/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 657 by ph10, Mon Aug 15 17:39:09 2011 UTC revision 1510 by ph10, Wed Nov 5 15:08:03 2014 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 8.13 02-Aug-2011  Version 8.37 xx-xxx-201x
5    ------------------------
6    
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".
13    
14    
15    
16    Version 8.36 26-September-2014
17    ------------------------------
18    
19    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
20        -Wmissing-field-initializers and -Wunused-parameter.
21    
22    2.  The tests for quantifiers being too big (greater than 65535) were being
23        applied after reading the number, and stupidly assuming that integer
24        overflow would give a negative number. The tests are now applied as the
25        numbers are read.
26    
27    3.  Tidy code in pcre_exec.c where two branches that used to be different are
28        now the same.
29    
30    4.  The JIT compiler did not generate match limit checks for certain
31        bracketed expressions with quantifiers. This may lead to exponential
32        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
33        issue should be resolved now.
34    
35    5.  Fixed an issue, which occures when nested alternatives are optimized
36        with table jumps.
37    
38    6.  Inserted two casts and changed some ints to size_t in the light of some
39        reported 64-bit compiler warnings (Bugzilla 1477).
40    
41    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
42        an empty string, which sometimes were behaving incorrectly in the
43        interpreter (though correctly in the JIT matcher). This pcretest input is
44        an example:
45    
46          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
47          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
48    
49        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
50        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
51        for an empty string was breaking the inner loop and carrying on at a lower
52        level, when possessive repeated groups should always return to a higher
53        level as they have no backtrack points in them. The empty string test now
54        occurs at the outer level.
55    
56    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
57        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
58    
59    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
60        doesn't).
61    
62    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
63        to hold a pattern to be incorrectly computed (too small) when there were
64        named back references to duplicated names. This could cause "internal
65        error: code overflow" or "double free or corruption" or other memory
66        handling errors.
67    
68    11. When named subpatterns had the same prefixes, back references could be
69        confused. For example, in this pattern:
70    
71          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
72    
73        the reference to 'Name' was incorrectly treated as a reference to a
74        duplicate name.
75    
76    12. A pattern such as /^s?c/mi8 where the optional character has more than
77        one "other case" was incorrectly compiled such that it would only try to
78        match starting at "c".
79    
80    13. When a pattern starting with \s was studied, VT was not included in the
81        list of possible starting characters; this should have been part of the
82        8.34/18 patch.
83    
84    14. If a character class started [\Qx]... where x is any character, the class
85        was incorrectly terminated at the ].
86    
87    15. If a pattern that started with a caseless match for a character with more
88        than one "other case" was studied, PCRE did not set up the starting code
89        unit bit map for the list of possible characters. Now it does. This is an
90        optimization improvement, not a bug fix.
91    
92    16. The Unicode data tables have been updated to Unicode 7.0.0.
93    
94    17. Fixed a number of memory leaks in pcregrep.
95    
96    18. Avoid a compiler warning (from some compilers) for a function call with
97        a cast that removes "const" from an lvalue by using an intermediate
98        variable (to which the compiler does not object).
99    
100    19. Incorrect code was compiled if a group that contained an internal recursive
101        back reference was optional (had quantifier with a minimum of zero). This
102        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
103        caused segmentation faults because of stack overflows at compile time.
104    
105    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
106        group that is quantified with an indefinite repeat, caused a compile-time
107        loop which used up all the system stack and provoked a segmentation fault.
108        This was not the same bug as 19 above.
109    
110    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
111        Patch by Mike Frysinger.
112    
113    
114    Version 8.35 04-April-2014
115    --------------------------
116    
117    1.  A new flag is set, when property checks are present in an XCLASS.
118        When this flag is not set, PCRE can perform certain optimizations
119        such as studying these XCLASS-es.
120    
121    2.  The auto-possessification of character sets were improved: a normal
122        and an extended character set can be compared now. Furthermore
123        the JIT compiler optimizes more character set checks.
124    
125    3.  Got rid of some compiler warnings for potentially uninitialized variables
126        that show up only when compiled with -O2.
127    
128    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
129        of a match later then the end of the match. The pcretest program was not
130        handling the case sensibly - it was outputting from the start to the next
131        binary zero. It now reports this situation in a message, and outputs the
132        text from the end to the start.
133    
134    5.  Fast forward search is improved in JIT. Instead of the first three
135        characters, any three characters with fixed position can be searched.
136        Search order: first, last, middle.
137    
138    6.  Improve character range checks in JIT. Characters are read by an inprecise
139        function now, which returns with an unknown value if the character code is
140        above a certain threshold (e.g: 256). The only limitation is that the value
141        must be bigger than the threshold as well. This function is useful when
142        the characters above the threshold are handled in the same way.
143    
144    7.  The macros whose names start with RAWUCHAR are placeholders for a future
145        mode in which only the bottom 21 bits of 32-bit data items are used. To
146        make this more memorable for those maintaining the code, the names have
147        been changed to start with UCHAR21, and an extensive comment has been added
148        to their definition.
149    
150    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
151        to the export list in Makefile.am (they were accidentally omitted from the
152        8.34 tarball).
153    
154    9.  The informational output from pcretest used the phrase "starting byte set"
155        which is inappropriate for the 16-bit and 32-bit libraries. As the output
156        for "first char" and "need char" really means "non-UTF-char", I've changed
157        "byte" to "char", and slightly reworded the output. The documentation about
158        these values has also been (I hope) clarified.
159    
160    10. Another JIT related optimization: use table jumps for selecting the correct
161        backtracking path, when more than four alternatives are present inside a
162        bracket.
163    
164    11. Empty match is not possible, when the minimum length is greater than zero,
165        and there is no \K in the pattern. JIT should avoid empty match checks in
166        such cases.
167    
168    12. In a caseless character class with UCP support, when a character with more
169        than one alternative case was not the first character of a range, not all
170        the alternative cases were added to the class. For example, s and \x{17f}
171        are both alternative cases for S: the class [RST] was handled correctly,
172        but [R-T] was not.
173    
174    13. The configure.ac file always checked for pthread support when JIT was
175        enabled. This is not used in Windows, so I have put this test inside a
176        check for the presence of windows.h (which was already tested for).
177    
178    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
179        The algorithm provides a way to skip certain starting offsets, and usually
180        faster than linear prefix searches.
181    
182    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
183        as for 'fr_FR' and 'french'. For some reason, however, it then used the
184        Windows-specific input and output files, which have 'french' screwed in.
185        So this could never have worked. One of the problems with locales is that
186        they aren't always the same. I have now updated RunTest so that it checks
187        the output of the locale test (test 3) against three different output
188        files, and it allows the test to pass if any one of them matches. With luck
189        this should make the test pass on some versions of Solaris where it was
190        failing. Because of the uncertainty, the script did not used to stop if
191        test 3 failed; it now does. If further versions of a French locale ever
192        come to light, they can now easily be added.
193    
194    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
195        there was a message during ./configure, but it did not stop. This now
196        provokes an error. The invalid example in README has been corrected.
197        If a value less than the minimum is given, the minimum value has always
198        been used, but now a warning is given.
199    
200    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
201        was a bug in the test system, which is now fixed. Also, the list of various
202        configurations that are tested for each release did not have one with both
203        16/32 bits and --enable-bar-anycrlf. It now does.
204    
205    18. pcretest was missing "-C bsr" for displaying the \R default setting.
206    
207    19. Little endian PowerPC systems are supported now by the JIT compiler.
208    
209    20. The fast forward newline mechanism could enter to an infinite loop on
210        certain invalid UTF-8 input. Although we don't support these cases
211        this issue can be fixed by a performance optimization.
212    
213    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
214        not take account if existing stack usage. There is now a new global
215        variable called pcre_stack_guard that can be set to point to an external
216        function to check stack availability. It is called at the start of
217        processing every parenthesized group.
218    
219    22. A typo in the code meant that in ungreedy mode the max/min qualifier
220        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
221        not match "ab".
222    
223    23. When UTF was disabled, the JIT program reported some incorrect compile
224        errors. These messages are silenced now.
225    
226    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
227        compiler.
228    
229    25. Change all the temporary files used in RunGrepTest to be different to those
230        used by RunTest so that the tests can be run simultaneously, for example by
231        "make -j check".
232    
233    
234    Version 8.34 15-December-2013
235    -----------------------------
236    
237    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
238        executable memory. Patch inspired by Carsten Klein.
239    
240    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
241        this macro is never tested and has no effect, because the work to support
242        coverage involves only compiling and linking options and special targets in
243        the Makefile. The comment in config.h implied that defining the macro would
244        enable coverage support, which is totally false. There was also support for
245        setting this macro in the CMake files (my fault, I just copied it from
246        configure). SUPPORT_GCOV has now been removed.
247    
248    3.  Make a small performance improvement in strlen16() and strlen32() in
249        pcretest.
250    
251    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
252        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
253        these cases). There was also one in pcretest.c.
254    
255    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
256    
257    6.  In UTF mode, the code for checking whether a group could match an empty
258        string (which is used for indefinitely repeated groups to allow for
259        breaking an infinite loop) was broken when the group contained a repeated
260        negated single-character class with a character that occupied more than one
261        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
262        UTF-8 mode). The effect was undefined: the group might or might not be
263        deemed as matching an empty string, or the program might have crashed.
264    
265    7.  The code for checking whether a group could match an empty string was not
266        recognizing that \h, \H, \v, \V, and \R must match a character.
267    
268    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
269        an empty string. If it can, pcretest shows this in its information output.
270    
271    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
272        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
273        matched by pcre_exec() without using JIT:
274    
275        (a) If the rest of the pattern did not match after a maximal run of
276            grapheme clusters, the code for backing up to try with fewer of them
277            did not always back up over a full grapheme when characters that do not
278            have the modifier quality were involved, e.g. Hangul syllables.
279    
280        (b) If the match point in a subject started with modifier character, and
281            there was no match, the code could incorrectly back up beyond the match
282            point, and potentially beyond the first character in the subject,
283            leading to a segfault or an incorrect match result.
284    
285    10. A conditional group with an assertion condition could lead to PCRE
286        recording an incorrect first data item for a match if no other first data
287        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
288        first data item, and therefore matched "ca" after "c" instead of at the
289        start.
290    
291    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
292        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
293    
294    12. The source of pcregrep now includes z/OS-specific code so that it can be
295        compiled for z/OS as part of the special z/OS distribution.
296    
297    13. Added the -T and -TM options to pcretest.
298    
299    14. The code in pcre_compile.c for creating the table of named capturing groups
300        has been refactored. Instead of creating the table dynamically during the
301        actual compiling pass, the information is remembered during the pre-compile
302        pass (on the stack unless there are more than 20 named groups, in which
303        case malloc() is used) and the whole table is created before the actual
304        compile happens. This has simplified the code (it is now nearly 150 lines
305        shorter) and prepared the way for better handling of references to groups
306        with duplicate names.
307    
308    15. A back reference to a named subpattern when there is more than one of the
309        same name now checks them in the order in which they appear in the pattern.
310        The first one that is set is used for the reference. Previously only the
311        first one was inspected. This change makes PCRE more compatible with Perl.
312    
313    16. Unicode character properties were updated from Unicode 6.3.0.
314    
315    17. The compile-time code for auto-possessification has been refactored, based
316        on a patch by Zoltan Herczeg. It now happens after instead of during
317        compilation. The code is cleaner, and more cases are handled. The option
318        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
319        options in pcretest are provided to set it. It can also be set by
320        (*NO_AUTO_POSSESS) at the start of a pattern.
321    
322    18. The character VT has been added to the default ("C" locale) set of
323        characters that match \s and are generally treated as white space,
324        following this same change in Perl 5.18. There is now no difference between
325        "Perl space" and "POSIX space". Whether VT is treated as white space in
326        other locales depends on the locale.
327    
328    19. The code for checking named groups as conditions, either for being set or
329        for being recursed, has been refactored (this is related to 14 and 15
330        above). Processing unduplicated named groups should now be as fast at
331        numerical groups, and processing duplicated groups should be faster than
332        before.
333    
334    20. Two patches to the CMake build system, by Alexander Barkov:
335    
336          (1) Replace the "source" command by "." in CMakeLists.txt because
337              "source" is a bash-ism.
338    
339          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
340              without these the CMake build does not work on Solaris.
341    
342    21. Perl has changed its handling of \8 and \9. If there is no previously
343        encountered capturing group of those numbers, they are treated as the
344        literal characters 8 and 9 instead of a binary zero followed by the
345        literals. PCRE now does the same.
346    
347    22. Following Perl, added \o{} to specify codepoints in octal, making it
348        possible to specify values greater than 0777 and also making them
349        unambiguous.
350    
351    23. Perl now gives an error for missing closing braces after \x{... instead of
352        treating the string as literal. PCRE now does the same.
353    
354    24. RunTest used to grumble if an inappropriate test was selected explicitly,
355        but just skip it when running all tests. This make it awkward to run ranges
356        of tests when one of them was inappropriate. Now it just skips any
357        inappropriate tests, as it always did when running all tests.
358    
359    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
360        character types such as \d or \w, too many callouts were inserted, and the
361        data that they returned was rubbish.
362    
363    26. In UCP mode, \s was not matching two of the characters that Perl matches,
364        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
365        were matched by \h. The code has now been refactored so that the lists of
366        the horizontal and vertical whitespace characters used for \h and \v (which
367        are defined only in one place) are now also used for \s.
368    
369    27. Add JIT support for the 64 bit TileGX architecture.
370        Patch by Jiong Wang (Tilera Corporation).
371    
372    28. Possessive quantifiers for classes (both explicit and automatically
373        generated) now use special opcodes instead of wrapping in ONCE brackets.
374    
375    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
376        (because it's meaningless), this was not happening when PCRE_CASELESS was
377        set. Not wrong, but inefficient.
378    
379    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
380        properties for \w, \d, etc) is present in a test regex. Otherwise if the
381        test contains no characters greater than 255, Perl doesn't realise it
382        should be using Unicode semantics.
383    
384    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
385        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
386        does in Unicode mode.
387    
388    32. Added the "forbid" facility to pcretest so that putting tests into the
389        wrong test files can sometimes be quickly detected.
390    
391    33. There is now a limit (default 250) on the depth of nesting of parentheses.
392        This limit is imposed to control the amount of system stack used at compile
393        time. It can be changed at build time by --with-parens-nest-limit=xxx or
394        the equivalent in CMake.
395    
396    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
397        errors. Perl warns for these when in warning mode, but PCRE has no facility
398        for giving warnings.
399    
400    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
401        However, this was not working for (?!) because it is optimized to (*FAIL),
402        for which PCRE does not allow quantifiers. The optimization is now disabled
403        when a quantifier follows (?!). I can't see any use for this, but it makes
404        things uniform.
405    
406    36. Perl no longer allows group names to start with digits, so I have made this
407        change also in PCRE. It simplifies the code a bit.
408    
409    37. In extended mode, Perl ignores spaces before a + that indicates a
410        possessive quantifier. PCRE allowed a space before the quantifier, but not
411        before the possessive +. It now does.
412    
413    38. The use of \K (reset reported match start) within a repeated possessive
414        group such as (a\Kb)*+ was not working.
415    
416    40. Document that the same character tables must be used at compile time and
417        run time, and that the facility to pass tables to pcre_exec() and
418        pcre_dfa_exec() is for use only with saved/restored patterns.
419    
420    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
421        features for Builds with MSVC:
422    
423        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
424           builds). The .rc files can be used to set FileDescription and many other
425           attributes.
426    
427        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
428           This allows higher-level build scripts which want .pdb files to avoid
429           hard-coding the exact files needed."
430    
431    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
432        mean "start of word" and "end of word", respectively, as a transition aid.
433    
434    43. A minimizing repeat of a class containing codepoints greater than 255 in
435        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
436        compiled to use the heap for recursion.
437    
438    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
439        is configured.
440    
441    
442    Version 8.33 28-May-2013
443    ------------------------
444    
445    1.  Added 'U' to some constants that are compared to unsigned integers, to
446        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
447        variables that are added to signed variables, to ensure the result is
448        signed and can be negated.
449    
450    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
451        CMake config files.
452    
453    3.  Revise the creation of config.h.generic so that all boolean macros are
454        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
455        overriding via -D on the command line possible.
456    
457    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
458        to unsigned int is reported to make a quite noticeable speed difference in
459        a specific Windows environment. Testing on Linux did also appear to show
460        some benefit (and it is clearly not harmful). Also fixed the definition of
461        Xop which should be unsigned.
462    
463    5.  Related to (4), changing the definition of the intermediate variable cc
464        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
465        improvements.
466    
467    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
468        unnecessary spaces.
469    
470    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
471        and later.
472    
473    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
474    
475    9.  Optimizing fast_forward_start_bits in JIT.
476    
477    10. Adding support for callouts in JIT, and fixing some issues revealed
478        during this work. Namely:
479    
480        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
481    
482        (b) Minimum length was not checked before the matching is started.
483    
484    11. The value of capture_last that is passed to callouts was incorrect in some
485        cases when there was a capture on one path that was subsequently abandoned
486        after a backtrack. Also, the capture_last value is now reset after a
487        recursion, since all captures are also reset in this case.
488    
489    12. The interpreter no longer returns the "too many substrings" error in the
490        case when an overflowing capture is in a branch that is subsequently
491        abandoned after a backtrack.
492    
493    13. In the pathological case when an offset vector of size 2 is used, pcretest
494        now prints out the matched string after a yield of 0 or 1.
495    
496    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
497        Only supported by the JIT compiler at the moment.
498    
499    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
500    
501    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
502        offset of the starting point of the matching process, provided the offsets
503        vector is large enough.
504    
505    17. The \A escape now records a lookbehind value of 1, though its execution
506        does not actually inspect the previous character. This is to ensure that,
507        in partial multi-segment matching, at least one character from the old
508        segment is retained when a new segment is processed. Otherwise, if there
509        are no lookbehinds in the pattern, \A might match incorrectly at the start
510        of a new segment.
511    
512    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
513    
514    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
515        gives some modest performance improvement in 8-bit mode.
516    
517    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
518        be expressed in certain programming languages using Universal Character
519        Names.
520    
521    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
522        which points out that "non characters" are not "characters that may not
523        appear in Unicode strings" but rather "characters that are reserved for
524        internal use and have only local meaning".
525    
526    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
527        there was a conditional group that depended on an assertion, if the
528        assertion was false, the callout that immediately followed the alternation
529        in the condition was skipped when pcre_exec() was used for matching.
530    
531    23. Allow an explicit callout to be inserted before an assertion that is the
532        condition for a conditional group, for compatibility with automatic
533        callouts, which always insert a callout at this point.
534    
535    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
536        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
537    
538    25. (*PRUNE) is now supported by the JIT compiler.
539    
540    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
541    
542    27. Fix the case where there are two or more SKIPs with arguments that may be
543        ignored.
544    
545    28. (*SKIP) is now supported by the JIT compiler.
546    
547    29. (*THEN) is now supported by the JIT compiler.
548    
549    30. Update RunTest with additional test selector options.
550    
551    31. The way PCRE handles backtracking verbs has been changed in two ways.
552    
553        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
554        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
555        backtracking. In some cases this makes it more Perl-compatible, but Perl's
556        rather obscure rules do not always do the same thing.
557    
558        (2) Previously, backtracking verbs were confined within assertions. This is
559        no longer the case for positive assertions, except for (*ACCEPT). Again,
560        this sometimes improves Perl compatibility, and sometimes does not.
561    
562    32. A number of tests that were in test 2 because Perl did things differently
563        have been moved to test 1, because either Perl or PCRE has changed, and
564        these tests are now compatible.
565    
566    32. Backtracking control verbs are now handled in the same way in JIT and
567        interpreter.
568    
569    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
570        contained a forward subroutine reference caused a compile error.
571    
572    34. Auto-detect and optimize limited repetitions in JIT.
573    
574    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
575        blocking (*UTF) etc.
576    
577    36. In the interpreter, maximizing pattern repetitions for characters and
578        character types now use tail recursion, which reduces stack usage.
579    
580    37. The value of the max lookbehind was not correctly preserved if a compiled
581        and saved regex was reloaded on a host of different endianness.
582    
583    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
584        of the compiled pattern block, expand the flags field from 16 to 32 bits
585        because it was almost full.
586    
587    39. Try madvise first before posix_madvise.
588    
589    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
590        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
591        original change locked out all matches of empty strings. This has been
592        changed so that one match of an empty string per line is recognized.
593        Subsequent searches on the same line (for colouring or for --only-matching,
594        for example) do not recognize empty strings.
595    
596    41. Applied a user patch to fix a number of spelling mistakes in comments.
597    
598    42. Data lines longer than 65536 caused pcretest to crash.
599    
600    43. Clarified the data type for length and startoffset arguments for pcre_exec
601        and pcre_dfa_exec in the function-specific man pages, where they were
602        explicitly stated to be in bytes, never having been updated. I also added
603        some clarification to the pcreapi man page.
604    
605    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
606        a segmentation fault.
607    
608    
609    Version 8.32 30-November-2012
610    -----------------------------
611    
612    1.  Improved JIT compiler optimizations for first character search and single
613        character iterators.
614    
615    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
616        Patch by Daniel Richard G.
617    
618    3.  Single character iterator optimizations in the JIT compiler.
619    
620    4.  Improved JIT compiler optimizations for character ranges.
621    
622    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
623        Reported by Giuseppe D'Angelo.
624    
625    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
626        of a line, was being set incorrectly in cases where .* appeared inside
627        atomic brackets at the start of a pattern, or where there was a subsequent
628        *PRUNE or *SKIP.
629    
630    7.  Improved instruction cache flush for POWER/PowerPC.
631        Patch by Daniel Richard G.
632    
633    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
634        grep:
635    
636        (a) There is now no limit to the number of patterns to be matched.
637    
638        (b) An error is given if a pattern is too long.
639    
640        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
641            are now supported.
642    
643        (d) --exclude-from and --include-from (multiple use) have been added.
644    
645        (e) Exclusions and inclusions now apply to all files and directories, not
646            just to those obtained from scanning a directory recursively.
647    
648        (f) Multiple uses of -f and --file-list are now supported.
649    
650        (g) In a Windows environment, the default for -d has been changed from
651            "read" (the GNU grep default) to "skip", because otherwise the presence
652            of a directory in the file list provokes an error.
653    
654        (h) The documentation has been revised and clarified in places.
655    
656    9.  Improve the matching speed of capturing brackets.
657    
658    10. Changed the meaning of \X so that it now matches a Unicode extended
659        grapheme cluster.
660    
661    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
662        out POSIX threads when JIT support is configured.
663    
664    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
665    
666    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
667        structure could go wrong in environments where size_t is not the same size
668        as int.
669    
670    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
671        set.
672    
673    15. The EBCDIC support had decayed; later updates to the code had included
674        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
675        general tidy up of EBCDIC-related issues, and the documentation was also
676        not quite right. There is now a test that can be run on ASCII systems to
677        check some of the EBCDIC-related things (but is it not a full test).
678    
679    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
680        in a small tidy to the code.
681    
682    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
683    
684    18. If the --only-matching (-o) option in pcregrep is specified multiple
685        times, each one causes appropriate output. For example, -o1 -o2 outputs the
686        substrings matched by the 1st and 2nd capturing parentheses. A separating
687        string can be specified by --om-separator (default empty).
688    
689    19. Improving the first n character searches.
690    
691    20. Turn case lists for horizontal and vertical white space into macros so that
692        they are defined only once.
693    
694    21. This set of changes together give more compatible Unicode case-folding
695        behaviour for characters that have more than one other case when UCP
696        support is available.
697    
698        (a) The Unicode property table now has offsets into a new table of sets of
699            three or more characters that are case-equivalent. The MultiStage2.py
700            script that generates these tables (the pcre_ucd.c file) now scans
701            CaseFolding.txt instead of UnicodeData.txt for character case
702            information.
703    
704        (b) The code for adding characters or ranges of characters to a character
705            class has been abstracted into a generalized function that also handles
706            case-independence. In UTF-mode with UCP support, this uses the new data
707            to handle characters with more than one other case.
708    
709        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
710            whose other case is greater than 256 are now correctly matched
711            caselessly. Previously, the high codepoint matched the low one, but not
712            vice versa.
713    
714        (d) The processing of \h, \H, \v, and \ in character classes now makes use
715            of the new class addition function, using character lists defined as
716            macros alongside the case definitions of 20 above.
717    
718        (e) Caseless back references now work with characters that have more than
719            one other case.
720    
721        (f) General caseless matching of characters with more than one other case
722            is supported.
723    
724    22. Unicode character properties were updated from Unicode 6.2.0
725    
726    23. Improved CMake support under Windows. Patch by Daniel Richard G.
727    
728    24. Add support for 32-bit character strings, and UTF-32
729    
730    25. Major JIT compiler update (code refactoring and bugfixing).
731        Experimental Sparc 32 support is added.
732    
733    26. Applied a modified version of Daniel Richard G's patch to create
734        pcre.h.generic and config.h.generic by "make" instead of in the
735        PrepareRelease script.
736    
737    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
738        pcre_compile.c when checking for a zero character.
739    
740    28. Introducing a native interface for JIT. Through this interface, the compiled
741        machine code can be directly executed. The purpose of this interface is to
742        provide fast pattern matching, so several sanity checks are not performed.
743        However, feature tests are still performed. The new interface provides
744        1.4x speedup compared to the old one.
745    
746    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
747        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
748        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
749    
750    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
751        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
752        the "old" RFC 2279). Instead, it ended up passing a negative length to
753        pcre_exec().
754    
755    31. Add support for GCC's visibility feature to hide internal functions.
756    
757    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
758        "unknown -C option" after outputting 0 or 1.
759    
760    33. There is now support for generating a code coverage report for the test
761        suite in environments where gcc is the compiler and lcov is installed. This
762        is mainly for the benefit of the developers.
763    
764    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
765        unaddressable using valgrind annotations, allowing valgrind to detect
766        invalid memory accesses. This is mainly for the benefit of the developers.
767    
768    25. (*UTF) can now be used to start a pattern in any of the three libraries.
769    
770    26. Give configure error if --enable-cpp but no C++ compiler found.
771    
772    
773    Version 8.31 06-July-2012
774    -------------------------
775    
776    1.  Fixing a wrong JIT test case and some compiler warnings.
777    
778    2.  Removed a bashism from the RunTest script.
779    
780    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
781        to unsigned type, result still unsigned" that was given by an MS compiler
782        on encountering the code "-sizeof(xxx)".
783    
784    4.  Partial matching support is added to the JIT compiler.
785    
786    5.  Fixed several bugs concerned with partial matching of items that consist
787        of more than one character:
788    
789        (a) /^(..)\1/ did not partially match "aba" because checking references was
790            done on an "all or nothing" basis. This also applied to repeated
791            references.
792    
793        (b) \R did not give a hard partial match if \r was found at the end of the
794            subject.
795    
796        (c) \X did not give a hard partial match after matching one or more
797            characters at the end of the subject.
798    
799        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
800            a partial match for the string "\r".
801    
802        (e) When newline was set to CRLF, the metacharacter "." did not recognize
803            a partial match for a CR character at the end of the subject string.
804    
805    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
806        running pcretest, the text "(JIT)" added to the output whenever JIT is
807        actually used to run the match.
808    
809    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
810        or /S+[+] with a digit between 1 and 7.
811    
812    8.  OP_NOT now supports any UTF character not just single-byte ones.
813    
814    9.  (*MARK) control verb is now supported by the JIT compiler.
815    
816    10. The command "./RunTest list" lists the available tests without actually
817        running any of them. (Because I keep forgetting what they all are.)
818    
819    11. Add PCRE_INFO_MAXLOOKBEHIND.
820    
821    12. Applied a (slightly modified) user-supplied patch that improves performance
822        when the heap is used for recursion (compiled with --disable-stack-for-
823        recursion). Instead of malloc and free for each heap frame each time a
824        logical recursion happens, frames are retained on a chain and re-used where
825        possible. This sometimes gives as much as 30% improvement.
826    
827    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
828        call.
829    
830    14. As documented, (*COMMIT) is now confined to within a positive assertion.
831    
832    15. It is now possible to link pcretest with libedit as an alternative to
833        libreadline.
834    
835    16. (*COMMIT) control verb is now supported by the JIT compiler.
836    
837    17. The Unicode data tables have been updated to Unicode 6.1.0.
838    
839    18. Added --file-list option to pcregrep.
840    
841    19. Added binary file support to pcregrep, including the -a, --binary-files,
842        -I, and --text options.
843    
844    20. The madvise function is renamed for posix_madvise for QNX compatibility
845        reasons. Fixed by Giuseppe D'Angelo.
846    
847    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
848    
849    22. Changed the option for creating symbolic links for 16-bit man pages from
850        -s to -sf so that re-installing does not cause issues.
851    
852    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
853    
854    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
855        matches in certain environments (the workspace was not being correctly
856        retained). Also added to pcre_dfa_exec() a simple plausibility check on
857        some of the workspace data at the beginning of a restart.
858    
859    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
860        was not doing so when it should - probably a typo introduced by SVN 528
861        (change 8.10/14).
862    
863    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
864        \w+ when the character tables indicated that \x{c4} was a word character.
865        There were several related cases, all because the tests for doing a table
866        lookup were testing for characters less than 127 instead of 255.
867    
868    27. If a pattern contains capturing parentheses that are not used in a match,
869        their slots in the ovector are set to -1. For those that are higher than
870        any matched groups, this happens at the end of processing. In the case when
871        there were back references that the ovector was too small to contain
872        (causing temporary malloc'd memory to be used during matching), and the
873        highest capturing number was not used, memory off the end of the ovector
874        was incorrectly being set to -1. (It was using the size of the temporary
875        memory instead of the true size.)
876    
877    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
878        ovector size, it uses memory at the end of the block that it has got.
879    
880    29. Check for an overlong MARK name and give an error at compile time. The
881        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
882    
883    30. JIT compiler update.
884    
885    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
886        Rill for the patch.
887    
888    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
889    
890    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
891    
892    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
893        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
894        was enabled.
895    
896    35. Improve JIT code generation for greedy plus quantifier.
897    
898    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
899        1 to "aa" instead of to an empty string. The bug affected repeated groups
900        that could potentially match an empty string.
901    
902    37. Optimizing single character iterators in JIT.
903    
904    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
905        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
906        codepoints that are too big for the mode are faulted, and in a UTF mode,
907        disallowed codepoints are also faulted.
908    
909    39. If PCRE was compiled with UTF support, in three places in the DFA
910        matcher there was code that should only have been obeyed in UTF mode, but
911        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
912        processing when bytes with values greater than 127 were present. In 16-bit
913        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
914        both cases the values are those that cannot be the first data item in a UTF
915        character. The three items that might have provoked this were recursions,
916        possessively repeated groups, and atomic groups.
917    
918    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
919        and pcregrep, because some OS require shared objects to be explicitly
920        passed to ld, causing the link step to fail if they are not.
921    
922    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
923        mode, patterns that started with \h* or \R* might be incorrectly matched.
924    
925    
926    Version 8.30 04-February-2012
927    -----------------------------
928    
929    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
930        name is defined in ctype.h.
931    
932    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
933        only in quite long subpatterns.
934    
935    3.  Removed the function pcre_info(), which has been obsolete and deprecated
936        since it was replaced by pcre_fullinfo() in February 2000.
937    
938    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
939        match a (*MARK), and the match failed at the start of the subject, a
940        reference to memory before the start of the subject could occur. This bug
941        was introduced by fix 17 of release 8.21.
942    
943    5.  A reference to an unset group with zero minimum repetition was giving
944        totally wrong answers (in non-JavaScript-compatibility mode). For example,
945        /(another)?(\1?)test/ matched against "hello world test". This bug was
946        introduced in release 8.13.
947    
948    6.  Add support for 16-bit character strings (a large amount of work involving
949        many changes and refactorings).
950    
951    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
952        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
953        from a file.
954    
955    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
956        rounding is not applied in this particular case).
957    
958    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
959        if they appear, or are escaped, in patterns.
960    
961    10. Get rid of a number of -Wunused-but-set-variable warnings.
962    
963    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
964        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
965        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
966        also returns the mark "x". This bug applied to capturing parentheses,
967        non-capturing parentheses, and atomic parentheses. It also applied to some
968        assertions.
969    
970    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
971        information out of configure.ac instead of relying on pcre.h.generic, which
972        is not stored in the repository.
973    
974    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
975        -lreadline.
976    
977    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
978    
979    15. Applied Graycode's patch to put the top-level frame on the stack rather
980        than the heap when not using the stack for recursion. This gives a
981        performance improvement in many cases when recursion is not deep.
982    
983    16. Experimental code added to "pcretest -C" to output the stack frame size.
984    
985    
986    Version 8.21 12-Dec-2011
987    ------------------------
988    
989    1.  Updating the JIT compiler.
990    
991    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
992        are added as well.
993    
994    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
995        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
996        calling _pcre_jit_exec. Some extra comments are added.
997    
998    4.  (*MARK) settings inside atomic groups that do not contain any capturing
999        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1000        was introduced by change 18 for 8.20.
1001    
1002    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1003        ECMA-262 standard.
1004    
1005    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1006        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1007        This bug was probably introduced by change 9 of 8.13.
1008    
1009    7.  While fixing 6 above, I noticed that a number of other items were being
1010        incorrectly rejected as "not fixed length". This arose partly because newer
1011        opcodes had not been added to the fixed-length checking code. I have (a)
1012        corrected the bug and added tests for these items, and (b) arranged for an
1013        error to occur if an unknown opcode is encountered while checking for fixed
1014        length instead of just assuming "not fixed length". The items that were
1015        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1016        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1017        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1018    
1019    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1020        being incorrectly compiled and would have given unpredicatble results.
1021    
1022    9.  A possessively repeated subpattern with minimum repeat count greater than
1023        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1024        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1025        the first (A) could occur when it should not.
1026    
1027    10. Add a cast and remove a redundant test from the code.
1028    
1029    11. JIT should use pcre_malloc/pcre_free for allocation.
1030    
1031    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1032        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1033        is anything other than /usr, -L is still shown).
1034    
1035    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1036    
1037    14. Perl does not support \N without a following name in a [] class; PCRE now
1038        also gives an error.
1039    
1040    15. If a forward reference was repeated with an upper limit of around 2000,
1041        it caused the error "internal error: overran compiling workspace". The
1042        maximum number of forward references (including repeats) was limited by the
1043        internal workspace, and dependent on the LINK_SIZE. The code has been
1044        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1045        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1046        of around 200,000 forward references. While doing this, I also speeded up
1047        the filling in of repeated forward references.
1048    
1049    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1050        incorrectly expecting the subject to contain another "a" after the start.
1051    
1052    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1053        in the match, the SKIP should be ignored. This was not happening; instead
1054        the SKIP was being treated as NOMATCH. For patterns such as
1055        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1056        tested.
1057    
1058    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1059        now much more compatible with Perl, in particular in cases where the result
1060        is a non-match for a non-anchored pattern. For example, if
1061        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1062        "m", where previously it did not return a name. A side effect of this
1063        change is that for partial matches, the last encountered mark name is
1064        returned, as for non matches. A number of tests that were previously not
1065        Perl-compatible have been moved into the Perl-compatible test files. The
1066        refactoring has had the pleasing side effect of removing one argument from
1067        the match() function, thus reducing its stack requirements.
1068    
1069    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1070        subsequent uses of /S (without +) incorrectly behaved like /S+.
1071    
1072    21. Retrieve executable code size support for the JIT compiler and fixing
1073        some warnings.
1074    
1075    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1076        not work when the shorter character appeared right at the end of the
1077        subject string.
1078    
1079    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1080        systems.
1081    
1082    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1083        output it when the /M option is used in pcretest.
1084    
1085    25. The CheckMan script was not being included in the distribution. Also, added
1086        an explicit "perl" to run Perl scripts from the PrepareRelease script
1087        because this is reportedly needed in Windows.
1088    
1089    26. If study data was being save in a file and studying had not found a set of
1090        "starts with" bytes for the pattern, the data written to the file (though
1091        never used) was taken from uninitialized memory and so caused valgrind to
1092        complain.
1093    
1094    27. Updated RunTest.bat as provided by Sheri Pierce.
1095    
1096    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1097    
1098    29. Computation of memory usage for the table of capturing group names was
1099        giving an unnecessarily large value.
1100    
1101    
1102    Version 8.20 21-Oct-2011
1103    ------------------------
1104    
1105    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1106        a POSIX class. After further experiments with Perl, which convinced me that
1107        Perl has bugs and confusions, a closing square bracket is no longer allowed
1108        in a POSIX name. This bug also affected patterns with classes that started
1109        with full stops.
1110    
1111    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1112        captured substring, but while checking the failing first alternative,
1113        substring 1 is temporarily captured. If the output vector supplied to
1114        pcre_exec() was not big enough for this capture, the yield of the function
1115        was still zero ("insufficient space for captured substrings"). This cannot
1116        be totally fixed without adding another stack variable, which seems a lot
1117        of expense for a edge case. However, I have improved the situation in cases
1118        such as /(a)(b)x|abc/ matched against "abc", where the return code
1119        indicates that fewer than the maximum number of slots in the ovector have
1120        been set.
1121    
1122    3.  Related to (2) above: when there are more back references in a pattern than
1123        slots in the output vector, pcre_exec() uses temporary memory during
1124        matching, and copies in the captures as far as possible afterwards. It was
1125        using the entire output vector, but this conflicts with the specification
1126        that only 2/3 is used for passing back captured substrings. Now it uses
1127        only the first 2/3, for compatibility. This is, of course, another edge
1128        case.
1129    
1130    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1131        main code base, and can be used by building with --enable-jit. When this is
1132        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1133        runtime --no-jit option is given.
1134    
1135    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1136        ovector, the return from the function was zero, implying that there were
1137        other matches that did not fit. The correct "exactly full" value is now
1138        returned.
1139    
1140    6.  If a subpattern that was called recursively or as a subroutine contained
1141        (*PRUNE) or any other control that caused it to give a non-standard return,
1142        invalid errors such as "Error -26 (nested recursion at the same subject
1143        position)" or even infinite loops could occur.
1144    
1145    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1146        computing the minimum length on reaching *ACCEPT, and so ended up with the
1147        wrong value of 1 rather than 0. Further investigation indicates that
1148        computing a minimum subject length in the presence of *ACCEPT is difficult
1149        (think back references, subroutine calls), and so I have changed the code
1150        so that no minimum is registered for a pattern that contains *ACCEPT.
1151    
1152    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1153        it was not handled as intended. [But see 16 below.]
1154    
1155    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1156        Sheri Pierce.
1157    
1158    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1159        the first byte in a match must be "a".
1160    
1161    11. Change 17 for 8.13 increased the recursion depth for patterns like
1162        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1163        pattern contains any instances of (*THEN). If it does not, the old
1164        optimizations are restored. It would be nice to do this on a per-group
1165        basis, but at the moment that is not feasible.
1166    
1167    12. In some environments, the output of pcretest -C is CRLF terminated. This
1168        broke RunTest's code that checks for the link size. A single white space
1169        character after the value is now allowed for.
1170    
1171    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1172        For "fr", it uses the Windows-specific input and output files.
1173    
1174    14. If (*THEN) appeared in a group that was called recursively or as a
1175        subroutine, it did not work as intended. [But see next item.]
1176    
1177    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1178        pattern fragments (but not containing any | characters). If A and B are
1179        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1180        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1181        D. In other words, Perl considers parentheses that do not contain any |
1182        characters to be part of a surrounding alternative, whereas PCRE was
1183        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1184        differently. PCRE now behaves in the same way as Perl, except in the case
1185        of subroutine/recursion calls such as (?1) which have in any case always
1186        been different (but PCRE had them first :-).
1187    
1188    16. Related to 15 above: Perl does not treat the | in a conditional group as
1189        creating alternatives. Such a group is treated in the same way as an
1190        ordinary group without any | characters when processing (*THEN). PCRE has
1191        been changed to match Perl's behaviour.
1192    
1193    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1194        RunGrepTest script failed.
1195    
1196    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1197        inevitable for groups that contain captures, but it can lead to a lot of
1198        stack use in large patterns. The old behaviour has been restored for atomic
1199        groups that do not contain any capturing parentheses.
1200    
1201    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1202        suppress the check for a minimum subject length at run time. (If it was
1203        given to pcre_exec() or pcre_dfa_exec() it did work.)
1204    
1205    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1206        fail to work when decoding hex characters in data strings in EBCDIC
1207        environments.
1208    
1209    21. It appears that in at least one Mac OS environment, the isxdigit() function
1210        is implemented as a macro that evaluates to its argument more than once,
1211        contravening the C 90 Standard (I haven't checked a later standard). There
1212        was an instance in pcretest which caused it to go wrong when processing
1213        \x{...} escapes in subject strings. The has been rewritten to avoid using
1214        things like p++ in the argument of isxdigit().
1215    
1216    
1217    Version 8.13 16-Aug-2011
1218  ------------------------  ------------------------
1219    
1220  1.  The Unicode data tables have been updated to Unicode 6.0.0.  1.  The Unicode data tables have been updated to Unicode 6.0.0.
# Line 98  Version 8.13 02-Aug-2011 Line 1311  Version 8.13 02-Aug-2011
1311      tail recursion to cut down on stack usage. Unfortunately, now that there is      tail recursion to cut down on stack usage. Unfortunately, now that there is
1312      the possibility of (*THEN) occurring in these branches, tail recursion is      the possibility of (*THEN) occurring in these branches, tail recursion is
1313      no longer possible because the return has to be checked for (*THEN). These      no longer possible because the return has to be checked for (*THEN). These
1314      two optimizations have therefore been removed.      two optimizations have therefore been removed. [But see 8.20/11 above.]
1315    
1316  18. If a pattern containing \R was studied, it was assumed that \R always  18. If a pattern containing \R was studied, it was assumed that \R always
1317      matched two bytes, thus causing the minimum subject length to be      matched two bytes, thus causing the minimum subject length to be
# Line 189  Version 8.13 02-Aug-2011 Line 1402  Version 8.13 02-Aug-2011
1402      For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,      For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1403      unescaped square brackets may also appear as part of class names. For      unescaped square brackets may also appear as part of class names. For
1404      example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves      example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1405      more like Perl.      more like Perl. (But see 8.20/1 above.)
1406    
1407  38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this  38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1408      was because it thought it was \N{name}, which is not supported).      was because it thought it was \N{name}, which is not supported).
# Line 224  Version 8.13 02-Aug-2011 Line 1437  Version 8.13 02-Aug-2011
1437      (a) The default value of the buffer size parameter has been increased from      (a) The default value of the buffer size parameter has been increased from
1438          8K to 20K. (The actual buffer used is three times this size.)          8K to 20K. (The actual buffer used is three times this size.)
1439    
1440      (b) The default can be changed by ./configure --with-pcregrep-bufsiz when      (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1441          PCRE is built.          PCRE is built.
1442    
1443      (c) A --buffer-size=n option has been added to pcregrep, to allow the size      (c) A --buffer-size=n option has been added to pcregrep, to allow the size
# Line 245  Version 8.13 02-Aug-2011 Line 1458  Version 8.13 02-Aug-2011
1458  47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a  47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1459      complete match instead of a partial match. This bug was dependent on both      complete match instead of a partial match. This bug was dependent on both
1460      the PCRE_UTF8 and PCRE_DOTALL options being set.      the PCRE_UTF8 and PCRE_DOTALL options being set.
1461    
1462  48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the  48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1463      starting byte set, because \b was not being ignored.      starting byte set, because \b was not being ignored.
1464    
1465    
1466  Version 8.12 15-Jan-2011  Version 8.12 15-Jan-2011
# Line 900  Version 7.9 11-Apr-09 Line 2113  Version 7.9 11-Apr-09
2113  7.  A pattern that could match an empty string could cause pcregrep to loop; it  7.  A pattern that could match an empty string could cause pcregrep to loop; it
2114      doesn't make sense to accept an empty string match in pcregrep, so I have      doesn't make sense to accept an empty string match in pcregrep, so I have
2115      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2116      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2117        8.33.]
2118    
2119  8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at  8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2120      start or after a newline", because the conditional assertion was not being      start or after a newline", because the conditional assertion was not being
# Line 1143  Version 7.7 07-May-08 Line 2357  Version 7.7 07-May-08
2357      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2358      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2359      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2360      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2361      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2362      was subsequently set up correctly.)      was subsequently set up correctly.)
2363    

Legend:
Removed from v.657  
changed lines
  Added in v.1510

  ViewVC Help
Powered by ViewVC 1.1.5