses.
408        This limit is imposed to control the amount of system stack used at compile
409        time. It can be changed at build time by --with-parens-nest-limit=xxx or
410        the equivalent in CMake.
412    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
413        errors. Perl warns for these when in warning mode, but PCRE has no facility
414        for giving warnings.
416    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
417        However, this was not working for (?!) because it is optimized to (*FAIL),
418        for which PCRE does not allow quantifiers. The optimization is now disabled
419        when a quantifier follows (?!). I can't see any use for this, but it makes
420        things uniform.
422    36. Perl no longer allows group names to start with digits, so I have made this
423        change also in PCRE. It simplifies the code a bit.
425    37. In extended mode, Perl ignores spaces before a + that indicates a
426        possessive quantifier. PCRE allowed a space before the quantifier, but not
427        before the possessive +. It now does.
429    38. The use of \K (reset reported match start) within a repeated possessive
430        group such as (a\Kb)*+ was not working.
432    40. Document that the same character tables must be used at compile time and
433        run time, and that the facility to pass tables to pcre_exec() and
434        pcre_dfa_exec() is for use only with saved/restored patterns.
436    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
437        features for Builds with MSVC:
439        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
440           builds). The .rc files can be used to set FileDescription and many other
441           attributes.
443        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
444           This allows higher-level build scripts which want .pdb files to avoid
445           hard-coding the exact files needed."
447    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
448        mean "start of word" and "end of word", respectively, as a transition aid.
450    43. A minimizing repeat of a class containing codepoints greater than 255 in
451        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
452        compiled to use the heap for recursion.
454    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
455        is configured.
458    Version 8.33 28-May-2013
459    ------------------------
461    1.  Added 'U' to some constants that are compared to unsigned integers, to
462        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
463        variables that are added to signed variables, to ensure the result is
464        signed and can be negated.
466    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
467        CMake config files.
469    3.  Revise the creation of config.h.generic so that all boolean macros are
470        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
471        overriding via -D on the command line possible.
473    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
474        to unsigned int is reported to make a quite noticeable speed difference in
475        a specific Windows environment. Testing on Linux did also appear to show
476        some benefit (and it is clearly not harmful). Also fixed the definition of
477        Xop which should be unsigned.
479    5.  Related to (4), changing the definition of the intermediate variable cc
480        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
481        improvements.
483    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
484        unnecessary spaces.
486    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
487        and later.
489    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
491    9.  Optimizing fast_forward_start_bits in JIT.
493    10. Adding support for callouts in JIT, and fixing some issues revealed
494        during this work. Namely:
496        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
498        (b) Minimum length was not checked before the matching is started.
500    11. The value of capture_last that is passed to callouts was incorrect in some
501        cases when there was a capture on one path that was subsequently abandoned
502        after a backtrack. Also, the capture_last value is now reset after a
503        recursion, since all captures are also reset in this case.
505    12. The interpreter no longer returns the "too many substrings" error in the
506        case when an overflowing capture is in a branch that is subsequently
507        abandoned after a backtrack.
509    13. In the pathological case when an offset vector of size 2 is used, pcretest
510        now prints out the matched string after a yield of 0 or 1.
512    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
513        Only supported by the JIT compiler at the moment.
515    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
517    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
518        offset of the starting point of the matching process, provided the offsets
519        vector is large enough.
521    17. The \A escape now records a lookbehind value of 1, though its execution
522        does not actually inspect the previous character. This is to ensure that,
523        in partial multi-segment matching, at least one character from the old
524        segment is retained when a new segment is processed. Otherwise, if there
525        are no lookbehinds in the pattern, \A might match incorrectly at the start
526        of a new segment.
528    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
530    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
531        gives some modest performance improvement in 8-bit mode.
533    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
534        be expressed in certain programming languages using Universal Character
535        Names.
537    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
538        which points out that "non characters" are not "characters that may not
539        appear in Unicode strings" but rather "characters that are reserved for
540        internal use and have only local meaning".
542    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
543        there was a conditional group that depended on an assertion, if the
544        assertion was false, the callout that immediately followed the alternation
545        in the condition was skipped when pcre_exec() was used for matching.
547    23. Allow an explicit callout to be inserted before an assertion that is the
548        condition for a conditional group, for compatibility with automatic
549        callouts, which always insert a callout at this point.
551    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
552        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
554    25. (*PRUNE) is now supported by the JIT compiler.
556    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
558    27. Fix the case where there are two or more SKIPs with arguments that may be
559        ignored.
561    28. (*SKIP) is now supported by the JIT compiler.
563    29. (*THEN) is now supported by the JIT compiler.
565    30. Update RunTest with additional test selector options.
567    31. The way PCRE handles backtracking verbs has been changed in two ways.
569        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
570        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
571        backtracking. In some cases this makes it more Perl-compatible, but Perl's
572        rather obscure rules do not always do the same thing.
574        (2) Previously, backtracking verbs were confined within assertions. This is
575        no longer the case for positive assertions, except for (*ACCEPT). Again,
576        this sometimes improves Perl compatibility, and sometimes does not.
578    32. A number of tests that were in test 2 because Perl did things differently
579        have been moved to test 1, because either Perl or PCRE has changed, and
580        these tests are now compatible.
582    32. Backtracking control verbs are now handled in the same way in JIT and
583        interpreter.
585    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
586        contained a forward subroutine reference caused a compile error.
588    34. Auto-detect and optimize limited repetitions in JIT.
590    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
591        blocking (*UTF) etc.
593    36. In the interpreter, maximizing pattern repetitions for characters and
594        character types now use tail recursion, which reduces stack usage.
596    37. The value of the max lookbehind was not correctly preserved if a compiled
597        and saved regex was reloaded on a host of different endianness.
599    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
600        of the compiled pattern block, expand the flags field from 16 to 32 bits
601        because it was almost full.
603    39. Try madvise first before posix_madvise.
605    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
606        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
607        original change locked out all matches of empty strings. This has been
608        changed so that one match of an empty string per line is recognized.
609        Subsequent searches on the same line (for colouring or for --only-matching,
610        for example) do not recognize empty strings.
612    41. Applied a user patch to fix a number of spelling mistakes in comments.
614    42. Data lines longer than 65536 caused pcretest to crash.
616    43. Clarified the data type for length and startoffset arguments for pcre_exec
617        and pcre_dfa_exec in the function-specific man pages, where they were
618        explicitly stated to be in bytes, never having been updated. I also added
619        some clarification to the pcreapi man page.
621    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
622        a segmentation fault.
625    Version 8.32 30-November-2012
626    -----------------------------
628    1.  Improved JIT compiler optimizations for first character search and single
629        character iterators.
631    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
632        Patch by Daniel Richard G.
634    3.  Single character iterator optimizations in the JIT compiler.
636    4.  Improved JIT compiler optimizations for character ranges.
638    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
639        Reported by Giuseppe D'Angelo.
641    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
642        of a line, was being set incorrectly in cases where .* appeared inside
643        atomic brackets at the start of a pattern, or where there was a subsequent
644        *PRUNE or *SKIP.
646    7.  Improved instruction cache flush for POWER/PowerPC.
647        Patch by Daniel Richard G.
649    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
650        grep:
652        (a) There is now no limit to the number of patterns to be matched.
654        (b) An error is given if a pattern is too long.
656        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
657            are now supported.
659        (d) --exclude-from and --include-from (multiple use) have been added.
661        (e) Exclusions and inclusions now apply to all files and directories, not
662            just to those obtained from scanning a directory recursively.
664        (f) Multiple uses of -f and --file-list are now supported.
666        (g) In a Windows environment, the default for -d has been changed from
667            "read" (the GNU grep default) to "skip", because otherwise the presence
668            of a directory in the file list provokes an error.
670        (h) The documentation has been revised and clarified in places.
672    9.  Improve the matching speed of capturing brackets.
674    10. Changed the meaning of \X so that it now matches a Unicode extended
675        grapheme cluster.
677    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
678        out POSIX threads when JIT support is configured.
680    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
682    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
683        structure could go wrong in environments where size_t is not the same size
684        as int.
686    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
687        set.
689    15. The EBCDIC support had decayed; later updates to the code had included
690        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
691        general tidy up of EBCDIC-related issues, and the documentation was also
692        not quite right. There is now a test that can be run on ASCII systems to
693        check some of the EBCDIC-related things (but is it not a full test).
695    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
696        in a small tidy to the code.
698    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
700    18. If the --only-matching (-o) option in pcregrep is specified multiple
701        times, each one causes appropriate output. For example, -o1 -o2 outputs the
702        substrings matched by the 1st and 2nd capturing parentheses. A separating
703        string can be specified by --om-separator (default empty).
705    19. Improving the first n character searches.
707    20. Turn case lists for horizontal and vertical white space into macros so that
708        they are defined only once.
710    21. This set of changes together give more compatible Unicode case-folding
711        behaviour for characters that have more than one other case when UCP
712        support is available.
714        (a) The Unicode property table now has offsets into a new table of sets of
715            three or more characters that are case-equivalent. The MultiStage2.py
716            script that generates these tables (the pcre_ucd.c file) now scans
717            CaseFolding.txt instead of UnicodeData.txt for character case
718            information.
720        (b) The code for adding characters or ranges of characters to a character
721            class has been abstracted into a generalized function that also handles
722            case-independence. In UTF-mode with UCP support, this uses the new data
723            to handle characters with more than one other case.
725        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
726            whose other case is greater than 256 are now correctly matched
727            caselessly. Previously, the high codepoint matched the low one, but not
728            vice versa.
730        (d) The processing of \h, \H, \v, and \ in character classes now makes use
731            of the new class addition function, using character lists defined as
732            macros alongside the case definitions of 20 above.
734        (e) Caseless back references now work with characters that have more than
735            one other case.
737        (f) General caseless matching of characters with more than one other case
738            is supported.
740    22. Unicode character properties were updated from Unicode 6.2.0
742    23. Improved CMake support under Windows. Patch by Daniel Richard G.
744    24. Add support for 32-bit character strings, and UTF-32
746    25. Major JIT compiler update (code refactoring and bugfixing).
747        Experimental Sparc 32 support is added.
749    26. Applied a modified version of Daniel Richard G's patch to create
750        pcre.h.generic and config.h.generic by "make" instead of in the
751        PrepareRelease script.
753    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
754        pcre_compile.c when checking for a zero character.
756    28. Introducing a native interface for JIT. Through this interface, the compiled
757        machine code can be directly executed. The purpose of this interface is to
758        provide fast pattern matching, so several sanity checks are not performed.
759        However, feature tests are still performed. The new interface provides
760        1.4x speedup compared to the old one.
762    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
763        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
764        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
766    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
767        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
768        the "old" RFC 2279). Instead, it ended up passing a negative length to
769        pcre_exec().
771    31. Add support for GCC's visibility feature to hide internal functions.
773    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
774        "unknown -C option" after outputting 0 or 1.
776    33. There is now support for generating a code coverage report for the test
777        suite in environments where gcc is the compiler and lcov is installed. This
778        is mainly for the benefit of the developers.
780    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
781        unaddressable using valgrind annotations, allowing valgrind to detect
782        invalid memory accesses. This is mainly for the benefit of the developers.
784    25. (*UTF) can now be used to start a pattern in any of the three libraries.
786    26. Give configure error if --enable-cpp but no C++ compiler found.
789    Version 8.31 06-July-2012
790    -------------------------
792    1.  Fixing a wrong JIT test case and some compiler warnings.
794    2.  Removed a bashism from the RunTest script.
796    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
797        to unsigned type, result still unsigned" that was given by an MS compiler
798        on encountering the code "-sizeof(xxx)".
800    4.  Partial matching support is added to the JIT compiler.
802    5.  Fixed several bugs concerned with partial matching of items that consist
803        of more than one character:
805        (a) /^(..)\1/ did not partially match "aba" because checking references was
806            done on an "all or nothing" basis. This also applied to repeated
807            references.
809        (b) \R did not give a hard partial match if \r was found at the end of the
810            subject.
812        (c) \X did not give a hard partial match after matching one or more
813            characters at the end of the subject.
815        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
816            a partial match for the string "\r".
818        (e) When newline was set to CRLF, the metacharacter "." did not recognize
819            a partial match for a CR character at the end of the subject string.
821    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
822        running pcretest, the text "(JIT)" added to the output whenever JIT is
823        actually used to run the match.
825    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
826        or /S+[+] with a digit between 1 and 7.
828    8.  OP_NOT now supports any UTF character not just single-byte ones.
830    9.  (*MARK) control verb is now supported by the JIT compiler.
832    10. The command "./RunTest list" lists the available tests without actually
833        running any of them. (Because I keep forgetting what they all are.)
837    12. Applied a (slightly modified) user-supplied patch that improves performance
838        when the heap is used for recursion (compiled with --disable-stack-for-
839        recursion). Instead of malloc and free for each heap frame each time a
840        logical recursion happens, frames are retained on a chain and re-used where
841        possible. This sometimes gives as much as 30% improvement.
843    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
844        call.
846    14. As documented, (*COMMIT) is now confined to within a positive assertion.
848    15. It is now possible to link pcretest with libedit as an alternative to
849        libreadline.
851    16. (*COMMIT) control verb is now supported by the JIT compiler.
853    17. The Unicode data tables have been updated to Unicode 6.1.0.
855    18. Added --file-list option to pcregrep.
857    19. Added binary file support to pcregrep, including the -a, --binary-files,
858        -I, and --text options.
860    20. The madvise function is renamed for posix_madvise for QNX compatibility
861        reasons. Fixed by Giuseppe D'Angelo.
863    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
865    22. Changed the option for creating symbolic links for 16-bit man pages from
866        -s to -sf so that re-installing does not cause issues.
868    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
870    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
871        matches in certain environments (the workspace was not being correctly
872        retained). Also added to pcre_dfa_exec() a simple plausibility check on
873        some of the workspace data at the beginning of a restart.
875    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
876        was not doing so when it should - probably a typo introduced by SVN 528
877        (change 8.10/14).
879    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
880        \w+ when the character tables indicated that \x{c4} was a word character.
881        There were several related cases, all because the tests for doing a table
882        lookup were testing for characters less than 127 instead of 255.
884    27. If a pattern contains capturing parentheses that are not used in a match,
885        their slots in the ovector are set to -1. For those that are higher than
886        any matched groups, this happens at the end of processing. In the case when
887        there were back references that the ovector was too small to contain
888        (causing temporary malloc'd memory to be used during matching), and the
889        highest capturing number was not used, memory off the end of the ovector
890        was incorrectly being set to -1. (It was using the size of the temporary
891        memory instead of the true size.)
893    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
894        ovector size, it uses memory at the end of the block that it has got.
896    29. Check for an overlong MARK name and give an error at compile time. The
897        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
899    30. JIT compiler update.
901    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
902        Rill for the patch.
904    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
906    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
908    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
909        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
910        was enabled.
912    35. Improve JIT code generation for greedy plus quantifier.
914    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
915        1 to "aa" instead of to an empty string. The bug affected repeated groups
916        that could potentially match an empty string.
918    37. Optimizing single character iterators in JIT.
920    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
921        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
922        codepoints that are too big for the mode are faulted, and in a UTF mode,
923        disallowed codepoints are also faulted.
925    39. If PCRE was compiled with UTF support, in three places in the DFA
926        matcher there was code that should only have been obeyed in UTF mode, but
927        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
928        processing when bytes with values greater than 127 were present. In 16-bit
929        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
930        both cases the values are those that cannot be the first data item in a UTF
931        character. The three items that might have provoked this were recursions,
932        possessively repeated groups, and atomic groups.
934    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
935        and pcregrep, because some OS require shared objects to be explicitly
936        passed to ld, causing the link step to fail if they are not.
938    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
939        mode, patterns that started with \h* or \R* might be incorrectly matched.
942    Version 8.30 04-February-2012
943    -----------------------------
945    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
946        name is defined in ctype.h.
948    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
949        only in quite long subpatterns.
951    3.  Removed the function pcre_info(), which has been obsolete and deprecated
952        since it was replaced by pcre_fullinfo() in February 2000.
954    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
955        match a (*MARK), and the match failed at the start of the subject, a
956        reference to memory before the start of the subject could occur. This bug
957        was introduced by fix 17 of release 8.21.
959    5.  A reference to an unset group with zero minimum repetition was giving
960        totally wrong answers (in non-JavaScript-compatibility mode). For example,
961        /(another)?(\1?)test/ matched against "hello world test". This bug was
962        introduced in release 8.13.
964    6.  Add support for 16-bit character strings (a large amount of work involving
965        many changes and refactorings).
967    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
968        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
969        from a file.
971    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
972        rounding is not applied in this particular case).
974    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
975        if they appear, or are escaped, in patterns.
977    10. Get rid of a number of -Wunused-but-set-variable warnings.
979    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
980        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
981        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
982        also returns the mark "x". This bug applied to capturing parentheses,
983        non-capturing parentheses, and atomic parentheses. It also applied to some
984        assertions.
986    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
987        information out of configure.ac instead of relying on pcre.h.generic, which
988        is not stored in the repository.
990    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
991        -lreadline.
993    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
995    15. Applied Graycode's patch to put the top-level frame on the stack rather
996        than the heap when not using the stack for recursion. This gives a
997        performance improvement in many cases when recursion is not deep.
999    16. Experimental code added to "pcretest -C" to output the stack frame size.
1002    Version 8.21 12-Dec-2011
1003    ------------------------
1005    1.  Updating the JIT compiler.
1007    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1008        are added as well.
1010    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1011        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1012        calling _pcre_jit_exec. Some extra comments are added.
1014    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1015        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1016        was introduced by change 18 for 8.20.
1018    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1019        ECMA-262 standard.
1021    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1022        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1023        This bug was probably introduced by change 9 of 8.13.
1025    7.  While fixing 6 above, I noticed that a number of other items were being
1026        incorrectly rejected as "not fixed length". This arose partly because newer
1027        opcodes had not been added to the fixed-length checking code. I have (a)
1028        corrected the bug and added tests for these items, and (b) arranged for an
1029        error to occur if an unknown opcode is encountered while checking for fixed
1030        length instead of just assuming "not fixed length". The items that were
1031        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1032        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1033        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1035    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1036        being incorrectly compiled and would have given unpredicatble results.
1038    9.  A possessively repeated subpattern with minimum repeat count greater than
1039        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1040        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1041        the first (A) could occur when it should not.
1043    10. Add a cast and remove a redundant test from the code.
1045    11. JIT should use pcre_malloc/pcre_free for allocation.
1047    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1048        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1049        is anything other than /usr, -L is still shown).
1051    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1053    14. Perl does not support \N without a following name in a [] class; PCRE now
1054        also gives an error.
1056    15. If a forward reference was repeated with an upper limit of around 2000,
1057        it caused the error "internal error: overran compiling workspace". The
1058        maximum number of forward references (including repeats) was limited by the
1059        internal workspace, and dependent on the LINK_SIZE. The code has been
1060        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1061        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1062        of around 200,000 forward references. While doing this, I also speeded up
1063        the filling in of repeated forward references.
1065    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1066        incorrectly expecting the subject to contain another "a" after the start.
1068    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1069        in the match, the SKIP should be ignored. This was not happening; instead
1070        the SKIP was being treated as NOMATCH. For patterns such as
1071        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1072        tested.
1074    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1075        now much more compatible with Perl, in particular in cases where the result
1076        is a non-match for a non-anchored pattern. For example, if
1077        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1078        "m", where previously it did not return a name. A side effect of this
1079        change is that for partial matches, the last encountered mark name is
1080        returned, as for non matches. A number of tests that were previously not
1081        Perl-compatible have been moved into the Perl-compatible test files. The
1082        refactoring has had the pleasing side effect of removing one argument from
1083        the match() function, thus reducing its stack requirements.
1085    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1086        subsequent uses of /S (without +) incorrectly behaved like /S+.
1088    21. Retrieve executable code size support for the JIT compiler and fixing
1089        some warnings.
1091    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1092        not work when the shorter character appeared right at the end of the
1093        subject string.
1095    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1096        systems.
1098    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1099        output it when the /M option is used in pcretest.
1101    25. The CheckMan script was not being included in the distribution. Also, added
1102        an explicit "perl" to run Perl scripts from the PrepareRelease script
1103        because this is reportedly needed in Windows.
1105    26. If study data was being save in a file and studying had not found a set of
1106        "starts with" bytes for the pattern, the data written to the file (though
1107        never used) was taken from uninitialized memory and so caused valgrind to
1108        complain.
1110    27. Updated RunTest.bat as provided by Sheri Pierce.
1112    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1114    29. Computation of memory usage for the table of capturing group names was
1115        giving an unnecessarily large value.
1118    Version 8.20 21-Oct-2011
1119    ------------------------
1121    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1122        a POSIX class. After further experiments with Perl, which convinced me that
1123        Perl has bugs and confusions, a closing square bracket is no longer allowed
1124        in a POSIX name. This bug also affected patterns with classes that started
1125        with full stops.
1127    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1128        captured substring, but while checking the failing first alternative,
1129        substring 1 is temporarily captured. If the output vector supplied to
1130        pcre_exec() was not big enough for this capture, the yield of the function
1131        was still zero ("insufficient space for captured substrings"). This cannot
1132        be totally fixed without adding another stack variable, which seems a lot
1133        of expense for a edge case. However, I have improved the situation in cases
1134        such as /(a)(b)x|abc/ matched against "abc", where the return code
1135        indicates that fewer than the maximum number of slots in the ovector have
1136        been set.
1138    3.  Related to (2) above: when there are more back references in a pattern than
1139        slots in the output vector, pcre_exec() uses temporary memory during
1140        matching, and copies in the captures as far as possible afterwards. It was
1141        using the entire output vector, but this conflicts with the specification
1142        that only 2/3 is used for passing back captured substrings. Now it uses
1143        only the first 2/3, for compatibility. This is, of course, another edge
1144        case.
1146    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1147        main code base, and can be used by building with --enable-jit. When this is
1148        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1149        runtime --no-jit option is given.
1151    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1152        ovector, the return from the function was zero, implying that there were
1153        other matches that did not fit. The correct "exactly full" value is now
1154        returned.
1156    6.  If a subpattern that was called recursively or as a subroutine contained
1157        (*PRUNE) or any other control that caused it to give a non-standard return,
1158        invalid errors such as "Error -26 (nested recursion at the same subject
1159        position)" or even infinite loops could occur.
1161    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1162        computing the minimum length on reaching *ACCEPT, and so ended up with the
1163        wrong value of 1 rather than 0. Further investigation indicates that
1164        computing a minimum subject length in the presence of *ACCEPT is difficult
1165        (think back references, subroutine calls), and so I have changed the code
1166        so that no minimum is registered for a pattern that contains *ACCEPT.
1168    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1169        it was not handled as intended. [But see 16 below.]
1171    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1172        Sheri Pierce.
1174    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1175        the first byte in a match must be "a".
1177    11. Change 17 for 8.13 increased the recursion depth for patterns like
1178        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1179        pattern contains any instances of (*THEN). If it does not, the old
1180        optimizations are restored. It would be nice to do this on a per-group
1181        basis, but at the moment that is not feasible.
1183    12. In some environments, the output of pcretest -C is CRLF terminated. This
1184        broke RunTest's code that checks for the link size. A single white space
1185        character after the value is now allowed for.
1187    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1188        For "fr", it uses the Windows-specific input and output files.
1190    14. If (*THEN) appeared in a group that was called recursively or as a
1191        subroutine, it did not work as intended. [But see next item.]
1193    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1194        pattern fragments (but not containing any | characters). If A and B are
1195        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1196        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1197        D. In other words, Perl considers parentheses that do not contain any |
1198        characters to be part of a surrounding alternative, whereas PCRE was
1199        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1200        differently. PCRE now behaves in the same way as Perl, except in the case
1201        of subroutine/recursion calls such as (?1) which have in any case always
1202        been different (but PCRE had them first :-).
1204    16. Related to 15 above: Perl does not treat the | in a conditional group as
1205        creating alternatives. Such a group is treated in the same way as an
1206        ordinary group without any | characters when processing (*THEN). PCRE has
1207        been changed to match Perl's behaviour.
1209    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1210        RunGrepTest script failed.
1212    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1213        inevitable for groups that contain captures, but it can lead to a lot of
1214        stack use in large patterns. The old behaviour has been restored for atomic
1215        groups that do not contain any capturing parentheses.
1217    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1218        suppress the check for a minimum subject length at run time. (If it was
1219        given to pcre_exec() or pcre_dfa_exec() it did work.)
1221    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1222        fail to work when decoding hex characters in data strings in EBCDIC
1223        environments.
1225    21. It appears that in at least one Mac OS environment, the isxdigit() function
1226        is implemented as a macro that evaluates to its argument more than once,
1227        contravening the C 90 Standard (I haven't checked a later standard). There
1228        was an instance in pcretest which caused it to go wrong when processing
1229        \x{...} escapes in subject strings. The has been rewritten to avoid using
1230        things like p++ in the argument of isxdigit().
1233    Version 8.13 16-Aug-2011
1234    ------------------------
1236    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1238    2.  Two minor typos in pcre_internal.h have been fixed.
1240    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1241        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1242        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1244    4.  There were a number of related bugs in the code for matching backrefences
1245        caselessly in UTF-8 mode when codes for the characters concerned were
1246        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1247        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1248        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1249        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1250        2-byte code at the end of the subject (it thought there wasn't enough data
1251        left).
1253    5.  Comprehensive information about what went wrong is now returned by
1254        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1255        as the output vector has at least 2 elements. The offset of the start of
1256        the failing character and a reason code are placed in the vector.
1258    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1259        now returned is for the first byte of the failing character, instead of the
1260        last byte inspected. This is an incompatible change, but I hope it is small
1261        enough not to be a problem. It makes the returned offset consistent with
1262        pcre_exec() and pcre_dfa_exec().
1264    7.  pcretest now gives a text phrase as well as the error number when
1265        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1266        failure, the offset and reason code are output.
1268    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1269        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1270        back over a single character (\n). This seems wrong (because it treated the
1271        two characters as a single entity when going forwards), conflicts with the
1272        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1273        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1274        has been changed.
1276    9.  Some internal refactoring has changed the processing so that the handling
1277        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1278        time (the PCRE_DOTALL option was changed this way some time ago: version
1279        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1280        which was always a bit of a fudge. It also means that there is one less
1281        argument for the match() function, which reduces its stack requirements
1282        slightly. This change also fixes an incompatibility with Perl: the pattern
1283        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1285    10. More internal refactoring has drastically reduced the number of recursive
1286        calls to match() for possessively repeated groups such as (abc)++ when
1287        using pcre_exec().
1289    11. While implementing 10, a number of bugs in the handling of groups were
1290        discovered and fixed:
1292        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1293        (a|)*(?1) gave a compile-time internal error.
1294        ((a|)+)+  did not notice that the outer group could match an empty string.
1295        (^a|^)+   was not marked as anchored.
1296        (.*a|.*)+ was not marked as matching at start or after a newline.
1298    12. Yet more internal refactoring has removed another argument from the match()
1299        function. Special calls to this function are now indicated by setting a
1300        value in a variable in the "match data" data block.
1302    13. Be more explicit in pcre_study() instead of relying on "default" for
1303        opcodes that mean there is no starting character; this means that when new
1304        ones are added and accidentally left out of pcre_study(), testing should
1305        pick them up.
1307    14. The -s option of pcretest has been documented for ages as being an old
1308        synonym of -m (show memory usage). I have changed it to mean "force study
1309        for every regex", that is, assume /S for every regex. This is similar to -i
1310        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1311        using it. It makes it easier to run collections of tests with and without
1312        study enabled, and thereby test pcre_study() more easily. All the standard
1313        tests are now run with and without -s (but some patterns can be marked as
1314        "never study" - see 20 below).
1316    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1317        restoration of the capturing data to the outer values was not happening
1318        correctly.
1320    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1321        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1322        pattern had matched an empty string, and so incorrectly returned a no
1323        match.
1325    17. There was optimizing code for the last branch of non-capturing parentheses,
1326        and also for the obeyed branch of a conditional subexpression, which used
1327        tail recursion to cut down on stack usage. Unfortunately, now that there is
1328        the possibility of (*THEN) occurring in these branches, tail recursion is
1329        no longer possible because the return has to be checked for (*THEN). These
1330        two optimizations have therefore been removed. [But see 8.20/11 above.]
1332    18. If a pattern containing \R was studied, it was assumed that \R always
1333        matched two bytes, thus causing the minimum subject length to be
1334        incorrectly computed because \R can also match just one byte.
1336    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1337        was incorrectly computed.
1339    20. If /S is present twice on a test pattern in pcretest input, it now
1340        *disables* studying, thereby overriding the use of -s on the command line
1341        (see 14 above). This is necessary for one or two tests to keep the output
1342        identical in both cases.
1344    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1345        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1347    22. When an atomic group that contained a capturing parenthesis was
1348        successfully matched, but the branch in which it appeared failed, the
1349        capturing was not being forgotten if a higher numbered group was later
1350        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1351        group 1 to "a", when in fact it should be unset. This applied to multi-
1352        branched capturing and non-capturing groups, repeated or not, and also to
1353        positive assertions (capturing in negative assertions does not happen
1354        in PCRE) and also to nested atomic groups.
1356    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1357        subject after a captured substring, to make it easier to tell which of a
1358        number of identical substrings has been captured.
1360    24. The way atomic groups are processed by pcre_exec() has been changed so that
1361        if they are repeated, backtracking one repetition now resets captured
1362        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1363        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1364        "aaa". Previously, it would have been "a". As part of this code
1365        refactoring, the way recursive calls are handled has also been changed.
1367    25. If an assertion condition captured any substrings, they were not passed
1368        back unless some other capturing happened later. For example, if
1369        (?(?=(a))a) was matched against "a", no capturing was returned.
1371    26. When studying a pattern that contained subroutine calls or assertions,
1372        the code for finding the minimum length of a possible match was handling
1373        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1374        group 1 called group 2 while simultaneously a separate group 2 called group
1375        1). A stack overflow occurred in this case. I have fixed this by limiting
1376        the recursion depth to 10.
1378    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1379        Fortmann. This supports explicit test numbers on the command line, and has
1380        argument validation and error reporting.
1382    28. An instance of \X with an unlimited repeat could fail if at any point the
1383        first character it looked at was a mark character.
1385    29. Some minor code refactoring concerning Unicode properties and scripts
1386        should reduce the stack requirement of match() slightly.
1388    30. Added the '=' option to pcretest to check the setting of unused capturing
1389        slots at the end of the pattern, which are documented as being -1, but are
1390        not included in the return count.
1392    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1393        compiled something random. Now it gives a compile-time error (as does
1394        Perl).
1396    32. A *MARK encountered during the processing of a positive assertion is now
1397        recorded and passed back (compatible with Perl).
1399    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1400        had alternative anchored branches, the search for a second match in a line
1401        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1402        matched the line "0102" twice. The same bug affected patterns that started
1403        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1404        twice.
1406    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1407        does, and because of capturing effects, quantifying parenthesized
1408        assertions may at times be useful. Quantifiers are now allowed for
1409        parenthesized assertions.
1411    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1413    36. \g was being checked for fancy things in a character class, when it should
1414        just be a literal "g".
1416    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1417        appearance of a nested POSIX class supersedes an apparent external class.
1418        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1419        unescaped square brackets may also appear as part of class names. For
1420        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1421        more like Perl. (But see 8.20/1 above.)
1423    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1424        was because it thought it was \N{name}, which is not supported).
1426    39. Add minix to OS list not supporting the -S option in pcretest.
1428    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1429        cannot analyze patterns in sufficient detail to catch mutual recursions
1430        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1431        subgroup is called recursively as a subpattern for a second time at the
1432        same position in the subject string. In previous releases this might have
1433        been caught by the recursion limit, or it might have run out of stack.
1435    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1436        happen only once. PCRE was, however incorrectly giving a compile time error
1437        "recursive call could loop indefinitely" because it cannot analyze the
1438        pattern in sufficient detail. The compile time test no longer happens when
1439        PCRE is compiling a conditional subpattern, but actual runaway loops are
1440        now caught at runtime (see 40 above).
1442    42. It seems that Perl allows any characters other than a closing parenthesis
1443        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1444        has been changed to be the same.
1446    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1447        as not to get warnings when autogen.sh is called. Also changed
1448        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1450    44. To help people who use pcregrep to scan files containing exceedingly long
1451        lines, the following changes have been made:
1453        (a) The default value of the buffer size parameter has been increased from
1454            8K to 20K. (The actual buffer used is three times this size.)
1456        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1457            PCRE is built.
1459        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1460            to be set at run time.
1462        (d) Numerical values in pcregrep options can be followed by K or M, for
1463            example --buffer-size=50K.
1465        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1466            given and the return code is set to 2.
1468    45. Add a pointer to the latest mark to the callout data block.
1470    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1471        partial match of an empty string instead of no match. This was specific to
1472        the use of ".".
1474    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1475        complete match instead of a partial match. This bug was dependent on both
1476        the PCRE_UTF8 and PCRE_DOTALL options being set.
1478    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1479        starting byte set, because \b was not being ignored.
1482    Version 8.12 15-Jan-2011
1483    ------------------------
1485    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1486        checks for such things as part of the documentation building process.
1488    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1489        --match-limit and --recursion-limit options (added for 8.11). In
1490        particular, this made one of the standard tests fail. (The integer value
1491        went into the wrong half of a long int.)
1493    3.  If the --colour option was given to pcregrep with -v (invert match), it
1494        did strange things, either producing crazy output, or crashing. It should,
1495        of course, ignore a request for colour when reporting lines that do not
1496        match.
1498    4.  Another pcregrep bug caused similar problems if --colour was specified with
1499        -M (multiline) and the pattern match finished with a line ending.
1501    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1502        matched in multiline mode, the following line was shown as part of the
1503        match. This seems wrong, so I have changed it.
1505    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1506        the check for further matches in the same line (so they could be coloured)
1507        to overrun the end of the current line. If another match was found, it was
1508        incorrectly shown (and then shown again when found in the next line).
1510    7.  If pcregrep was compiled under Windows, there was a reference to the
1511        function pcregrep_exit() before it was defined. I am assuming this was
1512        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1513        reported by a user. I've moved the definition above the reference.
1516    Version 8.11 10-Dec-2010
1517    ------------------------
1519    1.  (*THEN) was not working properly if there were untried alternatives prior
1520        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1521        backtracked to try for "b" instead of moving to the next alternative branch
1522        at the same level (in this case, to look for "c"). The Perl documentation
1523        is clear that when (*THEN) is backtracked onto, it goes to the "next
1524        alternative in the innermost enclosing group".
1526    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1527        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1528        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1529        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1530        (*THEN).
1532    3.  If \s appeared in a character class, it removed the VT character from
1533        the class, even if it had been included by some previous item, for example
1534        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1535        of \s, but is part of the POSIX "space" class.)
1537    4.  A partial match never returns an empty string (because you can always
1538        match an empty string at the end of the subject); however the checking for
1539        an empty string was starting at the "start of match" point. This has been
1540        changed to the "earliest inspected character" point, because the returned
1541        data for a partial match starts at this character. This means that, for
1542        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1543        (previously it gave "no match").
1545    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1546        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1547        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1548        has an implication that the given string is incomplete (because a partial
1549        match is preferred over a full match). For this reason, these items now
1550        give a partial match in this situation. [Aside: previously, the one case
1551        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1552        match rather than a full match, which was wrong by the old rules, but is
1553        now correct.]
1555    6.  There was a bug in the handling of #-introduced comments, recognized when
1556        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1557        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1558        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1559        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1560        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1561        places in pcre_compile().
1563    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1564        comments when looking ahead for named forward references to subpatterns,
1565        the only newline sequence it recognized was NL. It now handles newlines
1566        according to the set newline convention.
1568    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1569        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1570        cater for a lack of strerror(). These oversights have been fixed.
1572    9.  Added --match-limit and --recursion-limit to pcregrep.
1574    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1576    11. When the -o option was used, pcregrep was setting a return code of 1, even
1577        when matches were found, and --line-buffered was not being honoured.
1579    12. Added an optional parentheses number to the -o and --only-matching options
1580        of pcregrep.
1582    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1583        can match an empty string. The code to do it in pcretest and pcredemo
1584        needed fixing:
1586        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1587            only one byte after an empty string match just before CRLF (this case
1588            just got forgotten; "any" and "anycrlf" were OK).
1590        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1591            mode when an empty string match preceded an ASCII character followed by
1592            a non-ASCII character. (The code for advancing by one character rather
1593            than one byte was nonsense.)
1595        (c) The pcredemo.c sample program did not have any code at all to handle
1596            the cases when CRLF is a valid newline sequence.
1598    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1599        as a starting offset was within the subject string. There is now a new
1600        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1601        negative or greater than the length of the string. In order to test this,
1602        pcretest is extended to allow the setting of negative starting offsets.
1604    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1605        starting offset points to the beginning of a UTF-8 character was
1606        unnecessarily clumsy. I tidied it up.
1608    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1609        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1611    17. Nobody had reported that the --include_dir option, which was added in
1612        release 7.7 should have been called --include-dir (hyphen, not underscore)
1613        for compatibility with GNU grep. I have changed it to --include-dir, but
1614        left --include_dir as an undocumented synonym, and the same for
1615        --exclude-dir, though that is not available in GNU grep, at least as of
1616        release 2.5.4.
1618    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1619        characters from a string of bytes) have been redefined so as not to use
1620        loops, in order to improve performance in some environments. At the same
1621        time, I abstracted some of the common code into auxiliary macros to save
1622        repetition (this should not affect the compiled code).
1624    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1625        compile-time error is now given if \c is not followed by an ASCII
1626        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1627        different, and any byte value is allowed.)
1629    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1630        START_OPTIMIZE option, which is now allowed at compile time - but just
1631        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1632        to pcregrep and other applications that have no direct access to PCRE
1633        options. The new /Y option in pcretest sets this option when calling
1634        pcre_compile().
1636    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1637        back references. Groups containing recursive back references were forced to
1638        be atomic by that change, but in the case of named groups, the amount of
1639        memory required was incorrectly computed, leading to "Failed: internal
1640        error: code overflow". This has been fixed.
1642    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1643        pcretest.c, to avoid build problems in some Borland environments.
1646    Version 8.10 25-Jun-2010
1647    ------------------------
1649    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1650        THEN.
1652    2.  (*ACCEPT) was not working when inside an atomic group.
1654    3.  Inside a character class, \B is treated as a literal by default, but
1655        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1656        causes the error). The code is unchanged, but I tidied the documentation.
1658    4.  Inside a character class, PCRE always treated \R and \X as literals,
1659        whereas Perl faults them if its -w option is set. I have changed PCRE so
1660        that it faults them when PCRE_EXTRA is set.
1662    5.  Added support for \N, which always matches any character other than
1663        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1665    6.  When compiling pcregrep with newer versions of gcc which may have
1666        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1667        declared with attribute warn_unused_result" were given. Just casting the
1668        result to (void) does not stop the warnings; a more elaborate fudge is
1669        needed. I've used a macro to implement this.
1671    7.  Minor change to pcretest.c to avoid a compiler warning.
1673    8.  Added four artifical Unicode properties to help with an option to make
1674        \s etc use properties (see next item). The new properties are: Xan
1675        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1677    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1678        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1679        this option. Modified pcretest to add /W to test this facility. Added
1680        REG_UCP to make it available via the POSIX interface.
1682    10. Added --line-buffered to pcregrep.
1684    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1685        studied, and the match started with a letter with a code point greater than
1686        127 whose first byte was different to the first byte of the other case of
1687        the letter, the other case of this starting letter was not recognized
1688        (#976).
1690    12. If a pattern that was studied started with a repeated Unicode property
1691        test, for example, \p{Nd}+, there was the theoretical possibility of
1692        setting up an incorrect bitmap of starting bytes, but fortunately it could
1693        not have actually happened in practice until change 8 above was made (it
1694        added property types that matched character-matching opcodes).
1696    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1697        possible starting bytes for non-anchored patterns.
1699    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1700        \R, and also a number of cases that involve Unicode properties, both
1701        explicit and implicit when PCRE_UCP is set.
1703    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1704        input, it could crash or give wrong results if characters with values
1705        greater than 0xc0 were present in the subject string. (Detail: it assumed
1706        UTF-8 input when processing these items.)
1708    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1709        size_t is 64-bit (#991).
1711    17. Added a check for running out of memory when PCRE is compiled with
1712        --disable-stack-for-recursion (#990).
1714    18. If the last data line in a file for pcretest does not have a newline on
1715        the end, a newline was missing in the output.
1717    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1718        less than 128) in its various bitmaps. However, there is a facility for
1719        generating tables according to the current locale when PCRE is compiled. It
1720        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1721        characters, are recognized by isspace() and therefore were getting set in
1722        these tables, and indeed these tables seem to approximate to ISO 8859. This
1723        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1724        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1725        which of course cannot start UTF-8 characters. I have changed the code so
1726        that only real ASCII characters (less than 128) and the correct starting
1727        bytes for UTF-8 encodings are set for characters greater than 127 when in
1728        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1729        altogether.)
1731    20. Added the /T option to pcretest so as to be able to run tests with non-
1732        standard character tables, thus making it possible to include the tests
1733        used for 19 above in the standard set of tests.
1735    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1736        reference to a subpattern the other side of a comment that contains an
1737        opening parenthesis caused either an internal compiling error, or a
1738        reference to the wrong subpattern.
1741    Version 8.02 19-Mar-2010
1742    ------------------------
1744    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1746    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1747        configured.
1749    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1750        original author of that file, following a query about its status.
1752    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1753        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1755    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1756        quantifier applied to a forward-referencing subroutine call, could compile
1757        incorrect code or give the error "internal error: previously-checked
1758        referenced subpattern not found".
1760    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1761        variables to point to external functions. For these systems, therefore,
1762        pcre_malloc etc. are now initialized to local functions that call the
1763        relevant global functions.
1765    7.  There were two entries missing in the vectors called coptable and poptable
1766        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1767        I've fixed the data, and added a kludgy way of testing at compile time that
1768        the lengths are correct (equal to the number of opcodes).
1770    8.  Following on from 7, I added a similar kludge to check the length of the
1771        eint vector in pcreposix.c.
1773    9.  Error texts for pcre_compile() are held as one long string to avoid too
1774        much relocation at load time. To find a text, the string is searched,
1775        counting zeros. There was no check for running off the end of the string,
1776        which could happen if a new error number was added without updating the
1777        string.
1779    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1781    11. \K was not working if it appeared in an atomic group or in a group that
1782        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1783        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1784        the assertion is positive, but not if it is negative.
1786    12. Change 11 fortuitously reduced the size of the stack frame used in the
1787        "match()" function of pcre_exec.c by one pointer. Forthcoming
1788        implementation of support for (*MARK) will need an extra pointer on the
1789        stack; I have reserved it now, so that the stack frame size does not
1790        decrease.
1792    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1793        item in branch that calls a recursion is a subroutine call - as in the
1794        second branch in the above example - was incorrectly given the compile-
1795        time error "recursive call could loop indefinitely" because pcre_compile()
1796        was not correctly checking the subroutine for matching a non-empty string.
1798    14. The checks for overrunning compiling workspace could trigger after an
1799        overrun had occurred. This is a "should never occur" error, but it can be
1800        triggered by pathological patterns such as hundreds of nested parentheses.
1801        The checks now trigger 100 bytes before the end of the workspace.
1803    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1806    Version 8.01 19-Jan-2010
1807    ------------------------
1809    1.  If a pattern contained a conditional subpattern with only one branch (in
1810        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1811        computed the wrong minimum data length (which is of course zero for such
1812        subpatterns). This could cause incorrect "no match" results.
1814    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1815        the pattern is reset in the first branch, pcre_compile() failed with
1816        "internal error: code overflow at offset...". This happened only when
1817        the reset was to the original external option setting. (An optimization
1818        abstracts leading options settings into an external setting, which was the
1819        cause of this.)
1821    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1822        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1823        assertion pattern did not match (meaning that the assertion was true), it
1824        was incorrectly treated as false if the SKIP had been reached during the
1825        matching. This also applied to assertions used as conditions.
1827    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1828        assertion subpattern, including such a pattern used as a condition,
1829        unpredictable results occurred, instead of the error return
1832    5.  The C++ GlobalReplace function was not working like Perl for the special
1833        situation when an empty string is matched. It now does the fancy magic
1834        stuff that is necessary.
1836    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1837        removed. (These were left over from very, very early versions of PCRE.)
1839    7.  Some cosmetic changes to the code to make life easier when compiling it
1840        as part of something else:
1842        (a) Change DEBUG to PCRE_DEBUG.
1844        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1845            called "current" as "current_branch", to prevent a collision with the
1846            Linux macro when compiled as a kernel module.
1848        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1849            prevent a collision with the Linux macro when compiled as a kernel
1850            module.
1852    8.  In pcre_compile() there are some checks for integer overflows that used to
1853        cast potentially large values to (double). This has been changed to that
1854        when building, a check for int64_t is made, and if it is found, it is used
1855        instead, thus avoiding the use of floating point arithmetic. (There is no
1856        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1857        double.
1859    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1860        2005 (difference between two addresses compared to an unsigned value).
1862    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1863        custom one, because of the following reported problem in Windows:
1865          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1866              under Win32.
1867          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1868              therefore missing the function definition.
1869          - The compiler thus generates a "C" signature for the test function.
1870          - The linker fails to find the "C" function.
1871          - PCRE fails to configure if asked to do so against libbz2.
1873    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1874        messages were output:
1876          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1877          rerunning libtoolize, to keep the correct libtool macros in-tree.
1878          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1880        I have done both of these things.
1882    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1883        most of the time, it *can* run out if it is given a pattern that contains a
1884        runaway infinite recursion. I updated the discussion in the pcrestack man
1885        page.
1887    13. Now that we have gone to the x.xx style of version numbers, the minor
1888        version may start with zero. Using 08 or 09 is a bad idea because users
1889        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1890        interpreted as invalid octal numbers. I've updated the previous comment in
1891        configure.ac, and also added a check that gives an error if 08 or 09 are
1892        used.
1894    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1895        causing partial matching to fail when the end of the subject matched \W
1896        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1898    15. There were some discrepancies between the declarations in pcre_internal.h
1899        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1900        their definitions. The declarations used "const uschar *" and the
1901        definitions used USPTR. Even though USPTR is normally defined as "const
1902        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1903        reported that: "This difference in casting confuses some C++ compilers, for
1904        example, SunCC recognizes above declarations as different functions and
1905        generates broken code for hbpcre." I have changed the declarations to use
1906        USPTR.
1908    16. GNU libtool is named differently on some systems. The autogen.sh script now
1909        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1910        (FreeBSD).
1912    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1913        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1914        comment: "Figure out how to create a longlong from a string: strtoll and
1915        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1916        instance, but it only takes 2 args instead of 3!"
1918    18. A subtle bug concerned with back references has been fixed by a change of
1919        specification, with a corresponding code fix. A pattern such as
1920        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1921        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1922        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1923        same bug. Such groups have to be quantified to be useful, or contained
1924        inside another quantified group. (If there's no repetition, the reference
1925        can never match.) The problem arises because, having left the group and
1926        moved on to the rest of the pattern, a later failure that backtracks into
1927        the group uses the captured value from the final iteration of the group
1928        rather than the correct earlier one. I have fixed this in PCRE by forcing
1929        any group that contains a reference to itself to be an atomic group; that
1930        is, there cannot be any backtracking into it once it has completed. This is
1931        similar to recursive and subroutine calls.
1934    Version 8.00 19-Oct-09
1935    ----------------------
1937    1.  The table for translating pcre_compile() error codes into POSIX error codes
1938        was out-of-date, and there was no check on the pcre_compile() error code
1939        being within the table. This could lead to an OK return being given in
1940        error.
1942    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1943        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1944        in a Windows environment.
1946    3.  The pcregrep --count option prints the count for each file even when it is
1947        zero, as does GNU grep. However, pcregrep was also printing all files when
1948        --files-with-matches was added. Now, when both options are given, it prints
1949        counts only for those files that have at least one match. (GNU grep just
1950        prints the file name in this circumstance, but including the count seems
1951        more useful - otherwise, why use --count?) Also ensured that the
1952        combination -clh just lists non-zero counts, with no names.
1954    4.  The long form of the pcregrep -F option was incorrectly implemented as
1955        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1956        but it seems right to fix it, and I didn't think it was worth preserving
1957        the old behaviour.
1959    5.  The command line items --regex=pattern and --regexp=pattern were not
1960        recognized by pcregrep, which required --regex pattern or --regexp pattern
1961        (with a space rather than an '='). The man page documented the '=' forms,
1962        which are compatible with GNU grep; these now work.
1964    6.  No libpcreposix.pc file was created for pkg-config; there was just
1965        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1967    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1968        when UCP support is not needed, by modifying the Python script that
1969        generates it from Unicode data files. This should not matter if the module
1970        is correctly used as a library, but I received one complaint about 50K of
1971        unwanted data. My guess is that the person linked everything into his
1972        program rather than using a library. Anyway, it does no harm.
1974    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1975        was a minimum greater than 1 for a wide character in a possessive
1976        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1977        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1978        character. Chaos in the form of incorrect output or a compiling loop could
1979        result.
1981    9.  The restrictions on what a pattern can contain when partial matching is
1982        requested for pcre_exec() have been removed. All patterns can now be
1983        partially matched by this function. In addition, if there are at least two
1984        slots in the offset vector, the offset of the earliest inspected character
1985        for the match and the offset of the end of the subject are set in them when
1986        PCRE_ERROR_PARTIAL is returned.
1988    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1989        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1990        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1991        and may be more useful for multi-segment matching.
1993    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1994        used to be given if ever the end of the subject was reached; now it is
1995        given only if matching could not proceed because another character was
1996        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1997        string "Z", which now yields "no match" instead of "partial match". In the
1998        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1999        final character ended with (*FAIL).
2001    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
2002        if the pattern had a "must contain" character that was already found in the
2003        earlier partial match, unless partial matching was again requested. For
2004        example, with the pattern /dog.(body)?/, the "must contain" character is
2005        "g". If the first part-match was for the string "dog", restarting with
2006        "sbody" failed. This bug has been fixed.
2008    13. The string returned by pcre_dfa_exec() after a partial match has been
2009        changed so that it starts at the first inspected character rather than the
2010        first character of the match. This makes a difference only if the pattern
2011        starts with a lookbehind assertion or \b or \B (\K is not supported by
2012        pcre_dfa_exec()). It's an incompatible change, but it makes the two
2013        matching functions compatible, and I think it's the right thing to do.
2015    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
2016        so that the demonstration program is easily available in environments where
2017        PCRE has not been installed from source.
2019    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
2020        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
2021        library.
2023    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
2024        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
2025        is not the first non-POSIX option to be added. Clearly some people find
2026        these options useful.
2028    17. If a caller to the POSIX matching function regexec() passes a non-zero
2029        value for nmatch with a NULL value for pmatch, the value of
2030        nmatch is forced to zero.
2032    18. RunGrepTest did not have a test for the availability of the -u option of
2033        the diff command, as RunTest does. It now checks in the same way as
2034        RunTest, and also checks for the -b option.
2036    19. If an odd number of negated classes containing just a single character
2037        interposed, within parentheses, between a forward reference to a named
2038        subpattern and the definition of the subpattern, compilation crashed with
2039        an internal error, complaining that it could not find the referenced
2040        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
2041        [The bug was that it was starting one character too far in when skipping
2042        over the character class, thus treating the ] as data rather than
2043        terminating the class. This meant it could skip too much.]
2045    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
2046        /g option in pcretest when the pattern contains \K, which makes it possible
2047        to have an empty string match not at the start, even when the pattern is
2048        anchored. Updated pcretest and pcredemo to use this option.
2050    21. If the maximum number of capturing subpatterns in a recursion was greater
2051        than the maximum at the outer level, the higher number was returned, but
2052        with unset values at the outer level. The correct (outer level) value is
2053        now given.
2055    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
2056        PCRE did not set those parentheses (unlike Perl). I have now found a way to
2057        make it do so. The string so far is captured, making this feature
2058        compatible with Perl.
2060    23. The tests have been re-organized, adding tests 11 and 12, to make it
2061        possible to check the Perl 5.10 features against Perl 5.10.
2063    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
2064        pattern matches a fixed length string. PCRE did not allow this; now it
2065        does. Neither allows recursion.
2067    25. I finally figured out how to implement a request to provide the minimum
2068        length of subject string that was needed in order to match a given pattern.
2069        (It was back references and recursion that I had previously got hung up
2070        on.) This code has now been added to pcre_study(); it finds a lower bound
2071        to the length of subject needed. It is not necessarily the greatest lower
2072        bound, but using it to avoid searching strings that are too short does give
2073        some useful speed-ups. The value is available to calling programs via
2074        pcre_fullinfo().
2076    26. While implementing 25, I discovered to my embarrassment that pcretest had
2077        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
2078        study optimizations had never been tested with that matching function.
2079        Oops. What is worse, even when it was passed study data, there was a bug in
2080        pcre_dfa_exec() that meant it never actually used it. Double oops. There
2081        were also very few tests of studied patterns with pcre_dfa_exec().
2083    27. If (?| is used to create subpatterns with duplicate numbers, they are now
2084        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
2085        on the other side of the coin, they are no longer allowed to have different
2086        names, because these cannot be distinguished in PCRE, and this has caused
2087        confusion. (This is a difference from Perl.)
2089    28. When duplicate subpattern names are present (necessarily with different
2090        numbers, as required by 27 above), and a test is made by name in a
2091        conditional pattern, either for a subpattern having been matched, or for
2092        recursion in such a pattern, all the associated numbered subpatterns are
2093        tested, and the overall condition is true if the condition is true for any
2094        one of them. This is the way Perl works, and is also more like the way
2095        testing by number works.
2098    Version 7.9 11-Apr-09
2099    ---------------------
2101    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
2102        (pcretest), all targets were linked against these libraries. This included
2103        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
2104        libraries. This caused unwanted dependencies to be created. This problem
2105        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
2106        pcretest is linked with readline.
2108    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
2109        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
2110        moved outside it again, because FALSE and TRUE are already defined in AIX,
2111        but BOOL is not.
2113    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
2114        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
2116    4.  The pcregrep documentation said spaces were inserted as well as colons (or
2117        hyphens) following file names and line numbers when outputting matching
2118        lines. This is not true; no spaces are inserted. I have also clarified the
2119        wording for the --colour (or --color) option.
2121    5.  In pcregrep, when --colour was used with -o, the list of matching strings
2122        was not coloured; this is different to GNU grep, so I have changed it to be
2123        the same.
2125    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
2126        each matching line was coloured. Now it goes on to look for further matches
2127        of any of the test patterns, which is the same behaviour as GNU grep.
2129    7.  A pattern that could match an empty string could cause pcregrep to loop; it
2130        doesn't make sense to accept an empty string match in pcregrep, so I have
2131        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2132        seems to be how GNU grep behaves. [But see later change 40 for release
2133        8.33.]
2135    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2136        start or after a newline", because the conditional assertion was not being
2137        correctly handled. The rule now is that both the assertion and what follows
2138        in the first alternative must satisfy the test.
2140    9.  If auto-callout was enabled in a pattern with a conditional group whose
2141        condition was an assertion, PCRE could crash during matching, both with
2142        pcre_exec() and pcre_dfa_exec().
2144    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2145        used for matching.
2147    11. Unicode property support in character classes was not working for
2148        characters (bytes) greater than 127 when not in UTF-8 mode.
2150    12. Added the -M command line option to pcretest.
2152    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2154    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2156    16. Added comments and documentation about mis-use of no_arg in the C++
2157        wrapper.
2159    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2160        from Martin Jerabek that uses macro names for all relevant character and
2161        string constants.
2163    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2164        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2165        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2166        these, but not everybody uses configure.
2168    19. A conditional group that had only one branch was not being correctly
2169        recognized as an item that could match an empty string. This meant that an
2170        enclosing group might also not be so recognized, causing infinite looping
2171        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2172        with the subject "ab", where knowledge that the repeated group can match
2173        nothing is needed in order to break the loop.
2175    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2176        exec(), but without supplying a callout function, matching went wrong.
2178    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2179        leak if the size of the offset vector was greater than 30. When the vector
2180        is smaller, the saved offsets during recursion go onto a local stack
2181        vector, but for larger vectors malloc() is used. It was failing to free
2182        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2183        error, in fact).
2185    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2186        heapframe that is used only when UTF-8 support is enabled. This caused no
2187        problem, but was untidy.
2189    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2190        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2191        included within another project.
2193    24. Steven Van Ingelgem's patches to add more options to the CMake support,
2194        slightly modified by me:
2196          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2197              not building pcregrep.
2199          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2200              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2202    25. Forward references, both numeric and by name, in patterns that made use of
2203        duplicate group numbers, could behave incorrectly or give incorrect errors,
2204        because when scanning forward to find the reference group, PCRE was not
2205        taking into account the duplicate group numbers. A pattern such as
2206        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2208    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2209        the feature of a custom pointer more persuasive (as requested by a user).
2211    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2212        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2213        pre-defined.
2215    28. Added support for (*UTF8) at the start of a pattern.
2217    29. Arrange for flags added by the "release type" setting in CMake to be shown
2218        in the configuration summary.
2221    Version 7.8 05-Sep-08
2222    ---------------------
2224    1.  Replaced UCP searching code with optimized version as implemented for Ad
2225        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2226        stage table and inline lookup instead of a function, giving speed ups of 2
2227        to 5 times on some simple patterns that I tested. Permission was given to
2228        distribute the MultiStage2.py script that generates the tables (it's not in
2229        the tarball, but is in the Subversion repository).
2231    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2232        scripts.
2234    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2235        a group with a zero qualifier. The result of the study could be incorrect,
2236        or the function might crash, depending on the pattern.
2238    4.  Caseless matching was not working for non-ASCII characters in back
2239        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2240        It now works when Unicode Property Support is available.
2242    5.  In pcretest, an escape such as \x{de} in the data was always generating
2243        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2244        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2245        truncation.
2247    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2249    7.  Added two (int) casts to pcregrep when printing the difference of two
2250        pointers, in case they are 64-bit values.
2252    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2253        test 2 if it fails.
2255    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2256        and a #define of that name to empty if it is not externally set. This is to
2257        allow users of MSVC to set it if necessary.
2259    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2260        the convenience functions in the pcre_get.c source file.
2262    11. An option change at the start of a pattern that had top-level alternatives
2263        could cause overwriting and/or a crash. This command provoked a crash in
2264        some environments:
2266          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2268        This potential security problem was recorded as CVE-2008-2371.
2270    12. For a pattern where the match had to start at the beginning or immediately
2271        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2272        pcre_dfa_exec() could read past the end of the passed subject if there was
2273        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2274        pcretest so that it places the subject at the end of its malloc-ed buffer.
2276    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2277        exec() might read past the end of the data buffer in UTF-8 mode.
2279    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2280        the data contained the byte 0x85 as part of a UTF-8 character within its
2281        first line. This applied both to normal and DFA matching.
2283    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2284        /^[^d]*?$/8 failed to match "abc".
2286    16. Added a missing copyright notice to pcrecpp_internal.h.
2288    17. Make it more clear in the documentation that values returned from
2289        pcre_exec() in ovector are byte offsets, not character counts.
2291    18. Tidied a few places to stop certain compilers from issuing warnings.
2293    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2294        supplied by Stefan Weber. I made a further small update for 7.8 because
2295        there is a change of source arrangements: the pcre_searchfuncs.c module is
2296        replaced by pcre_ucd.c.
2299    Version 7.7 07-May-08
2300    ---------------------
2302    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2303        a string to a long long, pretend we don't even have a long long." This is
2304        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2306    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2307        pre-7.6 versions, which defined a global no_arg variable instead of putting
2308        it in the RE class. (See also #8 below.)
2310    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2311        Lopes.
2313    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2315        (1) The include/exclude patterns were being applied to the whole pathnames
2316            of files, instead of just to the final components.
2318        (2) If there was more than one level of directory, the subdirectories were
2319            skipped unless they satisfied the include/exclude conditions. This is
2320            inconsistent with GNU grep (and could even be seen as contrary to the
2321            pcregrep specification - which I improved to make it absolutely clear).
2322            The action now is always to scan all levels of directory, and just
2323            apply the include/exclude patterns to regular files.
2325    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2326        --exclude_dir in the tests to avoid scanning .svn directories.
2328    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2329        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2330        doesn't support NULs in patterns.
2332    7.  Added some missing "const"s to declarations of static tables in
2333        pcre_compile.c and pcre_dfa_exec.c.
2335    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2336        caused by fix #2  above. (Subsequently also a second patch to fix the
2337        first patch. And a third patch - this was a messy problem.)
2339    9.  Applied Craig's patch to remove the use of push_back().
2341    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2342        matching function regexec().
2344    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2345        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2346        references. PCRE supports relative numbers with this syntax (I don't think
2347        Oniguruma does).
2349    12. Previously, a group with a zero repeat such as (...){0} was completely
2350        omitted from the compiled regex. However, this means that if the group
2351        was called as a subroutine from elsewhere in the pattern, things went wrong
2352        (an internal error was given). Such groups are now left in the compiled
2353        pattern, with a new opcode that causes them to be skipped at execution
2354        time.
2356    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2357        to the way PCRE behaves:
2359        (a) A lone ] character is dis-allowed (Perl treats it as data).
2361        (b) A back reference to an unmatched subpattern matches an empty string
2362            (Perl fails the current match path).
2364        (c) A data ] in a character class must be notated as \] because if the
2365            first data character in a class is ], it defines an empty class. (In
2366            Perl it is not possible to have an empty class.) The empty class []
2367            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2368            The negative empty class [^] matches any one character, independently
2369            of the DOTALL setting.
2371    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2372        non-existent subpattern following a character class starting with ']' and
2373        containing () gave an internal compiling error instead of "reference to
2374        non-existent subpattern". Fortunately, when the pattern did exist, the
2375        compiled code was correct. (When scanning forwards to check for the
2376        existence of the subpattern, it was treating the data ']' as terminating
2377        the class, so got the count wrong. When actually compiling, the reference
2378        was subsequently set up correctly.)
2380    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2381        it was being rejected as not supported by pcre_dfa_exec(), even though
2382        other assertions are supported. I have made pcre_dfa_exec() support
2383        (*FAIL).
2385    16. The implementation of 13c above involved the invention of a new opcode,
2386        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2387        cannot be changed at match time, I realized I could make a small
2388        improvement to matching performance by compiling OP_ALLANY instead of
2389        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2390        on the OP_ANY path.
2392    17. Compiling pcretest on Windows with readline support failed without the
2393        following two fixes: (1) Make the unistd.h include conditional on
2394        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2396    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2397        ncurses library to be included for pcretest when ReadLine support is
2398        requested, but also to allow for it to be overridden. This patch came from
2399        Daniel Bergström.
2401    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2402        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2403        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2404        spotting this.
2407    Version 7.6 28-Jan-08
2408    ---------------------
2410    1.  A character class containing a very large number of characters with
2411        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2412        overflow.
2414    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2415        HAVE_LONG_LONG is not defined.
2417    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2418        bring them up to date and include new features. This patch includes:
2420        - Fixed PH's badly added libz and libbz2 support.
2421        - Fixed a problem with static linking.
2422        - Added pcredemo. [But later removed - see 7 below.]
2423        - Fixed dftables problem and added an option.
2424        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2425            HAVE_LONG_LONG.
2426        - Added readline support for pcretest.
2427        - Added an listing of the option settings after cmake has run.
2429    4.  A user submitted a patch to Makefile that makes it easy to create
2430        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2431        Makefile.am that cause it to include this special target, without
2432        affecting anything else. Note that the same mingw target plus all
2433        the other distribution libraries and programs are now supported
2434        when configuring with CMake (see 6 below) instead of with
2435        Configure/Make.
2437    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2438        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2439        exported in the Windows port". It has not yet been confirmed that the patch
2440        solves the problem, but it does no harm.
2442    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2443        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2444        with CMake, and also correct the comment about stack recursion.
2446    7.  Remove the automatic building of pcredemo from the ./configure system and
2447        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2448        of a program that users should build themselves after PCRE is installed, so
2449        building it automatically is not really right. What is more, it gave
2450        trouble in some build environments.
2452    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2455    Version 7.5 10-Jan-08
2456  ---------------------  ---------------------
2458  1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'  1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2459      values in parens when parsing an RE using the C++ wrapper."      values in parens when parsing an RE using the C++ wrapper."
2461  2.  Negative specials like \S did not work in character classes in UTF-8 mode.  2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2462      Characters greater than 255 were excluded from the class instead of being      Characters greater than 255 were excluded from the class instead of being
2463      included.      included.
2465  3.  The same bug as (2) above applied to negated POSIX classes such as  3.  The same bug as (2) above applied to negated POSIX classes such as
2466      [:^space:].      [:^space:].
2468  4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it  4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2469      defined or documented. It seems to have been a typo for PCRE_STATIC, so      defined or documented. It seems to have been a typo for PCRE_STATIC, so
2470      I have changed it.      I have changed it.
2472  5.  The construct (?&) was not diagnosed as a syntax error (it referenced the  5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2473      first named subpattern) and a construct such as (?&a) would reference the      first named subpattern) and a construct such as (?&a) would reference the
2474      first named subpattern whose name started with "a" (in other words, the      first named subpattern whose name started with "a" (in other words, the
2475      length check was missing).      length check was missing). Both these problems are fixed. "Subpattern name
2476        expected" is now given for (?&) (a zero-length name), and this patch also
2477        makes it give the same error for \k'' (previously it complained that that
2478        was a reference to a non-existent subpattern).
2480  6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;  6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2481      this is right because (?- can be followed by option settings as well as by      this is right because (?- can be followed by option settings as well as by
2482      digits. I have, however, made the messages clearer.      digits. I have, however, made the messages clearer.
2484  7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns  7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2485      than the number used in the conditional) now cause a compile-time error.      than the number used in the conditional) now cause a compile-time error.
2486      This is actually not compatible with Perl, which accepts such patterns, but      This is actually not compatible with Perl, which accepts such patterns, but
2487      treats the conditional as always being FALSE (as PCRE used to), but it      treats the conditional as always being FALSE (as PCRE used to), but it
2488      seems to me that giving a diagnostic is better.      seems to me that giving a diagnostic is better.
2490    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2491        and messages.
2493    9.  Fix two occurrences of "backslash" in comments that should have been
2494        "backspace".
2496    10. Remove two redundant lines of code that can never be obeyed (their function
2497        was moved elsewhere).
2499    11. The program that makes PCRE's Unicode character property table had a bug
2500        which caused it to generate incorrect table entries for sequences of
2501        characters that have the same character type, but are in different scripts.
2502        It amalgamated them into a single range, with the script of the first of
2503        them. In other words, some characters were in the wrong script. There were
2504        thirteen such cases, affecting characters in the following ranges:
2506          U+002b0 - U+002c1
2507          U+0060c - U+0060d
2508          U+0061e - U+00612
2509          U+0064b - U+0065e
2510          U+0074d - U+0076d
2511          U+01800 - U+01805
2512          U+01d00 - U+01d77
2513          U+01d9b - U+01dbf
2514          U+0200b - U+0200f
2515          U+030fc - U+030fe
2516          U+03260 - U+0327f
2517          U+0fb46 - U+0fbb1
2518          U+10450 - U+1049d
2520    12. The -o option (show only the matching part of a line) for pcregrep was not
2521        compatible with GNU grep in that, if there was more than one match in a
2522        line, it showed only the first of them. It now behaves in the same way as
2523        GNU grep.
2525    13. If the -o and -v options were combined for pcregrep, it printed a blank
2526        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2527        does the same. The return code can be used to tell if there were any
2528        non-matching lines.
2530    14. Added --file-offsets and --line-offsets to pcregrep.
2532    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2533        infinitely looping recursion. The bug was that positive lookaheads were not
2534        being skipped when checking for a possible empty match (negative lookaheads
2535        and both kinds of lookbehind were skipped).
2537    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2538        inclusion of <windows.h> to before rather than after the definition of
2539        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2541    17. Specifying a possessive quantifier with a specific limit for a Unicode
2542        character property caused pcre_compile() to compile bad code, which led at
2543        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2544        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2545        caused the error; without that there was no problem.
2547    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2549    19. Added --enable-pcretest-libreadline.
2551    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2552        RE::GlobalReplace(). As a result, the number of replacements returned was
2553        double what it should be. I removed one of the increments, but Craig sent a
2554        later patch that removed the other one (the right fix) and added unit tests
2555        that check the return values (which was not done before).
2557    21. Several CMake things:
2559        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2560            the names libpcre and libpcreposix, not just pcre and pcreposix.
2562        (2) The above change means that pcretest and pcregrep are now correctly
2563            linked with the newly-built libraries, not previously installed ones.
2567    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2568        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2569        UTF-8 newline character). The key issue is that the pattern starts .*;
2570        this means that the match must be either at the beginning, or after a
2571        newline. The bug was in the code for advancing after a failed match and
2572        checking that the new position followed a newline. It was not taking
2573        account of UTF-8 characters correctly.
2575    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2576        character classes. PCRE was not treating the sequence [:...:] as a
2577        character class unless the ... were all letters. Perl, however, seems to
2578        allow any characters between [: and :], though of course it rejects as
2579        unknown any "names" that contain non-letters, because all the known class
2580        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2581        for example, whereas PCRE did not - it did not recognize a POSIX character
2582        class. This seemed a bit dangerous, so the code has been changed to be
2583        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2584        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2585        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2586        Perl does, and where it didn't before.
2588    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2589        Windows environments %n is disabled by default.
2592  Version 7.4 21-Sep-07  Version 7.4 21-Sep-07
2593  ---------------------  ---------------------

Removed from v.270  
changed lines
  Added in v.1518

  ViewVC Help
Powered by ViewVC 1.1.5