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

Diff of /code/trunk/ChangeLog

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

revision 740 by zherczeg, Mon Oct 31 06:10:14 2011 UTC revision 1474 by zherczeg, Thu Apr 24 06:43:50 2014 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 8.21  Version 8.36 xx-xxx-2014
5  ------------  ------------------------
6    
7    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
8        -Wmissing-field-initializers and -Wunused-parameter.
9    
10    2.  The tests for quantifiers being too big (greater than 65535) were being
11        applied after reading the number, and stupidly assuming that integer
12        overflow would give a negative number. The tests are now applied as the
13        numbers are read.
14    
15    3.  Tidy code in pcre_exec.c where two branches that used to be different are
16        now the same.
17    
18    4.  The JIT compiler did not generate match limit checks for certain
19        bracketed expressions with quantifiers. This may lead to exponential
20        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
21        issue should be resolved now.
22    
23    
24    Version 8.35 04-April-2014
25    --------------------------
26    
27    1.  A new flag is set, when property checks are present in an XCLASS.
28        When this flag is not set, PCRE can perform certain optimizations
29        such as studying these XCLASS-es.
30    
31    2.  The auto-possessification of character sets were improved: a normal
32        and an extended character set can be compared now. Furthermore
33        the JIT compiler optimizes more character set checks.
34    
35    3.  Got rid of some compiler warnings for potentially uninitialized variables
36        that show up only when compiled with -O2.
37    
38    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
39        of a match later then the end of the match. The pcretest program was not
40        handling the case sensibly - it was outputting from the start to the next
41        binary zero. It now reports this situation in a message, and outputs the
42        text from the end to the start.
43    
44    5.  Fast forward search is improved in JIT. Instead of the first three
45        characters, any three characters with fixed position can be searched.
46        Search order: first, last, middle.
47    
48    6.  Improve character range checks in JIT. Characters are read by an inprecise
49        function now, which returns with an unknown value if the character code is
50        above a certain treshold (e.g: 256). The only limitation is that the value
51        must be bigger than the treshold as well. This function is useful, when
52        the characters above the treshold are handled in the same way.
53    
54    7.  The macros whose names start with RAWUCHAR are placeholders for a future
55        mode in which only the bottom 21 bits of 32-bit data items are used. To
56        make this more memorable for those maintaining the code, the names have
57        been changed to start with UCHAR21, and an extensive comment has been added
58        to their definition.
59    
60    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
61        to the export list in Makefile.am (they were accidentally omitted from the
62        8.34 tarball).
63    
64    9.  The informational output from pcretest used the phrase "starting byte set"
65        which is inappropriate for the 16-bit and 32-bit libraries. As the output
66        for "first char" and "need char" really means "non-UTF-char", I've changed
67        "byte" to "char", and slightly reworded the output. The documentation about
68        these values has also been (I hope) clarified.
69    
70    10. Another JIT related optimization: use table jumps for selecting the correct
71        backtracking path, when more than four alternatives are present inside a
72        bracket.
73    
74    11. Empty match is not possible, when the minimum length is greater than zero,
75        and there is no \K in the pattern. JIT should avoid empty match checks in
76        such cases.
77    
78    12. In a caseless character class with UCP support, when a character with more
79        than one alternative case was not the first character of a range, not all
80        the alternative cases were added to the class. For example, s and \x{17f}
81        are both alternative cases for S: the class [RST] was handled correctly,
82        but [R-T] was not.
83    
84    13. The configure.ac file always checked for pthread support when JIT was
85        enabled. This is not used in Windows, so I have put this test inside a
86        check for the presence of windows.h (which was already tested for).
87    
88    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
89        The algorithm provides a way to skip certain starting offsets, and usually
90        faster than linear prefix searches.
91    
92    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
93        as for 'fr_FR' and 'french'. For some reason, however, it then used the
94        Windows-specific input and output files, which have 'french' screwed in.
95        So this could never have worked. One of the problems with locales is that
96        they aren't always the same. I have now updated RunTest so that it checks
97        the output of the locale test (test 3) against three different output
98        files, and it allows the test to pass if any one of them matches. With luck
99        this should make the test pass on some versions of Solaris where it was
100        failing. Because of the uncertainty, the script did not used to stop if
101        test 3 failed; it now does. If further versions of a French locale ever
102        come to light, they can now easily be added.
103    
104    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
105        there was a message during ./configure, but it did not stop. This now
106        provokes an error. The invalid example in README has been corrected.
107        If a value less than the minimum is given, the minimum value has always
108        been used, but now a warning is given.
109    
110    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
111        was a bug in the test system, which is now fixed. Also, the list of various
112        configurations that are tested for each release did not have one with both
113        16/32 bits and --enable-bar-anycrlf. It now does.
114    
115    18. pcretest was missing "-C bsr" for displaying the \R default setting.
116    
117    19. Little endian PowerPC systems are supported now by the JIT compiler.
118    
119    20. The fast forward newline mechanism could enter to an infinite loop on
120        certain invalid UTF-8 input. Although we don't support these cases
121        this issue can be fixed by a performance optimization.
122    
123    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
124        not take account if existing stack usage. There is now a new global
125        variable called pcre_stack_guard that can be set to point to an external
126        function to check stack availability. It is called at the start of
127        processing every parenthesized group.
128    
129    22. A typo in the code meant that in ungreedy mode the max/min qualifier
130        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
131        not match "ab".
132    
133    23. When UTF was disabled, the JIT program reported some incorrect compile
134        errors. These messages are silenced now.
135    
136    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
137        compiler.
138    
139    25. Change all the temporary files used in RunGrepTest to be different to those
140        used by RunTest so that the tests can be run simultaneously, for example by
141        "make -j check".
142    
143    
144    Version 8.34 15-December-2013
145    -----------------------------
146    
147    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
148        executable memory. Patch inspired by Carsten Klein.
149    
150    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
151        this macro is never tested and has no effect, because the work to support
152        coverage involves only compiling and linking options and special targets in
153        the Makefile. The comment in config.h implied that defining the macro would
154        enable coverage support, which is totally false. There was also support for
155        setting this macro in the CMake files (my fault, I just copied it from
156        configure). SUPPORT_GCOV has now been removed.
157    
158    3.  Make a small performance improvement in strlen16() and strlen32() in
159        pcretest.
160    
161    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
162        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
163        these cases). There was also one in pcretest.c.
164    
165    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
166    
167    6.  In UTF mode, the code for checking whether a group could match an empty
168        string (which is used for indefinitely repeated groups to allow for
169        breaking an infinite loop) was broken when the group contained a repeated
170        negated single-character class with a character that occupied more than one
171        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
172        UTF-8 mode). The effect was undefined: the group might or might not be
173        deemed as matching an empty string, or the program might have crashed.
174    
175    7.  The code for checking whether a group could match an empty string was not
176        recognizing that \h, \H, \v, \V, and \R must match a character.
177    
178    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
179        an empty string. If it can, pcretest shows this in its information output.
180    
181    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
182        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
183        matched by pcre_exec() without using JIT:
184    
185        (a) If the rest of the pattern did not match after a maximal run of
186            grapheme clusters, the code for backing up to try with fewer of them
187            did not always back up over a full grapheme when characters that do not
188            have the modifier quality were involved, e.g. Hangul syllables.
189    
190        (b) If the match point in a subject started with modifier character, and
191            there was no match, the code could incorrectly back up beyond the match
192            point, and potentially beyond the first character in the subject,
193            leading to a segfault or an incorrect match result.
194    
195    10. A conditional group with an assertion condition could lead to PCRE
196        recording an incorrect first data item for a match if no other first data
197        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
198        first data item, and therefore matched "ca" after "c" instead of at the
199        start.
200    
201    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
202        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
203    
204    12. The source of pcregrep now includes z/OS-specific code so that it can be
205        compiled for z/OS as part of the special z/OS distribution.
206    
207    13. Added the -T and -TM options to pcretest.
208    
209    14. The code in pcre_compile.c for creating the table of named capturing groups
210        has been refactored. Instead of creating the table dynamically during the
211        actual compiling pass, the information is remembered during the pre-compile
212        pass (on the stack unless there are more than 20 named groups, in which
213        case malloc() is used) and the whole table is created before the actual
214        compile happens. This has simplified the code (it is now nearly 150 lines
215        shorter) and prepared the way for better handling of references to groups
216        with duplicate names.
217    
218    15. A back reference to a named subpattern when there is more than one of the
219        same name now checks them in the order in which they appear in the pattern.
220        The first one that is set is used for the reference. Previously only the
221        first one was inspected. This change makes PCRE more compatible with Perl.
222    
223    16. Unicode character properties were updated from Unicode 6.3.0.
224    
225    17. The compile-time code for auto-possessification has been refactored, based
226        on a patch by Zoltan Herczeg. It now happens after instead of during
227        compilation. The code is cleaner, and more cases are handled. The option
228        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
229        options in pcretest are provided to set it. It can also be set by
230        (*NO_AUTO_POSSESS) at the start of a pattern.
231    
232    18. The character VT has been added to the default ("C" locale) set of
233        characters that match \s and are generally treated as white space,
234        following this same change in Perl 5.18. There is now no difference between
235        "Perl space" and "POSIX space". Whether VT is treated as white space in
236        other locales depends on the locale.
237    
238    19. The code for checking named groups as conditions, either for being set or
239        for being recursed, has been refactored (this is related to 14 and 15
240        above). Processing unduplicated named groups should now be as fast at
241        numerical groups, and processing duplicated groups should be faster than
242        before.
243    
244    20. Two patches to the CMake build system, by Alexander Barkov:
245    
246          (1) Replace the "source" command by "." in CMakeLists.txt because
247              "source" is a bash-ism.
248    
249          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
250              without these the CMake build does not work on Solaris.
251    
252    21. Perl has changed its handling of \8 and \9. If there is no previously
253        encountered capturing group of those numbers, they are treated as the
254        literal characters 8 and 9 instead of a binary zero followed by the
255        literals. PCRE now does the same.
256    
257    22. Following Perl, added \o{} to specify codepoints in octal, making it
258        possible to specify values greater than 0777 and also making them
259        unambiguous.
260    
261    23. Perl now gives an error for missing closing braces after \x{... instead of
262        treating the string as literal. PCRE now does the same.
263    
264    24. RunTest used to grumble if an inappropriate test was selected explicitly,
265        but just skip it when running all tests. This make it awkward to run ranges
266        of tests when one of them was inappropriate. Now it just skips any
267        inappropriate tests, as it always did when running all tests.
268    
269    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
270        character types such as \d or \w, too many callouts were inserted, and the
271        data that they returned was rubbish.
272    
273    26. In UCP mode, \s was not matching two of the characters that Perl matches,
274        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
275        were matched by \h. The code has now been refactored so that the lists of
276        the horizontal and vertical whitespace characters used for \h and \v (which
277        are defined only in one place) are now also used for \s.
278    
279    27. Add JIT support for the 64 bit TileGX architecture.
280        Patch by Jiong Wang (Tilera Corporation).
281    
282    28. Possessive quantifiers for classes (both explicit and automatically
283        generated) now use special opcodes instead of wrapping in ONCE brackets.
284    
285    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
286        (because it's meaningless), this was not happening when PCRE_CASELESS was
287        set. Not wrong, but inefficient.
288    
289    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
290        properties for \w, \d, etc) is present in a test regex. Otherwise if the
291        test contains no characters greater than 255, Perl doesn't realise it
292        should be using Unicode semantics.
293    
294    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
295        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
296        does in Unicode mode.
297    
298    32. Added the "forbid" facility to pcretest so that putting tests into the
299        wrong test files can sometimes be quickly detected.
300    
301    33. There is now a limit (default 250) on the depth of nesting of parentheses.
302        This limit is imposed to control the amount of system stack used at compile
303        time. It can be changed at build time by --with-parens-nest-limit=xxx or
304        the equivalent in CMake.
305    
306    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
307        errors. Perl warns for these when in warning mode, but PCRE has no facility
308        for giving warnings.
309    
310    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
311        However, this was not working for (?!) because it is optimized to (*FAIL),
312        for which PCRE does not allow quantifiers. The optimization is now disabled
313        when a quantifier follows (?!). I can't see any use for this, but it makes
314        things uniform.
315    
316    36. Perl no longer allows group names to start with digits, so I have made this
317        change also in PCRE. It simplifies the code a bit.
318    
319    37. In extended mode, Perl ignores spaces before a + that indicates a
320        possessive quantifier. PCRE allowed a space before the quantifier, but not
321        before the possessive +. It now does.
322    
323    38. The use of \K (reset reported match start) within a repeated possessive
324        group such as (a\Kb)*+ was not working.
325    
326    40. Document that the same character tables must be used at compile time and
327        run time, and that the facility to pass tables to pcre_exec() and
328        pcre_dfa_exec() is for use only with saved/restored patterns.
329    
330    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
331        features for Builds with MSVC:
332    
333        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
334           builds). The .rc files can be used to set FileDescription and many other
335           attributes.
336    
337        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
338           This allows higher-level build scripts which want .pdb files to avoid
339           hard-coding the exact files needed."
340    
341    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
342        mean "start of word" and "end of word", respectively, as a transition aid.
343    
344    43. A minimizing repeat of a class containing codepoints greater than 255 in
345        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
346        compiled to use the heap for recursion.
347    
348    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
349        is configured.
350    
351    
352    Version 8.33 28-May-2013
353    ------------------------
354    
355    1.  Added 'U' to some constants that are compared to unsigned integers, to
356        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
357        variables that are added to signed variables, to ensure the result is
358        signed and can be negated.
359    
360    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
361        CMake config files.
362    
363    3.  Revise the creation of config.h.generic so that all boolean macros are
364        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
365        overriding via -D on the command line possible.
366    
367    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
368        to unsigned int is reported to make a quite noticeable speed difference in
369        a specific Windows environment. Testing on Linux did also appear to show
370        some benefit (and it is clearly not harmful). Also fixed the definition of
371        Xop which should be unsigned.
372    
373    5.  Related to (4), changing the definition of the intermediate variable cc
374        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
375        improvements.
376    
377    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
378        unnecessary spaces.
379    
380    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
381        and later.
382    
383    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
384    
385    9.  Optimizing fast_forward_start_bits in JIT.
386    
387    10. Adding support for callouts in JIT, and fixing some issues revealed
388        during this work. Namely:
389    
390        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
391    
392        (b) Minimum length was not checked before the matching is started.
393    
394    11. The value of capture_last that is passed to callouts was incorrect in some
395        cases when there was a capture on one path that was subsequently abandoned
396        after a backtrack. Also, the capture_last value is now reset after a
397        recursion, since all captures are also reset in this case.
398    
399    12. The interpreter no longer returns the "too many substrings" error in the
400        case when an overflowing capture is in a branch that is subsequently
401        abandoned after a backtrack.
402    
403    13. In the pathological case when an offset vector of size 2 is used, pcretest
404        now prints out the matched string after a yield of 0 or 1.
405    
406    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
407        Only supported by the JIT compiler at the moment.
408    
409    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
410    
411    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
412        offset of the starting point of the matching process, provided the offsets
413        vector is large enough.
414    
415    17. The \A escape now records a lookbehind value of 1, though its execution
416        does not actually inspect the previous character. This is to ensure that,
417        in partial multi-segment matching, at least one character from the old
418        segment is retained when a new segment is processed. Otherwise, if there
419        are no lookbehinds in the pattern, \A might match incorrectly at the start
420        of a new segment.
421    
422    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
423    
424    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
425        gives some modest performance improvement in 8-bit mode.
426    
427    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
428        be expressed in certain programming languages using Universal Character
429        Names.
430    
431    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
432        which points out that "non characters" are not "characters that may not
433        appear in Unicode strings" but rather "characters that are reserved for
434        internal use and have only local meaning".
435    
436    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
437        there was a conditional group that depended on an assertion, if the
438        assertion was false, the callout that immediately followed the alternation
439        in the condition was skipped when pcre_exec() was used for matching.
440    
441    23. Allow an explicit callout to be inserted before an assertion that is the
442        condition for a conditional group, for compatibility with automatic
443        callouts, which always insert a callout at this point.
444    
445    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
446        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
447    
448    25. (*PRUNE) is now supported by the JIT compiler.
449    
450    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
451    
452    27. Fix the case where there are two or more SKIPs with arguments that may be
453        ignored.
454    
455    28. (*SKIP) is now supported by the JIT compiler.
456    
457    29. (*THEN) is now supported by the JIT compiler.
458    
459    30. Update RunTest with additional test selector options.
460    
461    31. The way PCRE handles backtracking verbs has been changed in two ways.
462    
463        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
464        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
465        backtracking. In some cases this makes it more Perl-compatible, but Perl's
466        rather obscure rules do not always do the same thing.
467    
468        (2) Previously, backtracking verbs were confined within assertions. This is
469        no longer the case for positive assertions, except for (*ACCEPT). Again,
470        this sometimes improves Perl compatibility, and sometimes does not.
471    
472    32. A number of tests that were in test 2 because Perl did things differently
473        have been moved to test 1, because either Perl or PCRE has changed, and
474        these tests are now compatible.
475    
476    32. Backtracking control verbs are now handled in the same way in JIT and
477        interpreter.
478    
479    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
480        contained a forward subroutine reference caused a compile error.
481    
482    34. Auto-detect and optimize limited repetitions in JIT.
483    
484    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
485        blocking (*UTF) etc.
486    
487    36. In the interpreter, maximizing pattern repetitions for characters and
488        character types now use tail recursion, which reduces stack usage.
489    
490    37. The value of the max lookbehind was not correctly preserved if a compiled
491        and saved regex was reloaded on a host of different endianness.
492    
493    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
494        of the compiled pattern block, expand the flags field from 16 to 32 bits
495        because it was almost full.
496    
497    39. Try madvise first before posix_madvise.
498    
499    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
500        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
501        original change locked out all matches of empty strings. This has been
502        changed so that one match of an empty string per line is recognized.
503        Subsequent searches on the same line (for colouring or for --only-matching,
504        for example) do not recognize empty strings.
505    
506    41. Applied a user patch to fix a number of spelling mistakes in comments.
507    
508    42. Data lines longer than 65536 caused pcretest to crash.
509    
510    43. Clarified the data type for length and startoffset arguments for pcre_exec
511        and pcre_dfa_exec in the function-specific man pages, where they were
512        explicitly stated to be in bytes, never having been updated. I also added
513        some clarification to the pcreapi man page.
514    
515    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
516        a segmentation fault.
517    
518    
519    Version 8.32 30-November-2012
520    -----------------------------
521    
522    1.  Improved JIT compiler optimizations for first character search and single
523        character iterators.
524    
525    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
526        Patch by Daniel Richard G.
527    
528    3.  Single character iterator optimizations in the JIT compiler.
529    
530    4.  Improved JIT compiler optimizations for character ranges.
531    
532    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
533        Reported by Giuseppe D'Angelo.
534    
535    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
536        of a line, was being set incorrectly in cases where .* appeared inside
537        atomic brackets at the start of a pattern, or where there was a subsequent
538        *PRUNE or *SKIP.
539    
540    7.  Improved instruction cache flush for POWER/PowerPC.
541        Patch by Daniel Richard G.
542    
543    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
544        grep:
545    
546        (a) There is now no limit to the number of patterns to be matched.
547    
548        (b) An error is given if a pattern is too long.
549    
550        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
551            are now supported.
552    
553        (d) --exclude-from and --include-from (multiple use) have been added.
554    
555        (e) Exclusions and inclusions now apply to all files and directories, not
556            just to those obtained from scanning a directory recursively.
557    
558        (f) Multiple uses of -f and --file-list are now supported.
559    
560        (g) In a Windows environment, the default for -d has been changed from
561            "read" (the GNU grep default) to "skip", because otherwise the presence
562            of a directory in the file list provokes an error.
563    
564        (h) The documentation has been revised and clarified in places.
565    
566    9.  Improve the matching speed of capturing brackets.
567    
568    10. Changed the meaning of \X so that it now matches a Unicode extended
569        grapheme cluster.
570    
571    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
572        out POSIX threads when JIT support is configured.
573    
574    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
575    
576    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
577        structure could go wrong in environments where size_t is not the same size
578        as int.
579    
580    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
581        set.
582    
583    15. The EBCDIC support had decayed; later updates to the code had included
584        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
585        general tidy up of EBCDIC-related issues, and the documentation was also
586        not quite right. There is now a test that can be run on ASCII systems to
587        check some of the EBCDIC-related things (but is it not a full test).
588    
589    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
590        in a small tidy to the code.
591    
592    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
593    
594    18. If the --only-matching (-o) option in pcregrep is specified multiple
595        times, each one causes appropriate output. For example, -o1 -o2 outputs the
596        substrings matched by the 1st and 2nd capturing parentheses. A separating
597        string can be specified by --om-separator (default empty).
598    
599    19. Improving the first n character searches.
600    
601    20. Turn case lists for horizontal and vertical white space into macros so that
602        they are defined only once.
603    
604    21. This set of changes together give more compatible Unicode case-folding
605        behaviour for characters that have more than one other case when UCP
606        support is available.
607    
608        (a) The Unicode property table now has offsets into a new table of sets of
609            three or more characters that are case-equivalent. The MultiStage2.py
610            script that generates these tables (the pcre_ucd.c file) now scans
611            CaseFolding.txt instead of UnicodeData.txt for character case
612            information.
613    
614        (b) The code for adding characters or ranges of characters to a character
615            class has been abstracted into a generalized function that also handles
616            case-independence. In UTF-mode with UCP support, this uses the new data
617            to handle characters with more than one other case.
618    
619        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
620            whose other case is greater than 256 are now correctly matched
621            caselessly. Previously, the high codepoint matched the low one, but not
622            vice versa.
623    
624        (d) The processing of \h, \H, \v, and \ in character classes now makes use
625            of the new class addition function, using character lists defined as
626            macros alongside the case definitions of 20 above.
627    
628        (e) Caseless back references now work with characters that have more than
629            one other case.
630    
631        (f) General caseless matching of characters with more than one other case
632            is supported.
633    
634    22. Unicode character properties were updated from Unicode 6.2.0
635    
636    23. Improved CMake support under Windows. Patch by Daniel Richard G.
637    
638    24. Add support for 32-bit character strings, and UTF-32
639    
640    25. Major JIT compiler update (code refactoring and bugfixing).
641        Experimental Sparc 32 support is added.
642    
643    26. Applied a modified version of Daniel Richard G's patch to create
644        pcre.h.generic and config.h.generic by "make" instead of in the
645        PrepareRelease script.
646    
647    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
648        pcre_compile.c when checking for a zero character.
649    
650    28. Introducing a native interface for JIT. Through this interface, the compiled
651        machine code can be directly executed. The purpose of this interface is to
652        provide fast pattern matching, so several sanity checks are not performed.
653        However, feature tests are still performed. The new interface provides
654        1.4x speedup compared to the old one.
655    
656    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
657        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
658        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
659    
660    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
661        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
662        the "old" RFC 2279). Instead, it ended up passing a negative length to
663        pcre_exec().
664    
665    31. Add support for GCC's visibility feature to hide internal functions.
666    
667    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
668        "unknown -C option" after outputting 0 or 1.
669    
670    33. There is now support for generating a code coverage report for the test
671        suite in environments where gcc is the compiler and lcov is installed. This
672        is mainly for the benefit of the developers.
673    
674    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
675        unaddressable using valgrind annotations, allowing valgrind to detect
676        invalid memory accesses. This is mainly for the benefit of the developers.
677    
678    25. (*UTF) can now be used to start a pattern in any of the three libraries.
679    
680    26. Give configure error if --enable-cpp but no C++ compiler found.
681    
682    
683    Version 8.31 06-July-2012
684    -------------------------
685    
686    1.  Fixing a wrong JIT test case and some compiler warnings.
687    
688    2.  Removed a bashism from the RunTest script.
689    
690    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
691        to unsigned type, result still unsigned" that was given by an MS compiler
692        on encountering the code "-sizeof(xxx)".
693    
694    4.  Partial matching support is added to the JIT compiler.
695    
696    5.  Fixed several bugs concerned with partial matching of items that consist
697        of more than one character:
698    
699        (a) /^(..)\1/ did not partially match "aba" because checking references was
700            done on an "all or nothing" basis. This also applied to repeated
701            references.
702    
703        (b) \R did not give a hard partial match if \r was found at the end of the
704            subject.
705    
706        (c) \X did not give a hard partial match after matching one or more
707            characters at the end of the subject.
708    
709        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
710            a partial match for the string "\r".
711    
712        (e) When newline was set to CRLF, the metacharacter "." did not recognize
713            a partial match for a CR character at the end of the subject string.
714    
715    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
716        running pcretest, the text "(JIT)" added to the output whenever JIT is
717        actually used to run the match.
718    
719    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
720        or /S+[+] with a digit between 1 and 7.
721    
722    8.  OP_NOT now supports any UTF character not just single-byte ones.
723    
724    9.  (*MARK) control verb is now supported by the JIT compiler.
725    
726    10. The command "./RunTest list" lists the available tests without actually
727        running any of them. (Because I keep forgetting what they all are.)
728    
729    11. Add PCRE_INFO_MAXLOOKBEHIND.
730    
731    12. Applied a (slightly modified) user-supplied patch that improves performance
732        when the heap is used for recursion (compiled with --disable-stack-for-
733        recursion). Instead of malloc and free for each heap frame each time a
734        logical recursion happens, frames are retained on a chain and re-used where
735        possible. This sometimes gives as much as 30% improvement.
736    
737    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
738        call.
739    
740    14. As documented, (*COMMIT) is now confined to within a positive assertion.
741    
742    15. It is now possible to link pcretest with libedit as an alternative to
743        libreadline.
744    
745    16. (*COMMIT) control verb is now supported by the JIT compiler.
746    
747    17. The Unicode data tables have been updated to Unicode 6.1.0.
748    
749    18. Added --file-list option to pcregrep.
750    
751    19. Added binary file support to pcregrep, including the -a, --binary-files,
752        -I, and --text options.
753    
754    20. The madvise function is renamed for posix_madvise for QNX compatibility
755        reasons. Fixed by Giuseppe D'Angelo.
756    
757    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
758    
759    22. Changed the option for creating symbolic links for 16-bit man pages from
760        -s to -sf so that re-installing does not cause issues.
761    
762    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
763    
764    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
765        matches in certain environments (the workspace was not being correctly
766        retained). Also added to pcre_dfa_exec() a simple plausibility check on
767        some of the workspace data at the beginning of a restart.
768    
769    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
770        was not doing so when it should - probably a typo introduced by SVN 528
771        (change 8.10/14).
772    
773    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
774        \w+ when the character tables indicated that \x{c4} was a word character.
775        There were several related cases, all because the tests for doing a table
776        lookup were testing for characters less than 127 instead of 255.
777    
778    27. If a pattern contains capturing parentheses that are not used in a match,
779        their slots in the ovector are set to -1. For those that are higher than
780        any matched groups, this happens at the end of processing. In the case when
781        there were back references that the ovector was too small to contain
782        (causing temporary malloc'd memory to be used during matching), and the
783        highest capturing number was not used, memory off the end of the ovector
784        was incorrectly being set to -1. (It was using the size of the temporary
785        memory instead of the true size.)
786    
787    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
788        ovector size, it uses memory at the end of the block that it has got.
789    
790    29. Check for an overlong MARK name and give an error at compile time. The
791        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
792    
793    30. JIT compiler update.
794    
795    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
796        Rill for the patch.
797    
798    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
799    
800    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
801    
802    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
803        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
804        was enabled.
805    
806    35. Improve JIT code generation for greedy plus quantifier.
807    
808    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
809        1 to "aa" instead of to an empty string. The bug affected repeated groups
810        that could potentially match an empty string.
811    
812    37. Optimizing single character iterators in JIT.
813    
814    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
815        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
816        codepoints that are too big for the mode are faulted, and in a UTF mode,
817        disallowed codepoints are also faulted.
818    
819    39. If PCRE was compiled with UTF support, in three places in the DFA
820        matcher there was code that should only have been obeyed in UTF mode, but
821        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
822        processing when bytes with values greater than 127 were present. In 16-bit
823        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
824        both cases the values are those that cannot be the first data item in a UTF
825        character. The three items that might have provoked this were recursions,
826        possessively repeated groups, and atomic groups.
827    
828    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
829        and pcregrep, because some OS require shared objects to be explicitly
830        passed to ld, causing the link step to fail if they are not.
831    
832    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
833        mode, patterns that started with \h* or \R* might be incorrectly matched.
834    
835    
836    Version 8.30 04-February-2012
837    -----------------------------
838    
839    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
840        name is defined in ctype.h.
841    
842    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
843        only in quite long subpatterns.
844    
845    3.  Removed the function pcre_info(), which has been obsolete and deprecated
846        since it was replaced by pcre_fullinfo() in February 2000.
847    
848    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
849        match a (*MARK), and the match failed at the start of the subject, a
850        reference to memory before the start of the subject could occur. This bug
851        was introduced by fix 17 of release 8.21.
852    
853    5.  A reference to an unset group with zero minimum repetition was giving
854        totally wrong answers (in non-JavaScript-compatibility mode). For example,
855        /(another)?(\1?)test/ matched against "hello world test". This bug was
856        introduced in release 8.13.
857    
858    6.  Add support for 16-bit character strings (a large amount of work involving
859        many changes and refactorings).
860    
861    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
862        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
863        from a file.
864    
865    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
866        rounding is not applied in this particular case).
867    
868    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
869        if they appear, or are escaped, in patterns.
870    
871    10. Get rid of a number of -Wunused-but-set-variable warnings.
872    
873    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
874        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
875        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
876        also returns the mark "x". This bug applied to capturing parentheses,
877        non-capturing parentheses, and atomic parentheses. It also applied to some
878        assertions.
879    
880    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
881        information out of configure.ac instead of relying on pcre.h.generic, which
882        is not stored in the repository.
883    
884    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
885        -lreadline.
886    
887    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
888    
889    15. Applied Graycode's patch to put the top-level frame on the stack rather
890        than the heap when not using the stack for recursion. This gives a
891        performance improvement in many cases when recursion is not deep.
892    
893    16. Experimental code added to "pcretest -C" to output the stack frame size.
894    
895    
896    Version 8.21 12-Dec-2011
897    ------------------------
898    
899  1.  Updating the JIT compiler.  1.  Updating the JIT compiler.
900    
901    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
902        are added as well.
903    
904    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
905        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
906        calling _pcre_jit_exec. Some extra comments are added.
907    
908    4.  (*MARK) settings inside atomic groups that do not contain any capturing
909        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
910        was introduced by change 18 for 8.20.
911    
912    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
913        ECMA-262 standard.
914    
915    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
916        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
917        This bug was probably introduced by change 9 of 8.13.
918    
919    7.  While fixing 6 above, I noticed that a number of other items were being
920        incorrectly rejected as "not fixed length". This arose partly because newer
921        opcodes had not been added to the fixed-length checking code. I have (a)
922        corrected the bug and added tests for these items, and (b) arranged for an
923        error to occur if an unknown opcode is encountered while checking for fixed
924        length instead of just assuming "not fixed length". The items that were
925        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
926        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
927        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
928    
929    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
930        being incorrectly compiled and would have given unpredicatble results.
931    
932    9.  A possessively repeated subpattern with minimum repeat count greater than
933        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
934        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
935        the first (A) could occur when it should not.
936    
937    10. Add a cast and remove a redundant test from the code.
938    
939    11. JIT should use pcre_malloc/pcre_free for allocation.
940    
941    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
942        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
943        is anything other than /usr, -L is still shown).
944    
945    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
946    
947    14. Perl does not support \N without a following name in a [] class; PCRE now
948        also gives an error.
949    
950    15. If a forward reference was repeated with an upper limit of around 2000,
951        it caused the error "internal error: overran compiling workspace". The
952        maximum number of forward references (including repeats) was limited by the
953        internal workspace, and dependent on the LINK_SIZE. The code has been
954        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
955        the default depends on LINK_SIZE. There is a new upper limit (for safety)
956        of around 200,000 forward references. While doing this, I also speeded up
957        the filling in of repeated forward references.
958    
959    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
960        incorrectly expecting the subject to contain another "a" after the start.
961    
962    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
963        in the match, the SKIP should be ignored. This was not happening; instead
964        the SKIP was being treated as NOMATCH. For patterns such as
965        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
966        tested.
967    
968    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
969        now much more compatible with Perl, in particular in cases where the result
970        is a non-match for a non-anchored pattern. For example, if
971        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
972        "m", where previously it did not return a name. A side effect of this
973        change is that for partial matches, the last encountered mark name is
974        returned, as for non matches. A number of tests that were previously not
975        Perl-compatible have been moved into the Perl-compatible test files. The
976        refactoring has had the pleasing side effect of removing one argument from
977        the match() function, thus reducing its stack requirements.
978    
979    19. If the /S+ option was used in pcretest to study a pattern using JIT,
980        subsequent uses of /S (without +) incorrectly behaved like /S+.
981    
982    21. Retrieve executable code size support for the JIT compiler and fixing
983        some warnings.
984    
985    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
986        not work when the shorter character appeared right at the end of the
987        subject string.
988    
989    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
990        systems.
991    
992    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
993        output it when the /M option is used in pcretest.
994    
995    25. The CheckMan script was not being included in the distribution. Also, added
996        an explicit "perl" to run Perl scripts from the PrepareRelease script
997        because this is reportedly needed in Windows.
998    
999    26. If study data was being save in a file and studying had not found a set of
1000        "starts with" bytes for the pattern, the data written to the file (though
1001        never used) was taken from uninitialized memory and so caused valgrind to
1002        complain.
1003    
1004    27. Updated RunTest.bat as provided by Sheri Pierce.
1005    
1006    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1007    
1008    29. Computation of memory usage for the table of capturing group names was
1009        giving an unnecessarily large value.
1010    
1011    
1012  Version 8.20 21-Oct-2011  Version 8.20 21-Oct-2011
1013  ------------------------  ------------------------
# Line 1021  Version 7.9 11-Apr-09 Line 2023  Version 7.9 11-Apr-09
2023  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
2024      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
2025      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2026      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2027        8.33.]
2028    
2029  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
2030      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 1264  Version 7.7 07-May-08 Line 2267  Version 7.7 07-May-08
2267      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2268      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2269      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2270      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2271      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2272      was subsequently set up correctly.)      was subsequently set up correctly.)
2273    

Legend:
Removed from v.740  
changed lines
  Added in v.1474

  ViewVC Help
Powered by ViewVC 1.1.5