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

Diff of /code/trunk/ChangeLog

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

revision 425 by ph10, Mon Aug 17 14:48:48 2009 UTC revision 1449 by ph10, Fri Jan 17 16:59:52 2014 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 8.00 ??-???-??  Version 8.35-RC1 xx-xxxx-201x
5    -----------------------------
6    
7    1.  A new flag is set, when property checks are present in an XCLASS.
8        When this flag is not set, PCRE can perform certain optimizations
9        such as studying these XCLASS-es.
10    
11    2.  The auto-possessification of character sets were improved: a normal
12        and an extended character set can be compared now. Furthermore
13        the JIT compiler optimizes more character set checks.
14    
15    3.  Got rid of some compiler warnings for potentially uninitialized variables
16        that show up only when compiled with -O2.
17    
18    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
19        of a match later then the end of the match. The pcretest program was not
20        handling the case sensibly - it was outputting from the start to the next
21        binary zero. It now reports this situation in a message, and outputs the
22        text from the end to the start.
23    
24    5.  Fast forward search is improved in JIT. Instead of the first three
25        characters, any three characters with fixed position can be searched.
26        Search order: first, last, middle.
27    
28    6.  Improve character range checks in JIT. Characters are read by an inprecise
29        function now, which returns with an unknown value if the character code is
30        above a certain treshold (e.g: 256). The only limitation is that the value
31        must be bigger than the treshold as well. This function is useful, when
32        the characters above the treshold are handled in the same way.
33    
34    7.  The macros whose names start with RAWUCHAR are placeholders for a future
35        mode in which only the bottom 21 bits of 32-bit data items are used. To
36        make this more memorable for those maintaining the code, the names have
37        been changed to start with UCHAR21, and an extensive comment has been added
38        to their definition.
39    
40    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
41        to the export list in Makefile.am (they were accidentally omitted from the
42        8.34 tarball).
43    
44    9.  The informational output from pcretest used the phrase "starting byte set"
45        which is inappropriate for the 16-bit and 32-bit libraries. As the output
46        for "first char" and "need char" really means "non-UTF-char", I've changed
47        "byte" to "char", and slightly reworded the output. The documentation about
48        these values has also been (I hope) clarified.
49    
50    10. Another JIT related optimization: use table jumps for selecting the correct
51        backtracking path, when more than four alternatives are present inside a
52        bracket.
53    
54    11. Empty match is not possible, when the minimum length is greater than zero,
55        and there is no \K in the pattern. JIT should avoid empty match checks in
56        such cases.
57    
58    12. In a caseless character class with UCP support, when a character with more
59        than one alternative case was not the first character of a range, not all
60        the alternative cases were added to the class. For example, s and \x{17f}
61        are both alternative cases for S: the class [RST] was handled correctly,
62        but [R-T] was not.
63    
64    13. The configure.ac file always checked for pthread support when JIT was
65        enabled. This is not used in Windows, so I have put this test inside a
66        check for the presence of windows.h (which was already tested for).
67    
68    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
69        The algorithm provides a way to skip certain starting offsets, and usually
70        faster than linear prefix searches.
71    
72    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
73        as for 'fr_FR' and 'french'. For some reason, however, it then used the
74        Windows-specific input and output files, which have 'french' screwed in.
75        So this could never have worked. One of the problems with locales is that
76        they aren't always the same. I have now updated RunTest so that it checks
77        the output of the locale test (test 3) against three different output
78        files, and it allows the test to pass if any one of them matches. With luck
79        this should make the test pass on some versions of Solaris where it was
80        failing. Because of the uncertainty, the script did not used to stop if
81        test 3 failed; it now does. If further versions of a French locale ever
82        come to light, they can now easily be added.
83    
84    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
85        there was a message during ./configure, but it did not stop. This now
86        provokes an error. The invalid example in README has been corrected.
87        If a value less than the minimum is given, the minimum value has always
88        been used, but now a warning is given.
89    
90    
91    Version 8.34 15-December-2013
92    -----------------------------
93    
94    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
95        executable memory. Patch inspired by Carsten Klein.
96    
97    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
98        this macro is never tested and has no effect, because the work to support
99        coverage involves only compiling and linking options and special targets in
100        the Makefile. The comment in config.h implied that defining the macro would
101        enable coverage support, which is totally false. There was also support for
102        setting this macro in the CMake files (my fault, I just copied it from
103        configure). SUPPORT_GCOV has now been removed.
104    
105    3.  Make a small performance improvement in strlen16() and strlen32() in
106        pcretest.
107    
108    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
109        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
110        these cases). There was also one in pcretest.c.
111    
112    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
113    
114    6.  In UTF mode, the code for checking whether a group could match an empty
115        string (which is used for indefinitely repeated groups to allow for
116        breaking an infinite loop) was broken when the group contained a repeated
117        negated single-character class with a character that occupied more than one
118        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
119        UTF-8 mode). The effect was undefined: the group might or might not be
120        deemed as matching an empty string, or the program might have crashed.
121    
122    7.  The code for checking whether a group could match an empty string was not
123        recognizing that \h, \H, \v, \V, and \R must match a character.
124    
125    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
126        an empty string. If it can, pcretest shows this in its information output.
127    
128    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
129        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
130        matched by pcre_exec() without using JIT:
131    
132        (a) If the rest of the pattern did not match after a maximal run of
133            grapheme clusters, the code for backing up to try with fewer of them
134            did not always back up over a full grapheme when characters that do not
135            have the modifier quality were involved, e.g. Hangul syllables.
136    
137        (b) If the match point in a subject started with modifier character, and
138            there was no match, the code could incorrectly back up beyond the match
139            point, and potentially beyond the first character in the subject,
140            leading to a segfault or an incorrect match result.
141    
142    10. A conditional group with an assertion condition could lead to PCRE
143        recording an incorrect first data item for a match if no other first data
144        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
145        first data item, and therefore matched "ca" after "c" instead of at the
146        start.
147    
148    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
149        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
150    
151    12. The source of pcregrep now includes z/OS-specific code so that it can be
152        compiled for z/OS as part of the special z/OS distribution.
153    
154    13. Added the -T and -TM options to pcretest.
155    
156    14. The code in pcre_compile.c for creating the table of named capturing groups
157        has been refactored. Instead of creating the table dynamically during the
158        actual compiling pass, the information is remembered during the pre-compile
159        pass (on the stack unless there are more than 20 named groups, in which
160        case malloc() is used) and the whole table is created before the actual
161        compile happens. This has simplified the code (it is now nearly 150 lines
162        shorter) and prepared the way for better handling of references to groups
163        with duplicate names.
164    
165    15. A back reference to a named subpattern when there is more than one of the
166        same name now checks them in the order in which they appear in the pattern.
167        The first one that is set is used for the reference. Previously only the
168        first one was inspected. This change makes PCRE more compatible with Perl.
169    
170    16. Unicode character properties were updated from Unicode 6.3.0.
171    
172    17. The compile-time code for auto-possessification has been refactored, based
173        on a patch by Zoltan Herczeg. It now happens after instead of during
174        compilation. The code is cleaner, and more cases are handled. The option
175        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
176        options in pcretest are provided to set it. It can also be set by
177        (*NO_AUTO_POSSESS) at the start of a pattern.
178    
179    18. The character VT has been added to the default ("C" locale) set of
180        characters that match \s and are generally treated as white space,
181        following this same change in Perl 5.18. There is now no difference between
182        "Perl space" and "POSIX space". Whether VT is treated as white space in
183        other locales depends on the locale.
184    
185    19. The code for checking named groups as conditions, either for being set or
186        for being recursed, has been refactored (this is related to 14 and 15
187        above). Processing unduplicated named groups should now be as fast at
188        numerical groups, and processing duplicated groups should be faster than
189        before.
190    
191    20. Two patches to the CMake build system, by Alexander Barkov:
192    
193          (1) Replace the "source" command by "." in CMakeLists.txt because
194              "source" is a bash-ism.
195    
196          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
197              without these the CMake build does not work on Solaris.
198    
199    21. Perl has changed its handling of \8 and \9. If there is no previously
200        encountered capturing group of those numbers, they are treated as the
201        literal characters 8 and 9 instead of a binary zero followed by the
202        literals. PCRE now does the same.
203    
204    22. Following Perl, added \o{} to specify codepoints in octal, making it
205        possible to specify values greater than 0777 and also making them
206        unambiguous.
207    
208    23. Perl now gives an error for missing closing braces after \x{... instead of
209        treating the string as literal. PCRE now does the same.
210    
211    24. RunTest used to grumble if an inappropriate test was selected explicitly,
212        but just skip it when running all tests. This make it awkward to run ranges
213        of tests when one of them was inappropriate. Now it just skips any
214        inappropriate tests, as it always did when running all tests.
215    
216    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
217        character types such as \d or \w, too many callouts were inserted, and the
218        data that they returned was rubbish.
219    
220    26. In UCP mode, \s was not matching two of the characters that Perl matches,
221        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
222        were matched by \h. The code has now been refactored so that the lists of
223        the horizontal and vertical whitespace characters used for \h and \v (which
224        are defined only in one place) are now also used for \s.
225    
226    27. Add JIT support for the 64 bit TileGX architecture.
227        Patch by Jiong Wang (Tilera Corporation).
228    
229    28. Possessive quantifiers for classes (both explicit and automatically
230        generated) now use special opcodes instead of wrapping in ONCE brackets.
231    
232    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
233        (because it's meaningless), this was not happening when PCRE_CASELESS was
234        set. Not wrong, but inefficient.
235    
236    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
237        properties for \w, \d, etc) is present in a test regex. Otherwise if the
238        test contains no characters greater than 255, Perl doesn't realise it
239        should be using Unicode semantics.
240    
241    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
242        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
243        does in Unicode mode.
244    
245    32. Added the "forbid" facility to pcretest so that putting tests into the
246        wrong test files can sometimes be quickly detected.
247    
248    33. There is now a limit (default 250) on the depth of nesting of parentheses.
249        This limit is imposed to control the amount of system stack used at compile
250        time. It can be changed at build time by --with-parens-nest-limit=xxx or
251        the equivalent in CMake.
252    
253    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
254        errors. Perl warns for these when in warning mode, but PCRE has no facility
255        for giving warnings.
256    
257    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
258        However, this was not working for (?!) because it is optimized to (*FAIL),
259        for which PCRE does not allow quantifiers. The optimization is now disabled
260        when a quantifier follows (?!). I can't see any use for this, but it makes
261        things uniform.
262    
263    36. Perl no longer allows group names to start with digits, so I have made this
264        change also in PCRE. It simplifies the code a bit.
265    
266    37. In extended mode, Perl ignores spaces before a + that indicates a
267        possessive quantifier. PCRE allowed a space before the quantifier, but not
268        before the possessive +. It now does.
269    
270    38. The use of \K (reset reported match start) within a repeated possessive
271        group such as (a\Kb)*+ was not working.
272    
273    40. Document that the same character tables must be used at compile time and
274        run time, and that the facility to pass tables to pcre_exec() and
275        pcre_dfa_exec() is for use only with saved/restored patterns.
276    
277    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
278        features for Builds with MSVC:
279    
280        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
281           builds). The .rc files can be used to set FileDescription and many other
282           attributes.
283    
284        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
285           This allows higher-level build scripts which want .pdb files to avoid
286           hard-coding the exact files needed."
287    
288    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
289        mean "start of word" and "end of word", respectively, as a transition aid.
290    
291    43. A minimizing repeat of a class containing codepoints greater than 255 in
292        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
293        compiled to use the heap for recursion.
294    
295    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
296        is configured.
297    
298    
299    Version 8.33 28-May-2013
300    ------------------------
301    
302    1.  Added 'U' to some constants that are compared to unsigned integers, to
303        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
304        variables that are added to signed variables, to ensure the result is
305        signed and can be negated.
306    
307    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
308        CMake config files.
309    
310    3.  Revise the creation of config.h.generic so that all boolean macros are
311        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
312        overriding via -D on the command line possible.
313    
314    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
315        to unsigned int is reported to make a quite noticeable speed difference in
316        a specific Windows environment. Testing on Linux did also appear to show
317        some benefit (and it is clearly not harmful). Also fixed the definition of
318        Xop which should be unsigned.
319    
320    5.  Related to (4), changing the definition of the intermediate variable cc
321        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
322        improvements.
323    
324    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
325        unnecessary spaces.
326    
327    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
328        and later.
329    
330    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
331    
332    9.  Optimizing fast_forward_start_bits in JIT.
333    
334    10. Adding support for callouts in JIT, and fixing some issues revealed
335        during this work. Namely:
336    
337        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
338    
339        (b) Minimum length was not checked before the matching is started.
340    
341    11. The value of capture_last that is passed to callouts was incorrect in some
342        cases when there was a capture on one path that was subsequently abandoned
343        after a backtrack. Also, the capture_last value is now reset after a
344        recursion, since all captures are also reset in this case.
345    
346    12. The interpreter no longer returns the "too many substrings" error in the
347        case when an overflowing capture is in a branch that is subsequently
348        abandoned after a backtrack.
349    
350    13. In the pathological case when an offset vector of size 2 is used, pcretest
351        now prints out the matched string after a yield of 0 or 1.
352    
353    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
354        Only supported by the JIT compiler at the moment.
355    
356    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
357    
358    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
359        offset of the starting point of the matching process, provided the offsets
360        vector is large enough.
361    
362    17. The \A escape now records a lookbehind value of 1, though its execution
363        does not actually inspect the previous character. This is to ensure that,
364        in partial multi-segment matching, at least one character from the old
365        segment is retained when a new segment is processed. Otherwise, if there
366        are no lookbehinds in the pattern, \A might match incorrectly at the start
367        of a new segment.
368    
369    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
370    
371    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
372        gives some modest performance improvement in 8-bit mode.
373    
374    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
375        be expressed in certain programming languages using Universal Character
376        Names.
377    
378    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
379        which points out that "non characters" are not "characters that may not
380        appear in Unicode strings" but rather "characters that are reserved for
381        internal use and have only local meaning".
382    
383    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
384        there was a conditional group that depended on an assertion, if the
385        assertion was false, the callout that immediately followed the alternation
386        in the condition was skipped when pcre_exec() was used for matching.
387    
388    23. Allow an explicit callout to be inserted before an assertion that is the
389        condition for a conditional group, for compatibility with automatic
390        callouts, which always insert a callout at this point.
391    
392    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
393        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
394    
395    25. (*PRUNE) is now supported by the JIT compiler.
396    
397    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
398    
399    27. Fix the case where there are two or more SKIPs with arguments that may be
400        ignored.
401    
402    28. (*SKIP) is now supported by the JIT compiler.
403    
404    29. (*THEN) is now supported by the JIT compiler.
405    
406    30. Update RunTest with additional test selector options.
407    
408    31. The way PCRE handles backtracking verbs has been changed in two ways.
409    
410        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
411        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
412        backtracking. In some cases this makes it more Perl-compatible, but Perl's
413        rather obscure rules do not always do the same thing.
414    
415        (2) Previously, backtracking verbs were confined within assertions. This is
416        no longer the case for positive assertions, except for (*ACCEPT). Again,
417        this sometimes improves Perl compatibility, and sometimes does not.
418    
419    32. A number of tests that were in test 2 because Perl did things differently
420        have been moved to test 1, because either Perl or PCRE has changed, and
421        these tests are now compatible.
422    
423    32. Backtracking control verbs are now handled in the same way in JIT and
424        interpreter.
425    
426    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
427        contained a forward subroutine reference caused a compile error.
428    
429    34. Auto-detect and optimize limited repetitions in JIT.
430    
431    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
432        blocking (*UTF) etc.
433    
434    36. In the interpreter, maximizing pattern repetitions for characters and
435        character types now use tail recursion, which reduces stack usage.
436    
437    37. The value of the max lookbehind was not correctly preserved if a compiled
438        and saved regex was reloaded on a host of different endianness.
439    
440    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
441        of the compiled pattern block, expand the flags field from 16 to 32 bits
442        because it was almost full.
443    
444    39. Try madvise first before posix_madvise.
445    
446    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
447        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
448        original change locked out all matches of empty strings. This has been
449        changed so that one match of an empty string per line is recognized.
450        Subsequent searches on the same line (for colouring or for --only-matching,
451        for example) do not recognize empty strings.
452    
453    41. Applied a user patch to fix a number of spelling mistakes in comments.
454    
455    42. Data lines longer than 65536 caused pcretest to crash.
456    
457    43. Clarified the data type for length and startoffset arguments for pcre_exec
458        and pcre_dfa_exec in the function-specific man pages, where they were
459        explicitly stated to be in bytes, never having been updated. I also added
460        some clarification to the pcreapi man page.
461    
462    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
463        a segmentation fault.
464    
465    
466    Version 8.32 30-November-2012
467    -----------------------------
468    
469    1.  Improved JIT compiler optimizations for first character search and single
470        character iterators.
471    
472    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
473        Patch by Daniel Richard G.
474    
475    3.  Single character iterator optimizations in the JIT compiler.
476    
477    4.  Improved JIT compiler optimizations for character ranges.
478    
479    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
480        Reported by Giuseppe D'Angelo.
481    
482    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
483        of a line, was being set incorrectly in cases where .* appeared inside
484        atomic brackets at the start of a pattern, or where there was a subsequent
485        *PRUNE or *SKIP.
486    
487    7.  Improved instruction cache flush for POWER/PowerPC.
488        Patch by Daniel Richard G.
489    
490    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
491        grep:
492    
493        (a) There is now no limit to the number of patterns to be matched.
494    
495        (b) An error is given if a pattern is too long.
496    
497        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
498            are now supported.
499    
500        (d) --exclude-from and --include-from (multiple use) have been added.
501    
502        (e) Exclusions and inclusions now apply to all files and directories, not
503            just to those obtained from scanning a directory recursively.
504    
505        (f) Multiple uses of -f and --file-list are now supported.
506    
507        (g) In a Windows environment, the default for -d has been changed from
508            "read" (the GNU grep default) to "skip", because otherwise the presence
509            of a directory in the file list provokes an error.
510    
511        (h) The documentation has been revised and clarified in places.
512    
513    9.  Improve the matching speed of capturing brackets.
514    
515    10. Changed the meaning of \X so that it now matches a Unicode extended
516        grapheme cluster.
517    
518    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
519        out POSIX threads when JIT support is configured.
520    
521    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
522    
523    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
524        structure could go wrong in environments where size_t is not the same size
525        as int.
526    
527    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
528        set.
529    
530    15. The EBCDIC support had decayed; later updates to the code had included
531        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
532        general tidy up of EBCDIC-related issues, and the documentation was also
533        not quite right. There is now a test that can be run on ASCII systems to
534        check some of the EBCDIC-related things (but is it not a full test).
535    
536    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
537        in a small tidy to the code.
538    
539    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
540    
541    18. If the --only-matching (-o) option in pcregrep is specified multiple
542        times, each one causes appropriate output. For example, -o1 -o2 outputs the
543        substrings matched by the 1st and 2nd capturing parentheses. A separating
544        string can be specified by --om-separator (default empty).
545    
546    19. Improving the first n character searches.
547    
548    20. Turn case lists for horizontal and vertical white space into macros so that
549        they are defined only once.
550    
551    21. This set of changes together give more compatible Unicode case-folding
552        behaviour for characters that have more than one other case when UCP
553        support is available.
554    
555        (a) The Unicode property table now has offsets into a new table of sets of
556            three or more characters that are case-equivalent. The MultiStage2.py
557            script that generates these tables (the pcre_ucd.c file) now scans
558            CaseFolding.txt instead of UnicodeData.txt for character case
559            information.
560    
561        (b) The code for adding characters or ranges of characters to a character
562            class has been abstracted into a generalized function that also handles
563            case-independence. In UTF-mode with UCP support, this uses the new data
564            to handle characters with more than one other case.
565    
566        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
567            whose other case is greater than 256 are now correctly matched
568            caselessly. Previously, the high codepoint matched the low one, but not
569            vice versa.
570    
571        (d) The processing of \h, \H, \v, and \ in character classes now makes use
572            of the new class addition function, using character lists defined as
573            macros alongside the case definitions of 20 above.
574    
575        (e) Caseless back references now work with characters that have more than
576            one other case.
577    
578        (f) General caseless matching of characters with more than one other case
579            is supported.
580    
581    22. Unicode character properties were updated from Unicode 6.2.0
582    
583    23. Improved CMake support under Windows. Patch by Daniel Richard G.
584    
585    24. Add support for 32-bit character strings, and UTF-32
586    
587    25. Major JIT compiler update (code refactoring and bugfixing).
588        Experimental Sparc 32 support is added.
589    
590    26. Applied a modified version of Daniel Richard G's patch to create
591        pcre.h.generic and config.h.generic by "make" instead of in the
592        PrepareRelease script.
593    
594    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
595        pcre_compile.c when checking for a zero character.
596    
597    28. Introducing a native interface for JIT. Through this interface, the compiled
598        machine code can be directly executed. The purpose of this interface is to
599        provide fast pattern matching, so several sanity checks are not performed.
600        However, feature tests are still performed. The new interface provides
601        1.4x speedup compared to the old one.
602    
603    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
604        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
605        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
606    
607    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
608        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
609        the "old" RFC 2279). Instead, it ended up passing a negative length to
610        pcre_exec().
611    
612    31. Add support for GCC's visibility feature to hide internal functions.
613    
614    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
615        "unknown -C option" after outputting 0 or 1.
616    
617    33. There is now support for generating a code coverage report for the test
618        suite in environments where gcc is the compiler and lcov is installed. This
619        is mainly for the benefit of the developers.
620    
621    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
622        unaddressable using valgrind annotations, allowing valgrind to detect
623        invalid memory accesses. This is mainly for the benefit of the developers.
624    
625    25. (*UTF) can now be used to start a pattern in any of the three libraries.
626    
627    26. Give configure error if --enable-cpp but no C++ compiler found.
628    
629    
630    Version 8.31 06-July-2012
631    -------------------------
632    
633    1.  Fixing a wrong JIT test case and some compiler warnings.
634    
635    2.  Removed a bashism from the RunTest script.
636    
637    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
638        to unsigned type, result still unsigned" that was given by an MS compiler
639        on encountering the code "-sizeof(xxx)".
640    
641    4.  Partial matching support is added to the JIT compiler.
642    
643    5.  Fixed several bugs concerned with partial matching of items that consist
644        of more than one character:
645    
646        (a) /^(..)\1/ did not partially match "aba" because checking references was
647            done on an "all or nothing" basis. This also applied to repeated
648            references.
649    
650        (b) \R did not give a hard partial match if \r was found at the end of the
651            subject.
652    
653        (c) \X did not give a hard partial match after matching one or more
654            characters at the end of the subject.
655    
656        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
657            a partial match for the string "\r".
658    
659        (e) When newline was set to CRLF, the metacharacter "." did not recognize
660            a partial match for a CR character at the end of the subject string.
661    
662    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
663        running pcretest, the text "(JIT)" added to the output whenever JIT is
664        actually used to run the match.
665    
666    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
667        or /S+[+] with a digit between 1 and 7.
668    
669    8.  OP_NOT now supports any UTF character not just single-byte ones.
670    
671    9.  (*MARK) control verb is now supported by the JIT compiler.
672    
673    10. The command "./RunTest list" lists the available tests without actually
674        running any of them. (Because I keep forgetting what they all are.)
675    
676    11. Add PCRE_INFO_MAXLOOKBEHIND.
677    
678    12. Applied a (slightly modified) user-supplied patch that improves performance
679        when the heap is used for recursion (compiled with --disable-stack-for-
680        recursion). Instead of malloc and free for each heap frame each time a
681        logical recursion happens, frames are retained on a chain and re-used where
682        possible. This sometimes gives as much as 30% improvement.
683    
684    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
685        call.
686    
687    14. As documented, (*COMMIT) is now confined to within a positive assertion.
688    
689    15. It is now possible to link pcretest with libedit as an alternative to
690        libreadline.
691    
692    16. (*COMMIT) control verb is now supported by the JIT compiler.
693    
694    17. The Unicode data tables have been updated to Unicode 6.1.0.
695    
696    18. Added --file-list option to pcregrep.
697    
698    19. Added binary file support to pcregrep, including the -a, --binary-files,
699        -I, and --text options.
700    
701    20. The madvise function is renamed for posix_madvise for QNX compatibility
702        reasons. Fixed by Giuseppe D'Angelo.
703    
704    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
705    
706    22. Changed the option for creating symbolic links for 16-bit man pages from
707        -s to -sf so that re-installing does not cause issues.
708    
709    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
710    
711    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
712        matches in certain environments (the workspace was not being correctly
713        retained). Also added to pcre_dfa_exec() a simple plausibility check on
714        some of the workspace data at the beginning of a restart.
715    
716    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
717        was not doing so when it should - probably a typo introduced by SVN 528
718        (change 8.10/14).
719    
720    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
721        \w+ when the character tables indicated that \x{c4} was a word character.
722        There were several related cases, all because the tests for doing a table
723        lookup were testing for characters less than 127 instead of 255.
724    
725    27. If a pattern contains capturing parentheses that are not used in a match,
726        their slots in the ovector are set to -1. For those that are higher than
727        any matched groups, this happens at the end of processing. In the case when
728        there were back references that the ovector was too small to contain
729        (causing temporary malloc'd memory to be used during matching), and the
730        highest capturing number was not used, memory off the end of the ovector
731        was incorrectly being set to -1. (It was using the size of the temporary
732        memory instead of the true size.)
733    
734    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
735        ovector size, it uses memory at the end of the block that it has got.
736    
737    29. Check for an overlong MARK name and give an error at compile time. The
738        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
739    
740    30. JIT compiler update.
741    
742    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
743        Rill for the patch.
744    
745    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
746    
747    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
748    
749    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
750        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
751        was enabled.
752    
753    35. Improve JIT code generation for greedy plus quantifier.
754    
755    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
756        1 to "aa" instead of to an empty string. The bug affected repeated groups
757        that could potentially match an empty string.
758    
759    37. Optimizing single character iterators in JIT.
760    
761    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
762        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
763        codepoints that are too big for the mode are faulted, and in a UTF mode,
764        disallowed codepoints are also faulted.
765    
766    39. If PCRE was compiled with UTF support, in three places in the DFA
767        matcher there was code that should only have been obeyed in UTF mode, but
768        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
769        processing when bytes with values greater than 127 were present. In 16-bit
770        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
771        both cases the values are those that cannot be the first data item in a UTF
772        character. The three items that might have provoked this were recursions,
773        possessively repeated groups, and atomic groups.
774    
775    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
776        and pcregrep, because some OS require shared objects to be explicitly
777        passed to ld, causing the link step to fail if they are not.
778    
779    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
780        mode, patterns that started with \h* or \R* might be incorrectly matched.
781    
782    
783    Version 8.30 04-February-2012
784    -----------------------------
785    
786    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
787        name is defined in ctype.h.
788    
789    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
790        only in quite long subpatterns.
791    
792    3.  Removed the function pcre_info(), which has been obsolete and deprecated
793        since it was replaced by pcre_fullinfo() in February 2000.
794    
795    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
796        match a (*MARK), and the match failed at the start of the subject, a
797        reference to memory before the start of the subject could occur. This bug
798        was introduced by fix 17 of release 8.21.
799    
800    5.  A reference to an unset group with zero minimum repetition was giving
801        totally wrong answers (in non-JavaScript-compatibility mode). For example,
802        /(another)?(\1?)test/ matched against "hello world test". This bug was
803        introduced in release 8.13.
804    
805    6.  Add support for 16-bit character strings (a large amount of work involving
806        many changes and refactorings).
807    
808    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
809        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
810        from a file.
811    
812    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
813        rounding is not applied in this particular case).
814    
815    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
816        if they appear, or are escaped, in patterns.
817    
818    10. Get rid of a number of -Wunused-but-set-variable warnings.
819    
820    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
821        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
822        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
823        also returns the mark "x". This bug applied to capturing parentheses,
824        non-capturing parentheses, and atomic parentheses. It also applied to some
825        assertions.
826    
827    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
828        information out of configure.ac instead of relying on pcre.h.generic, which
829        is not stored in the repository.
830    
831    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
832        -lreadline.
833    
834    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
835    
836    15. Applied Graycode's patch to put the top-level frame on the stack rather
837        than the heap when not using the stack for recursion. This gives a
838        performance improvement in many cases when recursion is not deep.
839    
840    16. Experimental code added to "pcretest -C" to output the stack frame size.
841    
842    
843    Version 8.21 12-Dec-2011
844    ------------------------
845    
846    1.  Updating the JIT compiler.
847    
848    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
849        are added as well.
850    
851    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
852        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
853        calling _pcre_jit_exec. Some extra comments are added.
854    
855    4.  (*MARK) settings inside atomic groups that do not contain any capturing
856        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
857        was introduced by change 18 for 8.20.
858    
859    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
860        ECMA-262 standard.
861    
862    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
863        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
864        This bug was probably introduced by change 9 of 8.13.
865    
866    7.  While fixing 6 above, I noticed that a number of other items were being
867        incorrectly rejected as "not fixed length". This arose partly because newer
868        opcodes had not been added to the fixed-length checking code. I have (a)
869        corrected the bug and added tests for these items, and (b) arranged for an
870        error to occur if an unknown opcode is encountered while checking for fixed
871        length instead of just assuming "not fixed length". The items that were
872        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
873        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
874        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
875    
876    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
877        being incorrectly compiled and would have given unpredicatble results.
878    
879    9.  A possessively repeated subpattern with minimum repeat count greater than
880        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
881        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
882        the first (A) could occur when it should not.
883    
884    10. Add a cast and remove a redundant test from the code.
885    
886    11. JIT should use pcre_malloc/pcre_free for allocation.
887    
888    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
889        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
890        is anything other than /usr, -L is still shown).
891    
892    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
893    
894    14. Perl does not support \N without a following name in a [] class; PCRE now
895        also gives an error.
896    
897    15. If a forward reference was repeated with an upper limit of around 2000,
898        it caused the error "internal error: overran compiling workspace". The
899        maximum number of forward references (including repeats) was limited by the
900        internal workspace, and dependent on the LINK_SIZE. The code has been
901        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
902        the default depends on LINK_SIZE. There is a new upper limit (for safety)
903        of around 200,000 forward references. While doing this, I also speeded up
904        the filling in of repeated forward references.
905    
906    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
907        incorrectly expecting the subject to contain another "a" after the start.
908    
909    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
910        in the match, the SKIP should be ignored. This was not happening; instead
911        the SKIP was being treated as NOMATCH. For patterns such as
912        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
913        tested.
914    
915    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
916        now much more compatible with Perl, in particular in cases where the result
917        is a non-match for a non-anchored pattern. For example, if
918        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
919        "m", where previously it did not return a name. A side effect of this
920        change is that for partial matches, the last encountered mark name is
921        returned, as for non matches. A number of tests that were previously not
922        Perl-compatible have been moved into the Perl-compatible test files. The
923        refactoring has had the pleasing side effect of removing one argument from
924        the match() function, thus reducing its stack requirements.
925    
926    19. If the /S+ option was used in pcretest to study a pattern using JIT,
927        subsequent uses of /S (without +) incorrectly behaved like /S+.
928    
929    21. Retrieve executable code size support for the JIT compiler and fixing
930        some warnings.
931    
932    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
933        not work when the shorter character appeared right at the end of the
934        subject string.
935    
936    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
937        systems.
938    
939    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
940        output it when the /M option is used in pcretest.
941    
942    25. The CheckMan script was not being included in the distribution. Also, added
943        an explicit "perl" to run Perl scripts from the PrepareRelease script
944        because this is reportedly needed in Windows.
945    
946    26. If study data was being save in a file and studying had not found a set of
947        "starts with" bytes for the pattern, the data written to the file (though
948        never used) was taken from uninitialized memory and so caused valgrind to
949        complain.
950    
951    27. Updated RunTest.bat as provided by Sheri Pierce.
952    
953    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
954    
955    29. Computation of memory usage for the table of capturing group names was
956        giving an unnecessarily large value.
957    
958    
959    Version 8.20 21-Oct-2011
960    ------------------------
961    
962    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
963        a POSIX class. After further experiments with Perl, which convinced me that
964        Perl has bugs and confusions, a closing square bracket is no longer allowed
965        in a POSIX name. This bug also affected patterns with classes that started
966        with full stops.
967    
968    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
969        captured substring, but while checking the failing first alternative,
970        substring 1 is temporarily captured. If the output vector supplied to
971        pcre_exec() was not big enough for this capture, the yield of the function
972        was still zero ("insufficient space for captured substrings"). This cannot
973        be totally fixed without adding another stack variable, which seems a lot
974        of expense for a edge case. However, I have improved the situation in cases
975        such as /(a)(b)x|abc/ matched against "abc", where the return code
976        indicates that fewer than the maximum number of slots in the ovector have
977        been set.
978    
979    3.  Related to (2) above: when there are more back references in a pattern than
980        slots in the output vector, pcre_exec() uses temporary memory during
981        matching, and copies in the captures as far as possible afterwards. It was
982        using the entire output vector, but this conflicts with the specification
983        that only 2/3 is used for passing back captured substrings. Now it uses
984        only the first 2/3, for compatibility. This is, of course, another edge
985        case.
986    
987    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
988        main code base, and can be used by building with --enable-jit. When this is
989        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
990        runtime --no-jit option is given.
991    
992    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
993        ovector, the return from the function was zero, implying that there were
994        other matches that did not fit. The correct "exactly full" value is now
995        returned.
996    
997    6.  If a subpattern that was called recursively or as a subroutine contained
998        (*PRUNE) or any other control that caused it to give a non-standard return,
999        invalid errors such as "Error -26 (nested recursion at the same subject
1000        position)" or even infinite loops could occur.
1001    
1002    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1003        computing the minimum length on reaching *ACCEPT, and so ended up with the
1004        wrong value of 1 rather than 0. Further investigation indicates that
1005        computing a minimum subject length in the presence of *ACCEPT is difficult
1006        (think back references, subroutine calls), and so I have changed the code
1007        so that no minimum is registered for a pattern that contains *ACCEPT.
1008    
1009    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1010        it was not handled as intended. [But see 16 below.]
1011    
1012    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1013        Sheri Pierce.
1014    
1015    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1016        the first byte in a match must be "a".
1017    
1018    11. Change 17 for 8.13 increased the recursion depth for patterns like
1019        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1020        pattern contains any instances of (*THEN). If it does not, the old
1021        optimizations are restored. It would be nice to do this on a per-group
1022        basis, but at the moment that is not feasible.
1023    
1024    12. In some environments, the output of pcretest -C is CRLF terminated. This
1025        broke RunTest's code that checks for the link size. A single white space
1026        character after the value is now allowed for.
1027    
1028    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1029        For "fr", it uses the Windows-specific input and output files.
1030    
1031    14. If (*THEN) appeared in a group that was called recursively or as a
1032        subroutine, it did not work as intended. [But see next item.]
1033    
1034    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1035        pattern fragments (but not containing any | characters). If A and B are
1036        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1037        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1038        D. In other words, Perl considers parentheses that do not contain any |
1039        characters to be part of a surrounding alternative, whereas PCRE was
1040        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1041        differently. PCRE now behaves in the same way as Perl, except in the case
1042        of subroutine/recursion calls such as (?1) which have in any case always
1043        been different (but PCRE had them first :-).
1044    
1045    16. Related to 15 above: Perl does not treat the | in a conditional group as
1046        creating alternatives. Such a group is treated in the same way as an
1047        ordinary group without any | characters when processing (*THEN). PCRE has
1048        been changed to match Perl's behaviour.
1049    
1050    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1051        RunGrepTest script failed.
1052    
1053    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1054        inevitable for groups that contain captures, but it can lead to a lot of
1055        stack use in large patterns. The old behaviour has been restored for atomic
1056        groups that do not contain any capturing parentheses.
1057    
1058    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1059        suppress the check for a minimum subject length at run time. (If it was
1060        given to pcre_exec() or pcre_dfa_exec() it did work.)
1061    
1062    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1063        fail to work when decoding hex characters in data strings in EBCDIC
1064        environments.
1065    
1066    21. It appears that in at least one Mac OS environment, the isxdigit() function
1067        is implemented as a macro that evaluates to its argument more than once,
1068        contravening the C 90 Standard (I haven't checked a later standard). There
1069        was an instance in pcretest which caused it to go wrong when processing
1070        \x{...} escapes in subject strings. The has been rewritten to avoid using
1071        things like p++ in the argument of isxdigit().
1072    
1073    
1074    Version 8.13 16-Aug-2011
1075    ------------------------
1076    
1077    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1078    
1079    2.  Two minor typos in pcre_internal.h have been fixed.
1080    
1081    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1082        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1083        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1084    
1085    4.  There were a number of related bugs in the code for matching backrefences
1086        caselessly in UTF-8 mode when codes for the characters concerned were
1087        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1088        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1089        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1090        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1091        2-byte code at the end of the subject (it thought there wasn't enough data
1092        left).
1093    
1094    5.  Comprehensive information about what went wrong is now returned by
1095        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1096        as the output vector has at least 2 elements. The offset of the start of
1097        the failing character and a reason code are placed in the vector.
1098    
1099    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1100        now returned is for the first byte of the failing character, instead of the
1101        last byte inspected. This is an incompatible change, but I hope it is small
1102        enough not to be a problem. It makes the returned offset consistent with
1103        pcre_exec() and pcre_dfa_exec().
1104    
1105    7.  pcretest now gives a text phrase as well as the error number when
1106        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1107        failure, the offset and reason code are output.
1108    
1109    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1110        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1111        back over a single character (\n). This seems wrong (because it treated the
1112        two characters as a single entity when going forwards), conflicts with the
1113        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1114        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1115        has been changed.
1116    
1117    9.  Some internal refactoring has changed the processing so that the handling
1118        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1119        time (the PCRE_DOTALL option was changed this way some time ago: version
1120        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1121        which was always a bit of a fudge. It also means that there is one less
1122        argument for the match() function, which reduces its stack requirements
1123        slightly. This change also fixes an incompatibility with Perl: the pattern
1124        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1125    
1126    10. More internal refactoring has drastically reduced the number of recursive
1127        calls to match() for possessively repeated groups such as (abc)++ when
1128        using pcre_exec().
1129    
1130    11. While implementing 10, a number of bugs in the handling of groups were
1131        discovered and fixed:
1132    
1133        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1134        (a|)*(?1) gave a compile-time internal error.
1135        ((a|)+)+  did not notice that the outer group could match an empty string.
1136        (^a|^)+   was not marked as anchored.
1137        (.*a|.*)+ was not marked as matching at start or after a newline.
1138    
1139    12. Yet more internal refactoring has removed another argument from the match()
1140        function. Special calls to this function are now indicated by setting a
1141        value in a variable in the "match data" data block.
1142    
1143    13. Be more explicit in pcre_study() instead of relying on "default" for
1144        opcodes that mean there is no starting character; this means that when new
1145        ones are added and accidentally left out of pcre_study(), testing should
1146        pick them up.
1147    
1148    14. The -s option of pcretest has been documented for ages as being an old
1149        synonym of -m (show memory usage). I have changed it to mean "force study
1150        for every regex", that is, assume /S for every regex. This is similar to -i
1151        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1152        using it. It makes it easier to run collections of tests with and without
1153        study enabled, and thereby test pcre_study() more easily. All the standard
1154        tests are now run with and without -s (but some patterns can be marked as
1155        "never study" - see 20 below).
1156    
1157    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1158        restoration of the capturing data to the outer values was not happening
1159        correctly.
1160    
1161    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1162        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1163        pattern had matched an empty string, and so incorrectly returned a no
1164        match.
1165    
1166    17. There was optimizing code for the last branch of non-capturing parentheses,
1167        and also for the obeyed branch of a conditional subexpression, which used
1168        tail recursion to cut down on stack usage. Unfortunately, now that there is
1169        the possibility of (*THEN) occurring in these branches, tail recursion is
1170        no longer possible because the return has to be checked for (*THEN). These
1171        two optimizations have therefore been removed. [But see 8.20/11 above.]
1172    
1173    18. If a pattern containing \R was studied, it was assumed that \R always
1174        matched two bytes, thus causing the minimum subject length to be
1175        incorrectly computed because \R can also match just one byte.
1176    
1177    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1178        was incorrectly computed.
1179    
1180    20. If /S is present twice on a test pattern in pcretest input, it now
1181        *disables* studying, thereby overriding the use of -s on the command line
1182        (see 14 above). This is necessary for one or two tests to keep the output
1183        identical in both cases.
1184    
1185    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1186        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1187    
1188    22. When an atomic group that contained a capturing parenthesis was
1189        successfully matched, but the branch in which it appeared failed, the
1190        capturing was not being forgotten if a higher numbered group was later
1191        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1192        group 1 to "a", when in fact it should be unset. This applied to multi-
1193        branched capturing and non-capturing groups, repeated or not, and also to
1194        positive assertions (capturing in negative assertions does not happen
1195        in PCRE) and also to nested atomic groups.
1196    
1197    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1198        subject after a captured substring, to make it easier to tell which of a
1199        number of identical substrings has been captured.
1200    
1201    24. The way atomic groups are processed by pcre_exec() has been changed so that
1202        if they are repeated, backtracking one repetition now resets captured
1203        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1204        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1205        "aaa". Previously, it would have been "a". As part of this code
1206        refactoring, the way recursive calls are handled has also been changed.
1207    
1208    25. If an assertion condition captured any substrings, they were not passed
1209        back unless some other capturing happened later. For example, if
1210        (?(?=(a))a) was matched against "a", no capturing was returned.
1211    
1212    26. When studying a pattern that contained subroutine calls or assertions,
1213        the code for finding the minimum length of a possible match was handling
1214        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1215        group 1 called group 2 while simultaneously a separate group 2 called group
1216        1). A stack overflow occurred in this case. I have fixed this by limiting
1217        the recursion depth to 10.
1218    
1219    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1220        Fortmann. This supports explicit test numbers on the command line, and has
1221        argument validation and error reporting.
1222    
1223    28. An instance of \X with an unlimited repeat could fail if at any point the
1224        first character it looked at was a mark character.
1225    
1226    29. Some minor code refactoring concerning Unicode properties and scripts
1227        should reduce the stack requirement of match() slightly.
1228    
1229    30. Added the '=' option to pcretest to check the setting of unused capturing
1230        slots at the end of the pattern, which are documented as being -1, but are
1231        not included in the return count.
1232    
1233    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1234        compiled something random. Now it gives a compile-time error (as does
1235        Perl).
1236    
1237    32. A *MARK encountered during the processing of a positive assertion is now
1238        recorded and passed back (compatible with Perl).
1239    
1240    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1241        had alternative anchored branches, the search for a second match in a line
1242        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1243        matched the line "0102" twice. The same bug affected patterns that started
1244        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1245        twice.
1246    
1247    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1248        does, and because of capturing effects, quantifying parenthesized
1249        assertions may at times be useful. Quantifiers are now allowed for
1250        parenthesized assertions.
1251    
1252    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1253    
1254    36. \g was being checked for fancy things in a character class, when it should
1255        just be a literal "g".
1256    
1257    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1258        appearance of a nested POSIX class supersedes an apparent external class.
1259        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1260        unescaped square brackets may also appear as part of class names. For
1261        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1262        more like Perl. (But see 8.20/1 above.)
1263    
1264    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1265        was because it thought it was \N{name}, which is not supported).
1266    
1267    39. Add minix to OS list not supporting the -S option in pcretest.
1268    
1269    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1270        cannot analyze patterns in sufficient detail to catch mutual recursions
1271        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1272        subgroup is called recursively as a subpattern for a second time at the
1273        same position in the subject string. In previous releases this might have
1274        been caught by the recursion limit, or it might have run out of stack.
1275    
1276    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1277        happen only once. PCRE was, however incorrectly giving a compile time error
1278        "recursive call could loop indefinitely" because it cannot analyze the
1279        pattern in sufficient detail. The compile time test no longer happens when
1280        PCRE is compiling a conditional subpattern, but actual runaway loops are
1281        now caught at runtime (see 40 above).
1282    
1283    42. It seems that Perl allows any characters other than a closing parenthesis
1284        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1285        has been changed to be the same.
1286    
1287    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1288        as not to get warnings when autogen.sh is called. Also changed
1289        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1290    
1291    44. To help people who use pcregrep to scan files containing exceedingly long
1292        lines, the following changes have been made:
1293    
1294        (a) The default value of the buffer size parameter has been increased from
1295            8K to 20K. (The actual buffer used is three times this size.)
1296    
1297        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1298            PCRE is built.
1299    
1300        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1301            to be set at run time.
1302    
1303        (d) Numerical values in pcregrep options can be followed by K or M, for
1304            example --buffer-size=50K.
1305    
1306        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1307            given and the return code is set to 2.
1308    
1309    45. Add a pointer to the latest mark to the callout data block.
1310    
1311    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1312        partial match of an empty string instead of no match. This was specific to
1313        the use of ".".
1314    
1315    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1316        complete match instead of a partial match. This bug was dependent on both
1317        the PCRE_UTF8 and PCRE_DOTALL options being set.
1318    
1319    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1320        starting byte set, because \b was not being ignored.
1321    
1322    
1323    Version 8.12 15-Jan-2011
1324    ------------------------
1325    
1326    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1327        checks for such things as part of the documentation building process.
1328    
1329    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1330        --match-limit and --recursion-limit options (added for 8.11). In
1331        particular, this made one of the standard tests fail. (The integer value
1332        went into the wrong half of a long int.)
1333    
1334    3.  If the --colour option was given to pcregrep with -v (invert match), it
1335        did strange things, either producing crazy output, or crashing. It should,
1336        of course, ignore a request for colour when reporting lines that do not
1337        match.
1338    
1339    4.  Another pcregrep bug caused similar problems if --colour was specified with
1340        -M (multiline) and the pattern match finished with a line ending.
1341    
1342    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1343        matched in multiline mode, the following line was shown as part of the
1344        match. This seems wrong, so I have changed it.
1345    
1346    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1347        the check for further matches in the same line (so they could be coloured)
1348        to overrun the end of the current line. If another match was found, it was
1349        incorrectly shown (and then shown again when found in the next line).
1350    
1351    7.  If pcregrep was compiled under Windows, there was a reference to the
1352        function pcregrep_exit() before it was defined. I am assuming this was
1353        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1354        reported by a user. I've moved the definition above the reference.
1355    
1356    
1357    Version 8.11 10-Dec-2010
1358    ------------------------
1359    
1360    1.  (*THEN) was not working properly if there were untried alternatives prior
1361        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1362        backtracked to try for "b" instead of moving to the next alternative branch
1363        at the same level (in this case, to look for "c"). The Perl documentation
1364        is clear that when (*THEN) is backtracked onto, it goes to the "next
1365        alternative in the innermost enclosing group".
1366    
1367    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1368        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1369        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1370        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1371        (*THEN).
1372    
1373    3.  If \s appeared in a character class, it removed the VT character from
1374        the class, even if it had been included by some previous item, for example
1375        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1376        of \s, but is part of the POSIX "space" class.)
1377    
1378    4.  A partial match never returns an empty string (because you can always
1379        match an empty string at the end of the subject); however the checking for
1380        an empty string was starting at the "start of match" point. This has been
1381        changed to the "earliest inspected character" point, because the returned
1382        data for a partial match starts at this character. This means that, for
1383        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1384        (previously it gave "no match").
1385    
1386    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1387        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1388        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1389        has an implication that the given string is incomplete (because a partial
1390        match is preferred over a full match). For this reason, these items now
1391        give a partial match in this situation. [Aside: previously, the one case
1392        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1393        match rather than a full match, which was wrong by the old rules, but is
1394        now correct.]
1395    
1396    6.  There was a bug in the handling of #-introduced comments, recognized when
1397        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1398        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1399        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1400        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1401        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1402        places in pcre_compile().
1403    
1404    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1405        comments when looking ahead for named forward references to subpatterns,
1406        the only newline sequence it recognized was NL. It now handles newlines
1407        according to the set newline convention.
1408    
1409    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1410        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1411        cater for a lack of strerror(). These oversights have been fixed.
1412    
1413    9.  Added --match-limit and --recursion-limit to pcregrep.
1414    
1415    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1416    
1417    11. When the -o option was used, pcregrep was setting a return code of 1, even
1418        when matches were found, and --line-buffered was not being honoured.
1419    
1420    12. Added an optional parentheses number to the -o and --only-matching options
1421        of pcregrep.
1422    
1423    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1424        can match an empty string. The code to do it in pcretest and pcredemo
1425        needed fixing:
1426    
1427        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1428            only one byte after an empty string match just before CRLF (this case
1429            just got forgotten; "any" and "anycrlf" were OK).
1430    
1431        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1432            mode when an empty string match preceded an ASCII character followed by
1433            a non-ASCII character. (The code for advancing by one character rather
1434            than one byte was nonsense.)
1435    
1436        (c) The pcredemo.c sample program did not have any code at all to handle
1437            the cases when CRLF is a valid newline sequence.
1438    
1439    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1440        as a starting offset was within the subject string. There is now a new
1441        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1442        negative or greater than the length of the string. In order to test this,
1443        pcretest is extended to allow the setting of negative starting offsets.
1444    
1445    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1446        starting offset points to the beginning of a UTF-8 character was
1447        unnecessarily clumsy. I tidied it up.
1448    
1449    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1450        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1451    
1452    17. Nobody had reported that the --include_dir option, which was added in
1453        release 7.7 should have been called --include-dir (hyphen, not underscore)
1454        for compatibility with GNU grep. I have changed it to --include-dir, but
1455        left --include_dir as an undocumented synonym, and the same for
1456        --exclude-dir, though that is not available in GNU grep, at least as of
1457        release 2.5.4.
1458    
1459    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1460        characters from a string of bytes) have been redefined so as not to use
1461        loops, in order to improve performance in some environments. At the same
1462        time, I abstracted some of the common code into auxiliary macros to save
1463        repetition (this should not affect the compiled code).
1464    
1465    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1466        compile-time error is now given if \c is not followed by an ASCII
1467        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1468        different, and any byte value is allowed.)
1469    
1470    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1471        START_OPTIMIZE option, which is now allowed at compile time - but just
1472        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1473        to pcregrep and other applications that have no direct access to PCRE
1474        options. The new /Y option in pcretest sets this option when calling
1475        pcre_compile().
1476    
1477    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1478        back references. Groups containing recursive back references were forced to
1479        be atomic by that change, but in the case of named groups, the amount of
1480        memory required was incorrectly computed, leading to "Failed: internal
1481        error: code overflow". This has been fixed.
1482    
1483    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1484        pcretest.c, to avoid build problems in some Borland environments.
1485    
1486    
1487    Version 8.10 25-Jun-2010
1488    ------------------------
1489    
1490    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1491        THEN.
1492    
1493    2.  (*ACCEPT) was not working when inside an atomic group.
1494    
1495    3.  Inside a character class, \B is treated as a literal by default, but
1496        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1497        causes the error). The code is unchanged, but I tidied the documentation.
1498    
1499    4.  Inside a character class, PCRE always treated \R and \X as literals,
1500        whereas Perl faults them if its -w option is set. I have changed PCRE so
1501        that it faults them when PCRE_EXTRA is set.
1502    
1503    5.  Added support for \N, which always matches any character other than
1504        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1505    
1506    6.  When compiling pcregrep with newer versions of gcc which may have
1507        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1508        declared with attribute warn_unused_result" were given. Just casting the
1509        result to (void) does not stop the warnings; a more elaborate fudge is
1510        needed. I've used a macro to implement this.
1511    
1512    7.  Minor change to pcretest.c to avoid a compiler warning.
1513    
1514    8.  Added four artifical Unicode properties to help with an option to make
1515        \s etc use properties (see next item). The new properties are: Xan
1516        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1517    
1518    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1519        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1520        this option. Modified pcretest to add /W to test this facility. Added
1521        REG_UCP to make it available via the POSIX interface.
1522    
1523    10. Added --line-buffered to pcregrep.
1524    
1525    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1526        studied, and the match started with a letter with a code point greater than
1527        127 whose first byte was different to the first byte of the other case of
1528        the letter, the other case of this starting letter was not recognized
1529        (#976).
1530    
1531    12. If a pattern that was studied started with a repeated Unicode property
1532        test, for example, \p{Nd}+, there was the theoretical possibility of
1533        setting up an incorrect bitmap of starting bytes, but fortunately it could
1534        not have actually happened in practice until change 8 above was made (it
1535        added property types that matched character-matching opcodes).
1536    
1537    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1538        possible starting bytes for non-anchored patterns.
1539    
1540    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1541        \R, and also a number of cases that involve Unicode properties, both
1542        explicit and implicit when PCRE_UCP is set.
1543    
1544    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1545        input, it could crash or give wrong results if characters with values
1546        greater than 0xc0 were present in the subject string. (Detail: it assumed
1547        UTF-8 input when processing these items.)
1548    
1549    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1550        size_t is 64-bit (#991).
1551    
1552    17. Added a check for running out of memory when PCRE is compiled with
1553        --disable-stack-for-recursion (#990).
1554    
1555    18. If the last data line in a file for pcretest does not have a newline on
1556        the end, a newline was missing in the output.
1557    
1558    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1559        less than 128) in its various bitmaps. However, there is a facility for
1560        generating tables according to the current locale when PCRE is compiled. It
1561        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1562        characters, are recognized by isspace() and therefore were getting set in
1563        these tables, and indeed these tables seem to approximate to ISO 8859. This
1564        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1565        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1566        which of course cannot start UTF-8 characters. I have changed the code so
1567        that only real ASCII characters (less than 128) and the correct starting
1568        bytes for UTF-8 encodings are set for characters greater than 127 when in
1569        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1570        altogether.)
1571    
1572    20. Added the /T option to pcretest so as to be able to run tests with non-
1573        standard character tables, thus making it possible to include the tests
1574        used for 19 above in the standard set of tests.
1575    
1576    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1577        reference to a subpattern the other side of a comment that contains an
1578        opening parenthesis caused either an internal compiling error, or a
1579        reference to the wrong subpattern.
1580    
1581    
1582    Version 8.02 19-Mar-2010
1583    ------------------------
1584    
1585    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1586    
1587    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1588        configured.
1589    
1590    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1591        original author of that file, following a query about its status.
1592    
1593    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1594        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1595    
1596    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1597        quantifier applied to a forward-referencing subroutine call, could compile
1598        incorrect code or give the error "internal error: previously-checked
1599        referenced subpattern not found".
1600    
1601    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1602        variables to point to external functions. For these systems, therefore,
1603        pcre_malloc etc. are now initialized to local functions that call the
1604        relevant global functions.
1605    
1606    7.  There were two entries missing in the vectors called coptable and poptable
1607        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1608        I've fixed the data, and added a kludgy way of testing at compile time that
1609        the lengths are correct (equal to the number of opcodes).
1610    
1611    8.  Following on from 7, I added a similar kludge to check the length of the
1612        eint vector in pcreposix.c.
1613    
1614    9.  Error texts for pcre_compile() are held as one long string to avoid too
1615        much relocation at load time. To find a text, the string is searched,
1616        counting zeros. There was no check for running off the end of the string,
1617        which could happen if a new error number was added without updating the
1618        string.
1619    
1620    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1621    
1622    11. \K was not working if it appeared in an atomic group or in a group that
1623        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1624        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1625        the assertion is positive, but not if it is negative.
1626    
1627    12. Change 11 fortuitously reduced the size of the stack frame used in the
1628        "match()" function of pcre_exec.c by one pointer. Forthcoming
1629        implementation of support for (*MARK) will need an extra pointer on the
1630        stack; I have reserved it now, so that the stack frame size does not
1631        decrease.
1632    
1633    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1634        item in branch that calls a recursion is a subroutine call - as in the
1635        second branch in the above example - was incorrectly given the compile-
1636        time error "recursive call could loop indefinitely" because pcre_compile()
1637        was not correctly checking the subroutine for matching a non-empty string.
1638    
1639    14. The checks for overrunning compiling workspace could trigger after an
1640        overrun had occurred. This is a "should never occur" error, but it can be
1641        triggered by pathological patterns such as hundreds of nested parentheses.
1642        The checks now trigger 100 bytes before the end of the workspace.
1643    
1644    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1645    
1646    
1647    Version 8.01 19-Jan-2010
1648    ------------------------
1649    
1650    1.  If a pattern contained a conditional subpattern with only one branch (in
1651        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1652        computed the wrong minimum data length (which is of course zero for such
1653        subpatterns). This could cause incorrect "no match" results.
1654    
1655    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1656        the pattern is reset in the first branch, pcre_compile() failed with
1657        "internal error: code overflow at offset...". This happened only when
1658        the reset was to the original external option setting. (An optimization
1659        abstracts leading options settings into an external setting, which was the
1660        cause of this.)
1661    
1662    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1663        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1664        assertion pattern did not match (meaning that the assertion was true), it
1665        was incorrectly treated as false if the SKIP had been reached during the
1666        matching. This also applied to assertions used as conditions.
1667    
1668    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1669        assertion subpattern, including such a pattern used as a condition,
1670        unpredictable results occurred, instead of the error return
1671        PCRE_ERROR_DFA_UITEM.
1672    
1673    5.  The C++ GlobalReplace function was not working like Perl for the special
1674        situation when an empty string is matched. It now does the fancy magic
1675        stuff that is necessary.
1676    
1677    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1678        removed. (These were left over from very, very early versions of PCRE.)
1679    
1680    7.  Some cosmetic changes to the code to make life easier when compiling it
1681        as part of something else:
1682    
1683        (a) Change DEBUG to PCRE_DEBUG.
1684    
1685        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1686            called "current" as "current_branch", to prevent a collision with the
1687            Linux macro when compiled as a kernel module.
1688    
1689        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1690            prevent a collision with the Linux macro when compiled as a kernel
1691            module.
1692    
1693    8.  In pcre_compile() there are some checks for integer overflows that used to
1694        cast potentially large values to (double). This has been changed to that
1695        when building, a check for int64_t is made, and if it is found, it is used
1696        instead, thus avoiding the use of floating point arithmetic. (There is no
1697        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1698        double.
1699    
1700    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1701        2005 (difference between two addresses compared to an unsigned value).
1702    
1703    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1704        custom one, because of the following reported problem in Windows:
1705    
1706          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1707              under Win32.
1708          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1709              therefore missing the function definition.
1710          - The compiler thus generates a "C" signature for the test function.
1711          - The linker fails to find the "C" function.
1712          - PCRE fails to configure if asked to do so against libbz2.
1713    
1714    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1715        messages were output:
1716    
1717          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1718          rerunning libtoolize, to keep the correct libtool macros in-tree.
1719          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1720    
1721        I have done both of these things.
1722    
1723    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1724        most of the time, it *can* run out if it is given a pattern that contains a
1725        runaway infinite recursion. I updated the discussion in the pcrestack man
1726        page.
1727    
1728    13. Now that we have gone to the x.xx style of version numbers, the minor
1729        version may start with zero. Using 08 or 09 is a bad idea because users
1730        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1731        interpreted as invalid octal numbers. I've updated the previous comment in
1732        configure.ac, and also added a check that gives an error if 08 or 09 are
1733        used.
1734    
1735    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1736        causing partial matching to fail when the end of the subject matched \W
1737        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1738    
1739    15. There were some discrepancies between the declarations in pcre_internal.h
1740        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1741        their definitions. The declarations used "const uschar *" and the
1742        definitions used USPTR. Even though USPTR is normally defined as "const
1743        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1744        reported that: "This difference in casting confuses some C++ compilers, for
1745        example, SunCC recognizes above declarations as different functions and
1746        generates broken code for hbpcre." I have changed the declarations to use
1747        USPTR.
1748    
1749    16. GNU libtool is named differently on some systems. The autogen.sh script now
1750        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1751        (FreeBSD).
1752    
1753    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1754        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1755        comment: "Figure out how to create a longlong from a string: strtoll and
1756        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1757        instance, but it only takes 2 args instead of 3!"
1758    
1759    18. A subtle bug concerned with back references has been fixed by a change of
1760        specification, with a corresponding code fix. A pattern such as
1761        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1762        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1763        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1764        same bug. Such groups have to be quantified to be useful, or contained
1765        inside another quantified group. (If there's no repetition, the reference
1766        can never match.) The problem arises because, having left the group and
1767        moved on to the rest of the pattern, a later failure that backtracks into
1768        the group uses the captured value from the final iteration of the group
1769        rather than the correct earlier one. I have fixed this in PCRE by forcing
1770        any group that contains a reference to itself to be an atomic group; that
1771        is, there cannot be any backtracking into it once it has completed. This is
1772        similar to recursive and subroutine calls.
1773    
1774    
1775    Version 8.00 19-Oct-09
1776  ----------------------  ----------------------
1777    
1778  1.  The table for translating pcre_compile() error codes into POSIX error codes  1.  The table for translating pcre_compile() error codes into POSIX error codes
1779      was out-of-date, and there was no check on the pcre_compile() error code      was out-of-date, and there was no check on the pcre_compile() error code
1780      being within the table. This could lead to an OK return being given in      being within the table. This could lead to an OK return being given in
1781      error.      error.
1782    
1783  2.  Changed the call to open a subject file in pcregrep from fopen(pathname,  2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1784      "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests      "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1785      in a Windows environment.      in a Windows environment.
1786    
1787  3.  The pcregrep --count option prints the count for each file even when it is  3.  The pcregrep --count option prints the count for each file even when it is
1788      zero, as does GNU grep. However, pcregrep was also printing all files when      zero, as does GNU grep. However, pcregrep was also printing all files when
1789      --files-with-matches was added. Now, when both options are given, it prints      --files-with-matches was added. Now, when both options are given, it prints
1790      counts only for those files that have at least one match. (GNU grep just      counts only for those files that have at least one match. (GNU grep just
1791      prints the file name in this circumstance, but including the count seems      prints the file name in this circumstance, but including the count seems
1792      more useful - otherwise, why use --count?) Also ensured that the      more useful - otherwise, why use --count?) Also ensured that the
1793      combination -clh just lists non-zero counts, with no names.      combination -clh just lists non-zero counts, with no names.
1794    
1795  4.  The long form of the pcregrep -F option was incorrectly implemented as  4.  The long form of the pcregrep -F option was incorrectly implemented as
1796      --fixed_strings instead of --fixed-strings. This is an incompatible change,      --fixed_strings instead of --fixed-strings. This is an incompatible change,
1797      but it seems right to fix it, and I didn't think it was worth preserving      but it seems right to fix it, and I didn't think it was worth preserving
1798      the old behaviour.      the old behaviour.
1799    
1800  5.  The command line items --regex=pattern and --regexp=pattern were not  5.  The command line items --regex=pattern and --regexp=pattern were not
1801      recognized by pcregrep, which required --regex pattern or --regexp pattern      recognized by pcregrep, which required --regex pattern or --regexp pattern
1802      (with a space rather than an '='). The man page documented the '=' forms,      (with a space rather than an '='). The man page documented the '=' forms,
1803      which are compatible with GNU grep; these now work.      which are compatible with GNU grep; these now work.
1804    
1805  6.  No libpcreposix.pc file was created for pkg-config; there was just  6.  No libpcreposix.pc file was created for pkg-config; there was just
1806      libpcre.pc and libpcrecpp.pc. The omission has been rectified.      libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1807    
1808  7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size  7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1809      when UCP support is not needed, by modifying the Python script that      when UCP support is not needed, by modifying the Python script that
1810      generates it from Unicode data files. This should not matter if the module      generates it from Unicode data files. This should not matter if the module
1811      is correctly used as a library, but I received one complaint about 50K of      is correctly used as a library, but I received one complaint about 50K of
1812      unwanted data. My guess is that the person linked everything into his      unwanted data. My guess is that the person linked everything into his
1813      program rather than using a library. Anyway, it does no harm.      program rather than using a library. Anyway, it does no harm.
1814    
1815    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1816        was a minimum greater than 1 for a wide character in a possessive
1817        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1818        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1819        character. Chaos in the form of incorrect output or a compiling loop could
1820        result.
1821    
1822    9.  The restrictions on what a pattern can contain when partial matching is
1823        requested for pcre_exec() have been removed. All patterns can now be
1824        partially matched by this function. In addition, if there are at least two
1825        slots in the offset vector, the offset of the earliest inspected character
1826        for the match and the offset of the end of the subject are set in them when
1827        PCRE_ERROR_PARTIAL is returned.
1828    
1829    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1830        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1831        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1832        and may be more useful for multi-segment matching.
1833    
1834    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1835        used to be given if ever the end of the subject was reached; now it is
1836        given only if matching could not proceed because another character was
1837        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1838        string "Z", which now yields "no match" instead of "partial match". In the
1839        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1840        final character ended with (*FAIL).
1841    
1842    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1843        if the pattern had a "must contain" character that was already found in the
1844        earlier partial match, unless partial matching was again requested. For
1845        example, with the pattern /dog.(body)?/, the "must contain" character is
1846        "g". If the first part-match was for the string "dog", restarting with
1847        "sbody" failed. This bug has been fixed.
1848    
1849    13. The string returned by pcre_dfa_exec() after a partial match has been
1850        changed so that it starts at the first inspected character rather than the
1851        first character of the match. This makes a difference only if the pattern
1852        starts with a lookbehind assertion or \b or \B (\K is not supported by
1853        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1854        matching functions compatible, and I think it's the right thing to do.
1855    
1856    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1857        so that the demonstration program is easily available in environments where
1858        PCRE has not been installed from source.
1859    
1860    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1861        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1862        library.
1863    
1864    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1865        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1866        is not the first non-POSIX option to be added. Clearly some people find
1867        these options useful.
1868    
1869    17. If a caller to the POSIX matching function regexec() passes a non-zero
1870        value for nmatch with a NULL value for pmatch, the value of
1871        nmatch is forced to zero.
1872    
1873    18. RunGrepTest did not have a test for the availability of the -u option of
1874        the diff command, as RunTest does. It now checks in the same way as
1875        RunTest, and also checks for the -b option.
1876    
1877    19. If an odd number of negated classes containing just a single character
1878        interposed, within parentheses, between a forward reference to a named
1879        subpattern and the definition of the subpattern, compilation crashed with
1880        an internal error, complaining that it could not find the referenced
1881        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1882        [The bug was that it was starting one character too far in when skipping
1883        over the character class, thus treating the ] as data rather than
1884        terminating the class. This meant it could skip too much.]
1885    
1886    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1887        /g option in pcretest when the pattern contains \K, which makes it possible
1888        to have an empty string match not at the start, even when the pattern is
1889        anchored. Updated pcretest and pcredemo to use this option.
1890    
1891    21. If the maximum number of capturing subpatterns in a recursion was greater
1892        than the maximum at the outer level, the higher number was returned, but
1893        with unset values at the outer level. The correct (outer level) value is
1894        now given.
1895    
1896    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1897        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1898        make it do so. The string so far is captured, making this feature
1899        compatible with Perl.
1900    
1901    23. The tests have been re-organized, adding tests 11 and 12, to make it
1902        possible to check the Perl 5.10 features against Perl 5.10.
1903    
1904    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1905        pattern matches a fixed length string. PCRE did not allow this; now it
1906        does. Neither allows recursion.
1907    
1908    25. I finally figured out how to implement a request to provide the minimum
1909        length of subject string that was needed in order to match a given pattern.
1910        (It was back references and recursion that I had previously got hung up
1911        on.) This code has now been added to pcre_study(); it finds a lower bound
1912        to the length of subject needed. It is not necessarily the greatest lower
1913        bound, but using it to avoid searching strings that are too short does give
1914        some useful speed-ups. The value is available to calling programs via
1915        pcre_fullinfo().
1916    
1917    26. While implementing 25, I discovered to my embarrassment that pcretest had
1918        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1919        study optimizations had never been tested with that matching function.
1920        Oops. What is worse, even when it was passed study data, there was a bug in
1921        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1922        were also very few tests of studied patterns with pcre_dfa_exec().
1923    
1924    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1925        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1926        on the other side of the coin, they are no longer allowed to have different
1927        names, because these cannot be distinguished in PCRE, and this has caused
1928        confusion. (This is a difference from Perl.)
1929    
1930    28. When duplicate subpattern names are present (necessarily with different
1931        numbers, as required by 27 above), and a test is made by name in a
1932        conditional pattern, either for a subpattern having been matched, or for
1933        recursion in such a pattern, all the associated numbered subpatterns are
1934        tested, and the overall condition is true if the condition is true for any
1935        one of them. This is the way Perl works, and is also more like the way
1936        testing by number works.
1937    
1938    
1939  Version 7.9 11-Apr-09  Version 7.9 11-Apr-09
1940  ---------------------  ---------------------
# Line 76  Version 7.9 11-Apr-09 Line 1970  Version 7.9 11-Apr-09
1970  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
1971      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
1972      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1973      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
1974        8.33.]
1975    
1976  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
1977      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 319  Version 7.7 07-May-08 Line 2214  Version 7.7 07-May-08
2214      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2215      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2216      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2217      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2218      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2219      was subsequently set up correctly.)      was subsequently set up correctly.)
2220    

Legend:
Removed from v.425  
changed lines
  Added in v.1449

  ViewVC Help
Powered by ViewVC 1.1.5