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

Diff of /code/trunk/ChangeLog

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

revision 319 by ph10, Mon Jan 28 15:27:45 2008 UTC revision 1434 by zherczeg, Mon Jan 6 20:04:50 2014 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    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    
55    Version 8.34 15-December-2013
56    -----------------------------
57    
58    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
59        executable memory. Patch inspired by Carsten Klein.
60    
61    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
62        this macro is never tested and has no effect, because the work to support
63        coverage involves only compiling and linking options and special targets in
64        the Makefile. The comment in config.h implied that defining the macro would
65        enable coverage support, which is totally false. There was also support for
66        setting this macro in the CMake files (my fault, I just copied it from
67        configure). SUPPORT_GCOV has now been removed.
68    
69    3.  Make a small performance improvement in strlen16() and strlen32() in
70        pcretest.
71    
72    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
73        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
74        these cases). There was also one in pcretest.c.
75    
76    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
77    
78    6.  In UTF mode, the code for checking whether a group could match an empty
79        string (which is used for indefinitely repeated groups to allow for
80        breaking an infinite loop) was broken when the group contained a repeated
81        negated single-character class with a character that occupied more than one
82        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
83        UTF-8 mode). The effect was undefined: the group might or might not be
84        deemed as matching an empty string, or the program might have crashed.
85    
86    7.  The code for checking whether a group could match an empty string was not
87        recognizing that \h, \H, \v, \V, and \R must match a character.
88    
89    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
90        an empty string. If it can, pcretest shows this in its information output.
91    
92    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
93        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
94        matched by pcre_exec() without using JIT:
95    
96        (a) If the rest of the pattern did not match after a maximal run of
97            grapheme clusters, the code for backing up to try with fewer of them
98            did not always back up over a full grapheme when characters that do not
99            have the modifier quality were involved, e.g. Hangul syllables.
100    
101        (b) If the match point in a subject started with modifier character, and
102            there was no match, the code could incorrectly back up beyond the match
103            point, and potentially beyond the first character in the subject,
104            leading to a segfault or an incorrect match result.
105    
106    10. A conditional group with an assertion condition could lead to PCRE
107        recording an incorrect first data item for a match if no other first data
108        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
109        first data item, and therefore matched "ca" after "c" instead of at the
110        start.
111    
112    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
113        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
114    
115    12. The source of pcregrep now includes z/OS-specific code so that it can be
116        compiled for z/OS as part of the special z/OS distribution.
117    
118    13. Added the -T and -TM options to pcretest.
119    
120    14. The code in pcre_compile.c for creating the table of named capturing groups
121        has been refactored. Instead of creating the table dynamically during the
122        actual compiling pass, the information is remembered during the pre-compile
123        pass (on the stack unless there are more than 20 named groups, in which
124        case malloc() is used) and the whole table is created before the actual
125        compile happens. This has simplified the code (it is now nearly 150 lines
126        shorter) and prepared the way for better handling of references to groups
127        with duplicate names.
128    
129    15. A back reference to a named subpattern when there is more than one of the
130        same name now checks them in the order in which they appear in the pattern.
131        The first one that is set is used for the reference. Previously only the
132        first one was inspected. This change makes PCRE more compatible with Perl.
133    
134    16. Unicode character properties were updated from Unicode 6.3.0.
135    
136    17. The compile-time code for auto-possessification has been refactored, based
137        on a patch by Zoltan Herczeg. It now happens after instead of during
138        compilation. The code is cleaner, and more cases are handled. The option
139        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
140        options in pcretest are provided to set it. It can also be set by
141        (*NO_AUTO_POSSESS) at the start of a pattern.
142    
143    18. The character VT has been added to the default ("C" locale) set of
144        characters that match \s and are generally treated as white space,
145        following this same change in Perl 5.18. There is now no difference between
146        "Perl space" and "POSIX space". Whether VT is treated as white space in
147        other locales depends on the locale.
148    
149    19. The code for checking named groups as conditions, either for being set or
150        for being recursed, has been refactored (this is related to 14 and 15
151        above). Processing unduplicated named groups should now be as fast at
152        numerical groups, and processing duplicated groups should be faster than
153        before.
154    
155    20. Two patches to the CMake build system, by Alexander Barkov:
156    
157          (1) Replace the "source" command by "." in CMakeLists.txt because
158              "source" is a bash-ism.
159    
160          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
161              without these the CMake build does not work on Solaris.
162    
163    21. Perl has changed its handling of \8 and \9. If there is no previously
164        encountered capturing group of those numbers, they are treated as the
165        literal characters 8 and 9 instead of a binary zero followed by the
166        literals. PCRE now does the same.
167    
168    22. Following Perl, added \o{} to specify codepoints in octal, making it
169        possible to specify values greater than 0777 and also making them
170        unambiguous.
171    
172    23. Perl now gives an error for missing closing braces after \x{... instead of
173        treating the string as literal. PCRE now does the same.
174    
175    24. RunTest used to grumble if an inappropriate test was selected explicitly,
176        but just skip it when running all tests. This make it awkward to run ranges
177        of tests when one of them was inappropriate. Now it just skips any
178        inappropriate tests, as it always did when running all tests.
179    
180    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
181        character types such as \d or \w, too many callouts were inserted, and the
182        data that they returned was rubbish.
183    
184    26. In UCP mode, \s was not matching two of the characters that Perl matches,
185        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
186        were matched by \h. The code has now been refactored so that the lists of
187        the horizontal and vertical whitespace characters used for \h and \v (which
188        are defined only in one place) are now also used for \s.
189    
190    27. Add JIT support for the 64 bit TileGX architecture.
191        Patch by Jiong Wang (Tilera Corporation).
192    
193    28. Possessive quantifiers for classes (both explicit and automatically
194        generated) now use special opcodes instead of wrapping in ONCE brackets.
195    
196    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
197        (because it's meaningless), this was not happening when PCRE_CASELESS was
198        set. Not wrong, but inefficient.
199    
200    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
201        properties for \w, \d, etc) is present in a test regex. Otherwise if the
202        test contains no characters greater than 255, Perl doesn't realise it
203        should be using Unicode semantics.
204    
205    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
206        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
207        does in Unicode mode.
208    
209    32. Added the "forbid" facility to pcretest so that putting tests into the
210        wrong test files can sometimes be quickly detected.
211    
212    33. There is now a limit (default 250) on the depth of nesting of parentheses.
213        This limit is imposed to control the amount of system stack used at compile
214        time. It can be changed at build time by --with-parens-nest-limit=xxx or
215        the equivalent in CMake.
216    
217    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
218        errors. Perl warns for these when in warning mode, but PCRE has no facility
219        for giving warnings.
220    
221    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
222        However, this was not working for (?!) because it is optimized to (*FAIL),
223        for which PCRE does not allow quantifiers. The optimization is now disabled
224        when a quantifier follows (?!). I can't see any use for this, but it makes
225        things uniform.
226    
227    36. Perl no longer allows group names to start with digits, so I have made this
228        change also in PCRE. It simplifies the code a bit.
229    
230    37. In extended mode, Perl ignores spaces before a + that indicates a
231        possessive quantifier. PCRE allowed a space before the quantifier, but not
232        before the possessive +. It now does.
233    
234    38. The use of \K (reset reported match start) within a repeated possessive
235        group such as (a\Kb)*+ was not working.
236    
237    40. Document that the same character tables must be used at compile time and
238        run time, and that the facility to pass tables to pcre_exec() and
239        pcre_dfa_exec() is for use only with saved/restored patterns.
240    
241    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
242        features for Builds with MSVC:
243    
244        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
245           builds). The .rc files can be used to set FileDescription and many other
246           attributes.
247    
248        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
249           This allows higher-level build scripts which want .pdb files to avoid
250           hard-coding the exact files needed."
251    
252    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
253        mean "start of word" and "end of word", respectively, as a transition aid.
254    
255    43. A minimizing repeat of a class containing codepoints greater than 255 in
256        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
257        compiled to use the heap for recursion.
258    
259    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
260        is configured.
261    
262    
263    Version 8.33 28-May-2013
264    ------------------------
265    
266    1.  Added 'U' to some constants that are compared to unsigned integers, to
267        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
268        variables that are added to signed variables, to ensure the result is
269        signed and can be negated.
270    
271    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
272        CMake config files.
273    
274    3.  Revise the creation of config.h.generic so that all boolean macros are
275        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
276        overriding via -D on the command line possible.
277    
278    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
279        to unsigned int is reported to make a quite noticeable speed difference in
280        a specific Windows environment. Testing on Linux did also appear to show
281        some benefit (and it is clearly not harmful). Also fixed the definition of
282        Xop which should be unsigned.
283    
284    5.  Related to (4), changing the definition of the intermediate variable cc
285        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
286        improvements.
287    
288    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
289        unnecessary spaces.
290    
291    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
292        and later.
293    
294    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
295    
296    9.  Optimizing fast_forward_start_bits in JIT.
297    
298    10. Adding support for callouts in JIT, and fixing some issues revealed
299        during this work. Namely:
300    
301        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
302    
303        (b) Minimum length was not checked before the matching is started.
304    
305    11. The value of capture_last that is passed to callouts was incorrect in some
306        cases when there was a capture on one path that was subsequently abandoned
307        after a backtrack. Also, the capture_last value is now reset after a
308        recursion, since all captures are also reset in this case.
309    
310    12. The interpreter no longer returns the "too many substrings" error in the
311        case when an overflowing capture is in a branch that is subsequently
312        abandoned after a backtrack.
313    
314    13. In the pathological case when an offset vector of size 2 is used, pcretest
315        now prints out the matched string after a yield of 0 or 1.
316    
317    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
318        Only supported by the JIT compiler at the moment.
319    
320    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
321    
322    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
323        offset of the starting point of the matching process, provided the offsets
324        vector is large enough.
325    
326    17. The \A escape now records a lookbehind value of 1, though its execution
327        does not actually inspect the previous character. This is to ensure that,
328        in partial multi-segment matching, at least one character from the old
329        segment is retained when a new segment is processed. Otherwise, if there
330        are no lookbehinds in the pattern, \A might match incorrectly at the start
331        of a new segment.
332    
333    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
334    
335    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
336        gives some modest performance improvement in 8-bit mode.
337    
338    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
339        be expressed in certain programming languages using Universal Character
340        Names.
341    
342    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
343        which points out that "non characters" are not "characters that may not
344        appear in Unicode strings" but rather "characters that are reserved for
345        internal use and have only local meaning".
346    
347    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
348        there was a conditional group that depended on an assertion, if the
349        assertion was false, the callout that immediately followed the alternation
350        in the condition was skipped when pcre_exec() was used for matching.
351    
352    23. Allow an explicit callout to be inserted before an assertion that is the
353        condition for a conditional group, for compatibility with automatic
354        callouts, which always insert a callout at this point.
355    
356    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
357        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
358    
359    25. (*PRUNE) is now supported by the JIT compiler.
360    
361    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
362    
363    27. Fix the case where there are two or more SKIPs with arguments that may be
364        ignored.
365    
366    28. (*SKIP) is now supported by the JIT compiler.
367    
368    29. (*THEN) is now supported by the JIT compiler.
369    
370    30. Update RunTest with additional test selector options.
371    
372    31. The way PCRE handles backtracking verbs has been changed in two ways.
373    
374        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
375        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
376        backtracking. In some cases this makes it more Perl-compatible, but Perl's
377        rather obscure rules do not always do the same thing.
378    
379        (2) Previously, backtracking verbs were confined within assertions. This is
380        no longer the case for positive assertions, except for (*ACCEPT). Again,
381        this sometimes improves Perl compatibility, and sometimes does not.
382    
383    32. A number of tests that were in test 2 because Perl did things differently
384        have been moved to test 1, because either Perl or PCRE has changed, and
385        these tests are now compatible.
386    
387    32. Backtracking control verbs are now handled in the same way in JIT and
388        interpreter.
389    
390    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
391        contained a forward subroutine reference caused a compile error.
392    
393    34. Auto-detect and optimize limited repetitions in JIT.
394    
395    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
396        blocking (*UTF) etc.
397    
398    36. In the interpreter, maximizing pattern repetitions for characters and
399        character types now use tail recursion, which reduces stack usage.
400    
401    37. The value of the max lookbehind was not correctly preserved if a compiled
402        and saved regex was reloaded on a host of different endianness.
403    
404    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
405        of the compiled pattern block, expand the flags field from 16 to 32 bits
406        because it was almost full.
407    
408    39. Try madvise first before posix_madvise.
409    
410    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
411        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
412        original change locked out all matches of empty strings. This has been
413        changed so that one match of an empty string per line is recognized.
414        Subsequent searches on the same line (for colouring or for --only-matching,
415        for example) do not recognize empty strings.
416    
417    41. Applied a user patch to fix a number of spelling mistakes in comments.
418    
419    42. Data lines longer than 65536 caused pcretest to crash.
420    
421    43. Clarified the data type for length and startoffset arguments for pcre_exec
422        and pcre_dfa_exec in the function-specific man pages, where they were
423        explicitly stated to be in bytes, never having been updated. I also added
424        some clarification to the pcreapi man page.
425    
426    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
427        a segmentation fault.
428    
429    
430    Version 8.32 30-November-2012
431    -----------------------------
432    
433    1.  Improved JIT compiler optimizations for first character search and single
434        character iterators.
435    
436    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
437        Patch by Daniel Richard G.
438    
439    3.  Single character iterator optimizations in the JIT compiler.
440    
441    4.  Improved JIT compiler optimizations for character ranges.
442    
443    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
444        Reported by Giuseppe D'Angelo.
445    
446    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
447        of a line, was being set incorrectly in cases where .* appeared inside
448        atomic brackets at the start of a pattern, or where there was a subsequent
449        *PRUNE or *SKIP.
450    
451    7.  Improved instruction cache flush for POWER/PowerPC.
452        Patch by Daniel Richard G.
453    
454    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
455        grep:
456    
457        (a) There is now no limit to the number of patterns to be matched.
458    
459        (b) An error is given if a pattern is too long.
460    
461        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
462            are now supported.
463    
464        (d) --exclude-from and --include-from (multiple use) have been added.
465    
466        (e) Exclusions and inclusions now apply to all files and directories, not
467            just to those obtained from scanning a directory recursively.
468    
469        (f) Multiple uses of -f and --file-list are now supported.
470    
471        (g) In a Windows environment, the default for -d has been changed from
472            "read" (the GNU grep default) to "skip", because otherwise the presence
473            of a directory in the file list provokes an error.
474    
475        (h) The documentation has been revised and clarified in places.
476    
477    9.  Improve the matching speed of capturing brackets.
478    
479    10. Changed the meaning of \X so that it now matches a Unicode extended
480        grapheme cluster.
481    
482    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
483        out POSIX threads when JIT support is configured.
484    
485    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
486    
487    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
488        structure could go wrong in environments where size_t is not the same size
489        as int.
490    
491    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
492        set.
493    
494    15. The EBCDIC support had decayed; later updates to the code had included
495        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
496        general tidy up of EBCDIC-related issues, and the documentation was also
497        not quite right. There is now a test that can be run on ASCII systems to
498        check some of the EBCDIC-related things (but is it not a full test).
499    
500    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
501        in a small tidy to the code.
502    
503    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
504    
505    18. If the --only-matching (-o) option in pcregrep is specified multiple
506        times, each one causes appropriate output. For example, -o1 -o2 outputs the
507        substrings matched by the 1st and 2nd capturing parentheses. A separating
508        string can be specified by --om-separator (default empty).
509    
510    19. Improving the first n character searches.
511    
512    20. Turn case lists for horizontal and vertical white space into macros so that
513        they are defined only once.
514    
515    21. This set of changes together give more compatible Unicode case-folding
516        behaviour for characters that have more than one other case when UCP
517        support is available.
518    
519        (a) The Unicode property table now has offsets into a new table of sets of
520            three or more characters that are case-equivalent. The MultiStage2.py
521            script that generates these tables (the pcre_ucd.c file) now scans
522            CaseFolding.txt instead of UnicodeData.txt for character case
523            information.
524    
525        (b) The code for adding characters or ranges of characters to a character
526            class has been abstracted into a generalized function that also handles
527            case-independence. In UTF-mode with UCP support, this uses the new data
528            to handle characters with more than one other case.
529    
530        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
531            whose other case is greater than 256 are now correctly matched
532            caselessly. Previously, the high codepoint matched the low one, but not
533            vice versa.
534    
535        (d) The processing of \h, \H, \v, and \ in character classes now makes use
536            of the new class addition function, using character lists defined as
537            macros alongside the case definitions of 20 above.
538    
539        (e) Caseless back references now work with characters that have more than
540            one other case.
541    
542        (f) General caseless matching of characters with more than one other case
543            is supported.
544    
545    22. Unicode character properties were updated from Unicode 6.2.0
546    
547    23. Improved CMake support under Windows. Patch by Daniel Richard G.
548    
549    24. Add support for 32-bit character strings, and UTF-32
550    
551    25. Major JIT compiler update (code refactoring and bugfixing).
552        Experimental Sparc 32 support is added.
553    
554    26. Applied a modified version of Daniel Richard G's patch to create
555        pcre.h.generic and config.h.generic by "make" instead of in the
556        PrepareRelease script.
557    
558    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
559        pcre_compile.c when checking for a zero character.
560    
561    28. Introducing a native interface for JIT. Through this interface, the compiled
562        machine code can be directly executed. The purpose of this interface is to
563        provide fast pattern matching, so several sanity checks are not performed.
564        However, feature tests are still performed. The new interface provides
565        1.4x speedup compared to the old one.
566    
567    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
568        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
569        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
570    
571    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
572        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
573        the "old" RFC 2279). Instead, it ended up passing a negative length to
574        pcre_exec().
575    
576    31. Add support for GCC's visibility feature to hide internal functions.
577    
578    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
579        "unknown -C option" after outputting 0 or 1.
580    
581    33. There is now support for generating a code coverage report for the test
582        suite in environments where gcc is the compiler and lcov is installed. This
583        is mainly for the benefit of the developers.
584    
585    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
586        unaddressable using valgrind annotations, allowing valgrind to detect
587        invalid memory accesses. This is mainly for the benefit of the developers.
588    
589    25. (*UTF) can now be used to start a pattern in any of the three libraries.
590    
591    26. Give configure error if --enable-cpp but no C++ compiler found.
592    
593    
594    Version 8.31 06-July-2012
595    -------------------------
596    
597    1.  Fixing a wrong JIT test case and some compiler warnings.
598    
599    2.  Removed a bashism from the RunTest script.
600    
601    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
602        to unsigned type, result still unsigned" that was given by an MS compiler
603        on encountering the code "-sizeof(xxx)".
604    
605    4.  Partial matching support is added to the JIT compiler.
606    
607    5.  Fixed several bugs concerned with partial matching of items that consist
608        of more than one character:
609    
610        (a) /^(..)\1/ did not partially match "aba" because checking references was
611            done on an "all or nothing" basis. This also applied to repeated
612            references.
613    
614        (b) \R did not give a hard partial match if \r was found at the end of the
615            subject.
616    
617        (c) \X did not give a hard partial match after matching one or more
618            characters at the end of the subject.
619    
620        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
621            a partial match for the string "\r".
622    
623        (e) When newline was set to CRLF, the metacharacter "." did not recognize
624            a partial match for a CR character at the end of the subject string.
625    
626    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
627        running pcretest, the text "(JIT)" added to the output whenever JIT is
628        actually used to run the match.
629    
630    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
631        or /S+[+] with a digit between 1 and 7.
632    
633    8.  OP_NOT now supports any UTF character not just single-byte ones.
634    
635    9.  (*MARK) control verb is now supported by the JIT compiler.
636    
637    10. The command "./RunTest list" lists the available tests without actually
638        running any of them. (Because I keep forgetting what they all are.)
639    
640    11. Add PCRE_INFO_MAXLOOKBEHIND.
641    
642    12. Applied a (slightly modified) user-supplied patch that improves performance
643        when the heap is used for recursion (compiled with --disable-stack-for-
644        recursion). Instead of malloc and free for each heap frame each time a
645        logical recursion happens, frames are retained on a chain and re-used where
646        possible. This sometimes gives as much as 30% improvement.
647    
648    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
649        call.
650    
651    14. As documented, (*COMMIT) is now confined to within a positive assertion.
652    
653    15. It is now possible to link pcretest with libedit as an alternative to
654        libreadline.
655    
656    16. (*COMMIT) control verb is now supported by the JIT compiler.
657    
658    17. The Unicode data tables have been updated to Unicode 6.1.0.
659    
660    18. Added --file-list option to pcregrep.
661    
662    19. Added binary file support to pcregrep, including the -a, --binary-files,
663        -I, and --text options.
664    
665    20. The madvise function is renamed for posix_madvise for QNX compatibility
666        reasons. Fixed by Giuseppe D'Angelo.
667    
668    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
669    
670    22. Changed the option for creating symbolic links for 16-bit man pages from
671        -s to -sf so that re-installing does not cause issues.
672    
673    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
674    
675    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
676        matches in certain environments (the workspace was not being correctly
677        retained). Also added to pcre_dfa_exec() a simple plausibility check on
678        some of the workspace data at the beginning of a restart.
679    
680    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
681        was not doing so when it should - probably a typo introduced by SVN 528
682        (change 8.10/14).
683    
684    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
685        \w+ when the character tables indicated that \x{c4} was a word character.
686        There were several related cases, all because the tests for doing a table
687        lookup were testing for characters less than 127 instead of 255.
688    
689    27. If a pattern contains capturing parentheses that are not used in a match,
690        their slots in the ovector are set to -1. For those that are higher than
691        any matched groups, this happens at the end of processing. In the case when
692        there were back references that the ovector was too small to contain
693        (causing temporary malloc'd memory to be used during matching), and the
694        highest capturing number was not used, memory off the end of the ovector
695        was incorrectly being set to -1. (It was using the size of the temporary
696        memory instead of the true size.)
697    
698    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
699        ovector size, it uses memory at the end of the block that it has got.
700    
701    29. Check for an overlong MARK name and give an error at compile time. The
702        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
703    
704    30. JIT compiler update.
705    
706    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
707        Rill for the patch.
708    
709    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
710    
711    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
712    
713    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
714        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
715        was enabled.
716    
717    35. Improve JIT code generation for greedy plus quantifier.
718    
719    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
720        1 to "aa" instead of to an empty string. The bug affected repeated groups
721        that could potentially match an empty string.
722    
723    37. Optimizing single character iterators in JIT.
724    
725    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
726        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
727        codepoints that are too big for the mode are faulted, and in a UTF mode,
728        disallowed codepoints are also faulted.
729    
730    39. If PCRE was compiled with UTF support, in three places in the DFA
731        matcher there was code that should only have been obeyed in UTF mode, but
732        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
733        processing when bytes with values greater than 127 were present. In 16-bit
734        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
735        both cases the values are those that cannot be the first data item in a UTF
736        character. The three items that might have provoked this were recursions,
737        possessively repeated groups, and atomic groups.
738    
739    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
740        and pcregrep, because some OS require shared objects to be explicitly
741        passed to ld, causing the link step to fail if they are not.
742    
743    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
744        mode, patterns that started with \h* or \R* might be incorrectly matched.
745    
746    
747    Version 8.30 04-February-2012
748    -----------------------------
749    
750    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
751        name is defined in ctype.h.
752    
753    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
754        only in quite long subpatterns.
755    
756    3.  Removed the function pcre_info(), which has been obsolete and deprecated
757        since it was replaced by pcre_fullinfo() in February 2000.
758    
759    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
760        match a (*MARK), and the match failed at the start of the subject, a
761        reference to memory before the start of the subject could occur. This bug
762        was introduced by fix 17 of release 8.21.
763    
764    5.  A reference to an unset group with zero minimum repetition was giving
765        totally wrong answers (in non-JavaScript-compatibility mode). For example,
766        /(another)?(\1?)test/ matched against "hello world test". This bug was
767        introduced in release 8.13.
768    
769    6.  Add support for 16-bit character strings (a large amount of work involving
770        many changes and refactorings).
771    
772    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
773        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
774        from a file.
775    
776    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
777        rounding is not applied in this particular case).
778    
779    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
780        if they appear, or are escaped, in patterns.
781    
782    10. Get rid of a number of -Wunused-but-set-variable warnings.
783    
784    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
785        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
786        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
787        also returns the mark "x". This bug applied to capturing parentheses,
788        non-capturing parentheses, and atomic parentheses. It also applied to some
789        assertions.
790    
791    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
792        information out of configure.ac instead of relying on pcre.h.generic, which
793        is not stored in the repository.
794    
795    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
796        -lreadline.
797    
798    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
799    
800    15. Applied Graycode's patch to put the top-level frame on the stack rather
801        than the heap when not using the stack for recursion. This gives a
802        performance improvement in many cases when recursion is not deep.
803    
804    16. Experimental code added to "pcretest -C" to output the stack frame size.
805    
806    
807    Version 8.21 12-Dec-2011
808    ------------------------
809    
810    1.  Updating the JIT compiler.
811    
812    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
813        are added as well.
814    
815    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
816        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
817        calling _pcre_jit_exec. Some extra comments are added.
818    
819    4.  (*MARK) settings inside atomic groups that do not contain any capturing
820        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
821        was introduced by change 18 for 8.20.
822    
823    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
824        ECMA-262 standard.
825    
826    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
827        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
828        This bug was probably introduced by change 9 of 8.13.
829    
830    7.  While fixing 6 above, I noticed that a number of other items were being
831        incorrectly rejected as "not fixed length". This arose partly because newer
832        opcodes had not been added to the fixed-length checking code. I have (a)
833        corrected the bug and added tests for these items, and (b) arranged for an
834        error to occur if an unknown opcode is encountered while checking for fixed
835        length instead of just assuming "not fixed length". The items that were
836        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
837        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
838        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
839    
840    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
841        being incorrectly compiled and would have given unpredicatble results.
842    
843    9.  A possessively repeated subpattern with minimum repeat count greater than
844        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
845        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
846        the first (A) could occur when it should not.
847    
848    10. Add a cast and remove a redundant test from the code.
849    
850    11. JIT should use pcre_malloc/pcre_free for allocation.
851    
852    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
853        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
854        is anything other than /usr, -L is still shown).
855    
856    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
857    
858    14. Perl does not support \N without a following name in a [] class; PCRE now
859        also gives an error.
860    
861    15. If a forward reference was repeated with an upper limit of around 2000,
862        it caused the error "internal error: overran compiling workspace". The
863        maximum number of forward references (including repeats) was limited by the
864        internal workspace, and dependent on the LINK_SIZE. The code has been
865        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
866        the default depends on LINK_SIZE. There is a new upper limit (for safety)
867        of around 200,000 forward references. While doing this, I also speeded up
868        the filling in of repeated forward references.
869    
870    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
871        incorrectly expecting the subject to contain another "a" after the start.
872    
873    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
874        in the match, the SKIP should be ignored. This was not happening; instead
875        the SKIP was being treated as NOMATCH. For patterns such as
876        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
877        tested.
878    
879    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
880        now much more compatible with Perl, in particular in cases where the result
881        is a non-match for a non-anchored pattern. For example, if
882        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
883        "m", where previously it did not return a name. A side effect of this
884        change is that for partial matches, the last encountered mark name is
885        returned, as for non matches. A number of tests that were previously not
886        Perl-compatible have been moved into the Perl-compatible test files. The
887        refactoring has had the pleasing side effect of removing one argument from
888        the match() function, thus reducing its stack requirements.
889    
890    19. If the /S+ option was used in pcretest to study a pattern using JIT,
891        subsequent uses of /S (without +) incorrectly behaved like /S+.
892    
893    21. Retrieve executable code size support for the JIT compiler and fixing
894        some warnings.
895    
896    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
897        not work when the shorter character appeared right at the end of the
898        subject string.
899    
900    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
901        systems.
902    
903    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
904        output it when the /M option is used in pcretest.
905    
906    25. The CheckMan script was not being included in the distribution. Also, added
907        an explicit "perl" to run Perl scripts from the PrepareRelease script
908        because this is reportedly needed in Windows.
909    
910    26. If study data was being save in a file and studying had not found a set of
911        "starts with" bytes for the pattern, the data written to the file (though
912        never used) was taken from uninitialized memory and so caused valgrind to
913        complain.
914    
915    27. Updated RunTest.bat as provided by Sheri Pierce.
916    
917    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
918    
919    29. Computation of memory usage for the table of capturing group names was
920        giving an unnecessarily large value.
921    
922    
923    Version 8.20 21-Oct-2011
924    ------------------------
925    
926    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
927        a POSIX class. After further experiments with Perl, which convinced me that
928        Perl has bugs and confusions, a closing square bracket is no longer allowed
929        in a POSIX name. This bug also affected patterns with classes that started
930        with full stops.
931    
932    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
933        captured substring, but while checking the failing first alternative,
934        substring 1 is temporarily captured. If the output vector supplied to
935        pcre_exec() was not big enough for this capture, the yield of the function
936        was still zero ("insufficient space for captured substrings"). This cannot
937        be totally fixed without adding another stack variable, which seems a lot
938        of expense for a edge case. However, I have improved the situation in cases
939        such as /(a)(b)x|abc/ matched against "abc", where the return code
940        indicates that fewer than the maximum number of slots in the ovector have
941        been set.
942    
943    3.  Related to (2) above: when there are more back references in a pattern than
944        slots in the output vector, pcre_exec() uses temporary memory during
945        matching, and copies in the captures as far as possible afterwards. It was
946        using the entire output vector, but this conflicts with the specification
947        that only 2/3 is used for passing back captured substrings. Now it uses
948        only the first 2/3, for compatibility. This is, of course, another edge
949        case.
950    
951    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
952        main code base, and can be used by building with --enable-jit. When this is
953        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
954        runtime --no-jit option is given.
955    
956    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
957        ovector, the return from the function was zero, implying that there were
958        other matches that did not fit. The correct "exactly full" value is now
959        returned.
960    
961    6.  If a subpattern that was called recursively or as a subroutine contained
962        (*PRUNE) or any other control that caused it to give a non-standard return,
963        invalid errors such as "Error -26 (nested recursion at the same subject
964        position)" or even infinite loops could occur.
965    
966    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
967        computing the minimum length on reaching *ACCEPT, and so ended up with the
968        wrong value of 1 rather than 0. Further investigation indicates that
969        computing a minimum subject length in the presence of *ACCEPT is difficult
970        (think back references, subroutine calls), and so I have changed the code
971        so that no minimum is registered for a pattern that contains *ACCEPT.
972    
973    8.  If (*THEN) was present in the first (true) branch of a conditional group,
974        it was not handled as intended. [But see 16 below.]
975    
976    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
977        Sheri Pierce.
978    
979    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
980        the first byte in a match must be "a".
981    
982    11. Change 17 for 8.13 increased the recursion depth for patterns like
983        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
984        pattern contains any instances of (*THEN). If it does not, the old
985        optimizations are restored. It would be nice to do this on a per-group
986        basis, but at the moment that is not feasible.
987    
988    12. In some environments, the output of pcretest -C is CRLF terminated. This
989        broke RunTest's code that checks for the link size. A single white space
990        character after the value is now allowed for.
991    
992    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
993        For "fr", it uses the Windows-specific input and output files.
994    
995    14. If (*THEN) appeared in a group that was called recursively or as a
996        subroutine, it did not work as intended. [But see next item.]
997    
998    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
999        pattern fragments (but not containing any | characters). If A and B are
1000        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1001        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1002        D. In other words, Perl considers parentheses that do not contain any |
1003        characters to be part of a surrounding alternative, whereas PCRE was
1004        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1005        differently. PCRE now behaves in the same way as Perl, except in the case
1006        of subroutine/recursion calls such as (?1) which have in any case always
1007        been different (but PCRE had them first :-).
1008    
1009    16. Related to 15 above: Perl does not treat the | in a conditional group as
1010        creating alternatives. Such a group is treated in the same way as an
1011        ordinary group without any | characters when processing (*THEN). PCRE has
1012        been changed to match Perl's behaviour.
1013    
1014    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1015        RunGrepTest script failed.
1016    
1017    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1018        inevitable for groups that contain captures, but it can lead to a lot of
1019        stack use in large patterns. The old behaviour has been restored for atomic
1020        groups that do not contain any capturing parentheses.
1021    
1022    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1023        suppress the check for a minimum subject length at run time. (If it was
1024        given to pcre_exec() or pcre_dfa_exec() it did work.)
1025    
1026    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1027        fail to work when decoding hex characters in data strings in EBCDIC
1028        environments.
1029    
1030    21. It appears that in at least one Mac OS environment, the isxdigit() function
1031        is implemented as a macro that evaluates to its argument more than once,
1032        contravening the C 90 Standard (I haven't checked a later standard). There
1033        was an instance in pcretest which caused it to go wrong when processing
1034        \x{...} escapes in subject strings. The has been rewritten to avoid using
1035        things like p++ in the argument of isxdigit().
1036    
1037    
1038    Version 8.13 16-Aug-2011
1039    ------------------------
1040    
1041    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1042    
1043    2.  Two minor typos in pcre_internal.h have been fixed.
1044    
1045    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1046        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1047        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1048    
1049    4.  There were a number of related bugs in the code for matching backrefences
1050        caselessly in UTF-8 mode when codes for the characters concerned were
1051        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1052        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1053        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1054        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1055        2-byte code at the end of the subject (it thought there wasn't enough data
1056        left).
1057    
1058    5.  Comprehensive information about what went wrong is now returned by
1059        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1060        as the output vector has at least 2 elements. The offset of the start of
1061        the failing character and a reason code are placed in the vector.
1062    
1063    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1064        now returned is for the first byte of the failing character, instead of the
1065        last byte inspected. This is an incompatible change, but I hope it is small
1066        enough not to be a problem. It makes the returned offset consistent with
1067        pcre_exec() and pcre_dfa_exec().
1068    
1069    7.  pcretest now gives a text phrase as well as the error number when
1070        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1071        failure, the offset and reason code are output.
1072    
1073    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1074        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1075        back over a single character (\n). This seems wrong (because it treated the
1076        two characters as a single entity when going forwards), conflicts with the
1077        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1078        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1079        has been changed.
1080    
1081    9.  Some internal refactoring has changed the processing so that the handling
1082        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1083        time (the PCRE_DOTALL option was changed this way some time ago: version
1084        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1085        which was always a bit of a fudge. It also means that there is one less
1086        argument for the match() function, which reduces its stack requirements
1087        slightly. This change also fixes an incompatibility with Perl: the pattern
1088        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1089    
1090    10. More internal refactoring has drastically reduced the number of recursive
1091        calls to match() for possessively repeated groups such as (abc)++ when
1092        using pcre_exec().
1093    
1094    11. While implementing 10, a number of bugs in the handling of groups were
1095        discovered and fixed:
1096    
1097        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1098        (a|)*(?1) gave a compile-time internal error.
1099        ((a|)+)+  did not notice that the outer group could match an empty string.
1100        (^a|^)+   was not marked as anchored.
1101        (.*a|.*)+ was not marked as matching at start or after a newline.
1102    
1103    12. Yet more internal refactoring has removed another argument from the match()
1104        function. Special calls to this function are now indicated by setting a
1105        value in a variable in the "match data" data block.
1106    
1107    13. Be more explicit in pcre_study() instead of relying on "default" for
1108        opcodes that mean there is no starting character; this means that when new
1109        ones are added and accidentally left out of pcre_study(), testing should
1110        pick them up.
1111    
1112    14. The -s option of pcretest has been documented for ages as being an old
1113        synonym of -m (show memory usage). I have changed it to mean "force study
1114        for every regex", that is, assume /S for every regex. This is similar to -i
1115        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1116        using it. It makes it easier to run collections of tests with and without
1117        study enabled, and thereby test pcre_study() more easily. All the standard
1118        tests are now run with and without -s (but some patterns can be marked as
1119        "never study" - see 20 below).
1120    
1121    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1122        restoration of the capturing data to the outer values was not happening
1123        correctly.
1124    
1125    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1126        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1127        pattern had matched an empty string, and so incorrectly returned a no
1128        match.
1129    
1130    17. There was optimizing code for the last branch of non-capturing parentheses,
1131        and also for the obeyed branch of a conditional subexpression, which used
1132        tail recursion to cut down on stack usage. Unfortunately, now that there is
1133        the possibility of (*THEN) occurring in these branches, tail recursion is
1134        no longer possible because the return has to be checked for (*THEN). These
1135        two optimizations have therefore been removed. [But see 8.20/11 above.]
1136    
1137    18. If a pattern containing \R was studied, it was assumed that \R always
1138        matched two bytes, thus causing the minimum subject length to be
1139        incorrectly computed because \R can also match just one byte.
1140    
1141    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1142        was incorrectly computed.
1143    
1144    20. If /S is present twice on a test pattern in pcretest input, it now
1145        *disables* studying, thereby overriding the use of -s on the command line
1146        (see 14 above). This is necessary for one or two tests to keep the output
1147        identical in both cases.
1148    
1149    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1150        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1151    
1152    22. When an atomic group that contained a capturing parenthesis was
1153        successfully matched, but the branch in which it appeared failed, the
1154        capturing was not being forgotten if a higher numbered group was later
1155        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1156        group 1 to "a", when in fact it should be unset. This applied to multi-
1157        branched capturing and non-capturing groups, repeated or not, and also to
1158        positive assertions (capturing in negative assertions does not happen
1159        in PCRE) and also to nested atomic groups.
1160    
1161    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1162        subject after a captured substring, to make it easier to tell which of a
1163        number of identical substrings has been captured.
1164    
1165    24. The way atomic groups are processed by pcre_exec() has been changed so that
1166        if they are repeated, backtracking one repetition now resets captured
1167        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1168        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1169        "aaa". Previously, it would have been "a". As part of this code
1170        refactoring, the way recursive calls are handled has also been changed.
1171    
1172    25. If an assertion condition captured any substrings, they were not passed
1173        back unless some other capturing happened later. For example, if
1174        (?(?=(a))a) was matched against "a", no capturing was returned.
1175    
1176    26. When studying a pattern that contained subroutine calls or assertions,
1177        the code for finding the minimum length of a possible match was handling
1178        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1179        group 1 called group 2 while simultaneously a separate group 2 called group
1180        1). A stack overflow occurred in this case. I have fixed this by limiting
1181        the recursion depth to 10.
1182    
1183    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1184        Fortmann. This supports explicit test numbers on the command line, and has
1185        argument validation and error reporting.
1186    
1187    28. An instance of \X with an unlimited repeat could fail if at any point the
1188        first character it looked at was a mark character.
1189    
1190    29. Some minor code refactoring concerning Unicode properties and scripts
1191        should reduce the stack requirement of match() slightly.
1192    
1193    30. Added the '=' option to pcretest to check the setting of unused capturing
1194        slots at the end of the pattern, which are documented as being -1, but are
1195        not included in the return count.
1196    
1197    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1198        compiled something random. Now it gives a compile-time error (as does
1199        Perl).
1200    
1201    32. A *MARK encountered during the processing of a positive assertion is now
1202        recorded and passed back (compatible with Perl).
1203    
1204    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1205        had alternative anchored branches, the search for a second match in a line
1206        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1207        matched the line "0102" twice. The same bug affected patterns that started
1208        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1209        twice.
1210    
1211    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1212        does, and because of capturing effects, quantifying parenthesized
1213        assertions may at times be useful. Quantifiers are now allowed for
1214        parenthesized assertions.
1215    
1216    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1217    
1218    36. \g was being checked for fancy things in a character class, when it should
1219        just be a literal "g".
1220    
1221    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1222        appearance of a nested POSIX class supersedes an apparent external class.
1223        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1224        unescaped square brackets may also appear as part of class names. For
1225        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1226        more like Perl. (But see 8.20/1 above.)
1227    
1228    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1229        was because it thought it was \N{name}, which is not supported).
1230    
1231    39. Add minix to OS list not supporting the -S option in pcretest.
1232    
1233    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1234        cannot analyze patterns in sufficient detail to catch mutual recursions
1235        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1236        subgroup is called recursively as a subpattern for a second time at the
1237        same position in the subject string. In previous releases this might have
1238        been caught by the recursion limit, or it might have run out of stack.
1239    
1240    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1241        happen only once. PCRE was, however incorrectly giving a compile time error
1242        "recursive call could loop indefinitely" because it cannot analyze the
1243        pattern in sufficient detail. The compile time test no longer happens when
1244        PCRE is compiling a conditional subpattern, but actual runaway loops are
1245        now caught at runtime (see 40 above).
1246    
1247    42. It seems that Perl allows any characters other than a closing parenthesis
1248        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1249        has been changed to be the same.
1250    
1251    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1252        as not to get warnings when autogen.sh is called. Also changed
1253        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1254    
1255    44. To help people who use pcregrep to scan files containing exceedingly long
1256        lines, the following changes have been made:
1257    
1258        (a) The default value of the buffer size parameter has been increased from
1259            8K to 20K. (The actual buffer used is three times this size.)
1260    
1261        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1262            PCRE is built.
1263    
1264        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1265            to be set at run time.
1266    
1267        (d) Numerical values in pcregrep options can be followed by K or M, for
1268            example --buffer-size=50K.
1269    
1270        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1271            given and the return code is set to 2.
1272    
1273    45. Add a pointer to the latest mark to the callout data block.
1274    
1275    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1276        partial match of an empty string instead of no match. This was specific to
1277        the use of ".".
1278    
1279    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1280        complete match instead of a partial match. This bug was dependent on both
1281        the PCRE_UTF8 and PCRE_DOTALL options being set.
1282    
1283    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1284        starting byte set, because \b was not being ignored.
1285    
1286    
1287    Version 8.12 15-Jan-2011
1288    ------------------------
1289    
1290    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1291        checks for such things as part of the documentation building process.
1292    
1293    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1294        --match-limit and --recursion-limit options (added for 8.11). In
1295        particular, this made one of the standard tests fail. (The integer value
1296        went into the wrong half of a long int.)
1297    
1298    3.  If the --colour option was given to pcregrep with -v (invert match), it
1299        did strange things, either producing crazy output, or crashing. It should,
1300        of course, ignore a request for colour when reporting lines that do not
1301        match.
1302    
1303    4.  Another pcregrep bug caused similar problems if --colour was specified with
1304        -M (multiline) and the pattern match finished with a line ending.
1305    
1306    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1307        matched in multiline mode, the following line was shown as part of the
1308        match. This seems wrong, so I have changed it.
1309    
1310    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1311        the check for further matches in the same line (so they could be coloured)
1312        to overrun the end of the current line. If another match was found, it was
1313        incorrectly shown (and then shown again when found in the next line).
1314    
1315    7.  If pcregrep was compiled under Windows, there was a reference to the
1316        function pcregrep_exit() before it was defined. I am assuming this was
1317        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1318        reported by a user. I've moved the definition above the reference.
1319    
1320    
1321    Version 8.11 10-Dec-2010
1322    ------------------------
1323    
1324    1.  (*THEN) was not working properly if there were untried alternatives prior
1325        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1326        backtracked to try for "b" instead of moving to the next alternative branch
1327        at the same level (in this case, to look for "c"). The Perl documentation
1328        is clear that when (*THEN) is backtracked onto, it goes to the "next
1329        alternative in the innermost enclosing group".
1330    
1331    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1332        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1333        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1334        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1335        (*THEN).
1336    
1337    3.  If \s appeared in a character class, it removed the VT character from
1338        the class, even if it had been included by some previous item, for example
1339        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1340        of \s, but is part of the POSIX "space" class.)
1341    
1342    4.  A partial match never returns an empty string (because you can always
1343        match an empty string at the end of the subject); however the checking for
1344        an empty string was starting at the "start of match" point. This has been
1345        changed to the "earliest inspected character" point, because the returned
1346        data for a partial match starts at this character. This means that, for
1347        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1348        (previously it gave "no match").
1349    
1350    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1351        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1352        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1353        has an implication that the given string is incomplete (because a partial
1354        match is preferred over a full match). For this reason, these items now
1355        give a partial match in this situation. [Aside: previously, the one case
1356        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1357        match rather than a full match, which was wrong by the old rules, but is
1358        now correct.]
1359    
1360    6.  There was a bug in the handling of #-introduced comments, recognized when
1361        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1362        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1363        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1364        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1365        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1366        places in pcre_compile().
1367    
1368    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1369        comments when looking ahead for named forward references to subpatterns,
1370        the only newline sequence it recognized was NL. It now handles newlines
1371        according to the set newline convention.
1372    
1373    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1374        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1375        cater for a lack of strerror(). These oversights have been fixed.
1376    
1377    9.  Added --match-limit and --recursion-limit to pcregrep.
1378    
1379    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1380    
1381    11. When the -o option was used, pcregrep was setting a return code of 1, even
1382        when matches were found, and --line-buffered was not being honoured.
1383    
1384    12. Added an optional parentheses number to the -o and --only-matching options
1385        of pcregrep.
1386    
1387    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1388        can match an empty string. The code to do it in pcretest and pcredemo
1389        needed fixing:
1390    
1391        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1392            only one byte after an empty string match just before CRLF (this case
1393            just got forgotten; "any" and "anycrlf" were OK).
1394    
1395        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1396            mode when an empty string match preceded an ASCII character followed by
1397            a non-ASCII character. (The code for advancing by one character rather
1398            than one byte was nonsense.)
1399    
1400        (c) The pcredemo.c sample program did not have any code at all to handle
1401            the cases when CRLF is a valid newline sequence.
1402    
1403    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1404        as a starting offset was within the subject string. There is now a new
1405        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1406        negative or greater than the length of the string. In order to test this,
1407        pcretest is extended to allow the setting of negative starting offsets.
1408    
1409    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1410        starting offset points to the beginning of a UTF-8 character was
1411        unnecessarily clumsy. I tidied it up.
1412    
1413    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1414        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1415    
1416    17. Nobody had reported that the --include_dir option, which was added in
1417        release 7.7 should have been called --include-dir (hyphen, not underscore)
1418        for compatibility with GNU grep. I have changed it to --include-dir, but
1419        left --include_dir as an undocumented synonym, and the same for
1420        --exclude-dir, though that is not available in GNU grep, at least as of
1421        release 2.5.4.
1422    
1423    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1424        characters from a string of bytes) have been redefined so as not to use
1425        loops, in order to improve performance in some environments. At the same
1426        time, I abstracted some of the common code into auxiliary macros to save
1427        repetition (this should not affect the compiled code).
1428    
1429    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1430        compile-time error is now given if \c is not followed by an ASCII
1431        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1432        different, and any byte value is allowed.)
1433    
1434    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1435        START_OPTIMIZE option, which is now allowed at compile time - but just
1436        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1437        to pcregrep and other applications that have no direct access to PCRE
1438        options. The new /Y option in pcretest sets this option when calling
1439        pcre_compile().
1440    
1441    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1442        back references. Groups containing recursive back references were forced to
1443        be atomic by that change, but in the case of named groups, the amount of
1444        memory required was incorrectly computed, leading to "Failed: internal
1445        error: code overflow". This has been fixed.
1446    
1447    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1448        pcretest.c, to avoid build problems in some Borland environments.
1449    
1450    
1451    Version 8.10 25-Jun-2010
1452    ------------------------
1453    
1454    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1455        THEN.
1456    
1457    2.  (*ACCEPT) was not working when inside an atomic group.
1458    
1459    3.  Inside a character class, \B is treated as a literal by default, but
1460        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1461        causes the error). The code is unchanged, but I tidied the documentation.
1462    
1463    4.  Inside a character class, PCRE always treated \R and \X as literals,
1464        whereas Perl faults them if its -w option is set. I have changed PCRE so
1465        that it faults them when PCRE_EXTRA is set.
1466    
1467    5.  Added support for \N, which always matches any character other than
1468        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1469    
1470    6.  When compiling pcregrep with newer versions of gcc which may have
1471        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1472        declared with attribute warn_unused_result" were given. Just casting the
1473        result to (void) does not stop the warnings; a more elaborate fudge is
1474        needed. I've used a macro to implement this.
1475    
1476    7.  Minor change to pcretest.c to avoid a compiler warning.
1477    
1478    8.  Added four artifical Unicode properties to help with an option to make
1479        \s etc use properties (see next item). The new properties are: Xan
1480        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1481    
1482    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1483        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1484        this option. Modified pcretest to add /W to test this facility. Added
1485        REG_UCP to make it available via the POSIX interface.
1486    
1487    10. Added --line-buffered to pcregrep.
1488    
1489    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1490        studied, and the match started with a letter with a code point greater than
1491        127 whose first byte was different to the first byte of the other case of
1492        the letter, the other case of this starting letter was not recognized
1493        (#976).
1494    
1495    12. If a pattern that was studied started with a repeated Unicode property
1496        test, for example, \p{Nd}+, there was the theoretical possibility of
1497        setting up an incorrect bitmap of starting bytes, but fortunately it could
1498        not have actually happened in practice until change 8 above was made (it
1499        added property types that matched character-matching opcodes).
1500    
1501    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1502        possible starting bytes for non-anchored patterns.
1503    
1504    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1505        \R, and also a number of cases that involve Unicode properties, both
1506        explicit and implicit when PCRE_UCP is set.
1507    
1508    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1509        input, it could crash or give wrong results if characters with values
1510        greater than 0xc0 were present in the subject string. (Detail: it assumed
1511        UTF-8 input when processing these items.)
1512    
1513    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1514        size_t is 64-bit (#991).
1515    
1516    17. Added a check for running out of memory when PCRE is compiled with
1517        --disable-stack-for-recursion (#990).
1518    
1519    18. If the last data line in a file for pcretest does not have a newline on
1520        the end, a newline was missing in the output.
1521    
1522    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1523        less than 128) in its various bitmaps. However, there is a facility for
1524        generating tables according to the current locale when PCRE is compiled. It
1525        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1526        characters, are recognized by isspace() and therefore were getting set in
1527        these tables, and indeed these tables seem to approximate to ISO 8859. This
1528        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1529        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1530        which of course cannot start UTF-8 characters. I have changed the code so
1531        that only real ASCII characters (less than 128) and the correct starting
1532        bytes for UTF-8 encodings are set for characters greater than 127 when in
1533        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1534        altogether.)
1535    
1536    20. Added the /T option to pcretest so as to be able to run tests with non-
1537        standard character tables, thus making it possible to include the tests
1538        used for 19 above in the standard set of tests.
1539    
1540    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1541        reference to a subpattern the other side of a comment that contains an
1542        opening parenthesis caused either an internal compiling error, or a
1543        reference to the wrong subpattern.
1544    
1545    
1546    Version 8.02 19-Mar-2010
1547    ------------------------
1548    
1549    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1550    
1551    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1552        configured.
1553    
1554    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1555        original author of that file, following a query about its status.
1556    
1557    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1558        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1559    
1560    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1561        quantifier applied to a forward-referencing subroutine call, could compile
1562        incorrect code or give the error "internal error: previously-checked
1563        referenced subpattern not found".
1564    
1565    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1566        variables to point to external functions. For these systems, therefore,
1567        pcre_malloc etc. are now initialized to local functions that call the
1568        relevant global functions.
1569    
1570    7.  There were two entries missing in the vectors called coptable and poptable
1571        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1572        I've fixed the data, and added a kludgy way of testing at compile time that
1573        the lengths are correct (equal to the number of opcodes).
1574    
1575    8.  Following on from 7, I added a similar kludge to check the length of the
1576        eint vector in pcreposix.c.
1577    
1578    9.  Error texts for pcre_compile() are held as one long string to avoid too
1579        much relocation at load time. To find a text, the string is searched,
1580        counting zeros. There was no check for running off the end of the string,
1581        which could happen if a new error number was added without updating the
1582        string.
1583    
1584    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1585    
1586    11. \K was not working if it appeared in an atomic group or in a group that
1587        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1588        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1589        the assertion is positive, but not if it is negative.
1590    
1591    12. Change 11 fortuitously reduced the size of the stack frame used in the
1592        "match()" function of pcre_exec.c by one pointer. Forthcoming
1593        implementation of support for (*MARK) will need an extra pointer on the
1594        stack; I have reserved it now, so that the stack frame size does not
1595        decrease.
1596    
1597    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1598        item in branch that calls a recursion is a subroutine call - as in the
1599        second branch in the above example - was incorrectly given the compile-
1600        time error "recursive call could loop indefinitely" because pcre_compile()
1601        was not correctly checking the subroutine for matching a non-empty string.
1602    
1603    14. The checks for overrunning compiling workspace could trigger after an
1604        overrun had occurred. This is a "should never occur" error, but it can be
1605        triggered by pathological patterns such as hundreds of nested parentheses.
1606        The checks now trigger 100 bytes before the end of the workspace.
1607    
1608    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1609    
1610    
1611    Version 8.01 19-Jan-2010
1612    ------------------------
1613    
1614    1.  If a pattern contained a conditional subpattern with only one branch (in
1615        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1616        computed the wrong minimum data length (which is of course zero for such
1617        subpatterns). This could cause incorrect "no match" results.
1618    
1619    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1620        the pattern is reset in the first branch, pcre_compile() failed with
1621        "internal error: code overflow at offset...". This happened only when
1622        the reset was to the original external option setting. (An optimization
1623        abstracts leading options settings into an external setting, which was the
1624        cause of this.)
1625    
1626    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1627        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1628        assertion pattern did not match (meaning that the assertion was true), it
1629        was incorrectly treated as false if the SKIP had been reached during the
1630        matching. This also applied to assertions used as conditions.
1631    
1632    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1633        assertion subpattern, including such a pattern used as a condition,
1634        unpredictable results occurred, instead of the error return
1635        PCRE_ERROR_DFA_UITEM.
1636    
1637    5.  The C++ GlobalReplace function was not working like Perl for the special
1638        situation when an empty string is matched. It now does the fancy magic
1639        stuff that is necessary.
1640    
1641    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1642        removed. (These were left over from very, very early versions of PCRE.)
1643    
1644    7.  Some cosmetic changes to the code to make life easier when compiling it
1645        as part of something else:
1646    
1647        (a) Change DEBUG to PCRE_DEBUG.
1648    
1649        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1650            called "current" as "current_branch", to prevent a collision with the
1651            Linux macro when compiled as a kernel module.
1652    
1653        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1654            prevent a collision with the Linux macro when compiled as a kernel
1655            module.
1656    
1657    8.  In pcre_compile() there are some checks for integer overflows that used to
1658        cast potentially large values to (double). This has been changed to that
1659        when building, a check for int64_t is made, and if it is found, it is used
1660        instead, thus avoiding the use of floating point arithmetic. (There is no
1661        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1662        double.
1663    
1664    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1665        2005 (difference between two addresses compared to an unsigned value).
1666    
1667    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1668        custom one, because of the following reported problem in Windows:
1669    
1670          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1671              under Win32.
1672          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1673              therefore missing the function definition.
1674          - The compiler thus generates a "C" signature for the test function.
1675          - The linker fails to find the "C" function.
1676          - PCRE fails to configure if asked to do so against libbz2.
1677    
1678    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1679        messages were output:
1680    
1681          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1682          rerunning libtoolize, to keep the correct libtool macros in-tree.
1683          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1684    
1685        I have done both of these things.
1686    
1687    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1688        most of the time, it *can* run out if it is given a pattern that contains a
1689        runaway infinite recursion. I updated the discussion in the pcrestack man
1690        page.
1691    
1692    13. Now that we have gone to the x.xx style of version numbers, the minor
1693        version may start with zero. Using 08 or 09 is a bad idea because users
1694        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1695        interpreted as invalid octal numbers. I've updated the previous comment in
1696        configure.ac, and also added a check that gives an error if 08 or 09 are
1697        used.
1698    
1699    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1700        causing partial matching to fail when the end of the subject matched \W
1701        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1702    
1703    15. There were some discrepancies between the declarations in pcre_internal.h
1704        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1705        their definitions. The declarations used "const uschar *" and the
1706        definitions used USPTR. Even though USPTR is normally defined as "const
1707        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1708        reported that: "This difference in casting confuses some C++ compilers, for
1709        example, SunCC recognizes above declarations as different functions and
1710        generates broken code for hbpcre." I have changed the declarations to use
1711        USPTR.
1712    
1713    16. GNU libtool is named differently on some systems. The autogen.sh script now
1714        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1715        (FreeBSD).
1716    
1717    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1718        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1719        comment: "Figure out how to create a longlong from a string: strtoll and
1720        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1721        instance, but it only takes 2 args instead of 3!"
1722    
1723    18. A subtle bug concerned with back references has been fixed by a change of
1724        specification, with a corresponding code fix. A pattern such as
1725        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1726        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1727        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1728        same bug. Such groups have to be quantified to be useful, or contained
1729        inside another quantified group. (If there's no repetition, the reference
1730        can never match.) The problem arises because, having left the group and
1731        moved on to the rest of the pattern, a later failure that backtracks into
1732        the group uses the captured value from the final iteration of the group
1733        rather than the correct earlier one. I have fixed this in PCRE by forcing
1734        any group that contains a reference to itself to be an atomic group; that
1735        is, there cannot be any backtracking into it once it has completed. This is
1736        similar to recursive and subroutine calls.
1737    
1738    
1739    Version 8.00 19-Oct-09
1740    ----------------------
1741    
1742    1.  The table for translating pcre_compile() error codes into POSIX error codes
1743        was out-of-date, and there was no check on the pcre_compile() error code
1744        being within the table. This could lead to an OK return being given in
1745        error.
1746    
1747    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1748        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1749        in a Windows environment.
1750    
1751    3.  The pcregrep --count option prints the count for each file even when it is
1752        zero, as does GNU grep. However, pcregrep was also printing all files when
1753        --files-with-matches was added. Now, when both options are given, it prints
1754        counts only for those files that have at least one match. (GNU grep just
1755        prints the file name in this circumstance, but including the count seems
1756        more useful - otherwise, why use --count?) Also ensured that the
1757        combination -clh just lists non-zero counts, with no names.
1758    
1759    4.  The long form of the pcregrep -F option was incorrectly implemented as
1760        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1761        but it seems right to fix it, and I didn't think it was worth preserving
1762        the old behaviour.
1763    
1764    5.  The command line items --regex=pattern and --regexp=pattern were not
1765        recognized by pcregrep, which required --regex pattern or --regexp pattern
1766        (with a space rather than an '='). The man page documented the '=' forms,
1767        which are compatible with GNU grep; these now work.
1768    
1769    6.  No libpcreposix.pc file was created for pkg-config; there was just
1770        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1771    
1772    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1773        when UCP support is not needed, by modifying the Python script that
1774        generates it from Unicode data files. This should not matter if the module
1775        is correctly used as a library, but I received one complaint about 50K of
1776        unwanted data. My guess is that the person linked everything into his
1777        program rather than using a library. Anyway, it does no harm.
1778    
1779    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1780        was a minimum greater than 1 for a wide character in a possessive
1781        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1782        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1783        character. Chaos in the form of incorrect output or a compiling loop could
1784        result.
1785    
1786    9.  The restrictions on what a pattern can contain when partial matching is
1787        requested for pcre_exec() have been removed. All patterns can now be
1788        partially matched by this function. In addition, if there are at least two
1789        slots in the offset vector, the offset of the earliest inspected character
1790        for the match and the offset of the end of the subject are set in them when
1791        PCRE_ERROR_PARTIAL is returned.
1792    
1793    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1794        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1795        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1796        and may be more useful for multi-segment matching.
1797    
1798    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1799        used to be given if ever the end of the subject was reached; now it is
1800        given only if matching could not proceed because another character was
1801        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1802        string "Z", which now yields "no match" instead of "partial match". In the
1803        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1804        final character ended with (*FAIL).
1805    
1806    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1807        if the pattern had a "must contain" character that was already found in the
1808        earlier partial match, unless partial matching was again requested. For
1809        example, with the pattern /dog.(body)?/, the "must contain" character is
1810        "g". If the first part-match was for the string "dog", restarting with
1811        "sbody" failed. This bug has been fixed.
1812    
1813    13. The string returned by pcre_dfa_exec() after a partial match has been
1814        changed so that it starts at the first inspected character rather than the
1815        first character of the match. This makes a difference only if the pattern
1816        starts with a lookbehind assertion or \b or \B (\K is not supported by
1817        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1818        matching functions compatible, and I think it's the right thing to do.
1819    
1820    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1821        so that the demonstration program is easily available in environments where
1822        PCRE has not been installed from source.
1823    
1824    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1825        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1826        library.
1827    
1828    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1829        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1830        is not the first non-POSIX option to be added. Clearly some people find
1831        these options useful.
1832    
1833    17. If a caller to the POSIX matching function regexec() passes a non-zero
1834        value for nmatch with a NULL value for pmatch, the value of
1835        nmatch is forced to zero.
1836    
1837    18. RunGrepTest did not have a test for the availability of the -u option of
1838        the diff command, as RunTest does. It now checks in the same way as
1839        RunTest, and also checks for the -b option.
1840    
1841    19. If an odd number of negated classes containing just a single character
1842        interposed, within parentheses, between a forward reference to a named
1843        subpattern and the definition of the subpattern, compilation crashed with
1844        an internal error, complaining that it could not find the referenced
1845        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1846        [The bug was that it was starting one character too far in when skipping
1847        over the character class, thus treating the ] as data rather than
1848        terminating the class. This meant it could skip too much.]
1849    
1850    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1851        /g option in pcretest when the pattern contains \K, which makes it possible
1852        to have an empty string match not at the start, even when the pattern is
1853        anchored. Updated pcretest and pcredemo to use this option.
1854    
1855    21. If the maximum number of capturing subpatterns in a recursion was greater
1856        than the maximum at the outer level, the higher number was returned, but
1857        with unset values at the outer level. The correct (outer level) value is
1858        now given.
1859    
1860    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1861        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1862        make it do so. The string so far is captured, making this feature
1863        compatible with Perl.
1864    
1865    23. The tests have been re-organized, adding tests 11 and 12, to make it
1866        possible to check the Perl 5.10 features against Perl 5.10.
1867    
1868    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1869        pattern matches a fixed length string. PCRE did not allow this; now it
1870        does. Neither allows recursion.
1871    
1872    25. I finally figured out how to implement a request to provide the minimum
1873        length of subject string that was needed in order to match a given pattern.
1874        (It was back references and recursion that I had previously got hung up
1875        on.) This code has now been added to pcre_study(); it finds a lower bound
1876        to the length of subject needed. It is not necessarily the greatest lower
1877        bound, but using it to avoid searching strings that are too short does give
1878        some useful speed-ups. The value is available to calling programs via
1879        pcre_fullinfo().
1880    
1881    26. While implementing 25, I discovered to my embarrassment that pcretest had
1882        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1883        study optimizations had never been tested with that matching function.
1884        Oops. What is worse, even when it was passed study data, there was a bug in
1885        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1886        were also very few tests of studied patterns with pcre_dfa_exec().
1887    
1888    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1889        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1890        on the other side of the coin, they are no longer allowed to have different
1891        names, because these cannot be distinguished in PCRE, and this has caused
1892        confusion. (This is a difference from Perl.)
1893    
1894    28. When duplicate subpattern names are present (necessarily with different
1895        numbers, as required by 27 above), and a test is made by name in a
1896        conditional pattern, either for a subpattern having been matched, or for
1897        recursion in such a pattern, all the associated numbered subpatterns are
1898        tested, and the overall condition is true if the condition is true for any
1899        one of them. This is the way Perl works, and is also more like the way
1900        testing by number works.
1901    
1902    
1903    Version 7.9 11-Apr-09
1904    ---------------------
1905    
1906    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1907        (pcretest), all targets were linked against these libraries. This included
1908        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1909        libraries. This caused unwanted dependencies to be created. This problem
1910        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1911        pcretest is linked with readline.
1912    
1913    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1914        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1915        moved outside it again, because FALSE and TRUE are already defined in AIX,
1916        but BOOL is not.
1917    
1918    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1919        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1920    
1921    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1922        hyphens) following file names and line numbers when outputting matching
1923        lines. This is not true; no spaces are inserted. I have also clarified the
1924        wording for the --colour (or --color) option.
1925    
1926    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1927        was not coloured; this is different to GNU grep, so I have changed it to be
1928        the same.
1929    
1930    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1931        each matching line was coloured. Now it goes on to look for further matches
1932        of any of the test patterns, which is the same behaviour as GNU grep.
1933    
1934    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1935        doesn't make sense to accept an empty string match in pcregrep, so I have
1936        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1937        seems to be how GNU grep behaves. [But see later change 40 for release
1938        8.33.]
1939    
1940    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1941        start or after a newline", because the conditional assertion was not being
1942        correctly handled. The rule now is that both the assertion and what follows
1943        in the first alternative must satisfy the test.
1944    
1945    9.  If auto-callout was enabled in a pattern with a conditional group whose
1946        condition was an assertion, PCRE could crash during matching, both with
1947        pcre_exec() and pcre_dfa_exec().
1948    
1949    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1950        used for matching.
1951    
1952    11. Unicode property support in character classes was not working for
1953        characters (bytes) greater than 127 when not in UTF-8 mode.
1954    
1955    12. Added the -M command line option to pcretest.
1956    
1957    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1958    
1959    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1960    
1961    16. Added comments and documentation about mis-use of no_arg in the C++
1962        wrapper.
1963    
1964    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1965        from Martin Jerabek that uses macro names for all relevant character and
1966        string constants.
1967    
1968    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1969        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1970        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1971        these, but not everybody uses configure.
1972    
1973    19. A conditional group that had only one branch was not being correctly
1974        recognized as an item that could match an empty string. This meant that an
1975        enclosing group might also not be so recognized, causing infinite looping
1976        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1977        with the subject "ab", where knowledge that the repeated group can match
1978        nothing is needed in order to break the loop.
1979    
1980    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1981        exec(), but without supplying a callout function, matching went wrong.
1982    
1983    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1984        leak if the size of the offset vector was greater than 30. When the vector
1985        is smaller, the saved offsets during recursion go onto a local stack
1986        vector, but for larger vectors malloc() is used. It was failing to free
1987        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1988        error, in fact).
1989    
1990    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1991        heapframe that is used only when UTF-8 support is enabled. This caused no
1992        problem, but was untidy.
1993    
1994    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1995        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1996        included within another project.
1997    
1998    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1999        slightly modified by me:
2000    
2001          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2002              not building pcregrep.
2003    
2004          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2005              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2006    
2007    25. Forward references, both numeric and by name, in patterns that made use of
2008        duplicate group numbers, could behave incorrectly or give incorrect errors,
2009        because when scanning forward to find the reference group, PCRE was not
2010        taking into account the duplicate group numbers. A pattern such as
2011        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2012    
2013    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2014        the feature of a custom pointer more persuasive (as requested by a user).
2015    
2016    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2017        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2018        pre-defined.
2019    
2020    28. Added support for (*UTF8) at the start of a pattern.
2021    
2022    29. Arrange for flags added by the "release type" setting in CMake to be shown
2023        in the configuration summary.
2024    
2025    
2026    Version 7.8 05-Sep-08
2027    ---------------------
2028    
2029    1.  Replaced UCP searching code with optimized version as implemented for Ad
2030        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2031        stage table and inline lookup instead of a function, giving speed ups of 2
2032        to 5 times on some simple patterns that I tested. Permission was given to
2033        distribute the MultiStage2.py script that generates the tables (it's not in
2034        the tarball, but is in the Subversion repository).
2035    
2036    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2037        scripts.
2038    
2039    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2040        a group with a zero qualifier. The result of the study could be incorrect,
2041        or the function might crash, depending on the pattern.
2042    
2043    4.  Caseless matching was not working for non-ASCII characters in back
2044        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2045        It now works when Unicode Property Support is available.
2046    
2047    5.  In pcretest, an escape such as \x{de} in the data was always generating
2048        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2049        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2050        truncation.
2051    
2052    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2053    
2054    7.  Added two (int) casts to pcregrep when printing the difference of two
2055        pointers, in case they are 64-bit values.
2056    
2057    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2058        test 2 if it fails.
2059    
2060    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2061        and a #define of that name to empty if it is not externally set. This is to
2062        allow users of MSVC to set it if necessary.
2063    
2064    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2065        the convenience functions in the pcre_get.c source file.
2066    
2067    11. An option change at the start of a pattern that had top-level alternatives
2068        could cause overwriting and/or a crash. This command provoked a crash in
2069        some environments:
2070    
2071          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2072    
2073        This potential security problem was recorded as CVE-2008-2371.
2074    
2075    12. For a pattern where the match had to start at the beginning or immediately
2076        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2077        pcre_dfa_exec() could read past the end of the passed subject if there was
2078        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2079        pcretest so that it places the subject at the end of its malloc-ed buffer.
2080    
2081    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2082        exec() might read past the end of the data buffer in UTF-8 mode.
2083    
2084    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2085        the data contained the byte 0x85 as part of a UTF-8 character within its
2086        first line. This applied both to normal and DFA matching.
2087    
2088    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2089        /^[^d]*?$/8 failed to match "abc".
2090    
2091    16. Added a missing copyright notice to pcrecpp_internal.h.
2092    
2093    17. Make it more clear in the documentation that values returned from
2094        pcre_exec() in ovector are byte offsets, not character counts.
2095    
2096    18. Tidied a few places to stop certain compilers from issuing warnings.
2097    
2098    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2099        supplied by Stefan Weber. I made a further small update for 7.8 because
2100        there is a change of source arrangements: the pcre_searchfuncs.c module is
2101        replaced by pcre_ucd.c.
2102    
2103    
2104    Version 7.7 07-May-08
2105    ---------------------
2106    
2107    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2108        a string to a long long, pretend we don't even have a long long." This is
2109        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2110    
2111    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2112        pre-7.6 versions, which defined a global no_arg variable instead of putting
2113        it in the RE class. (See also #8 below.)
2114    
2115    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2116        Lopes.
2117    
2118    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2119    
2120        (1) The include/exclude patterns were being applied to the whole pathnames
2121            of files, instead of just to the final components.
2122    
2123        (2) If there was more than one level of directory, the subdirectories were
2124            skipped unless they satisfied the include/exclude conditions. This is
2125            inconsistent with GNU grep (and could even be seen as contrary to the
2126            pcregrep specification - which I improved to make it absolutely clear).
2127            The action now is always to scan all levels of directory, and just
2128            apply the include/exclude patterns to regular files.
2129    
2130    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2131        --exclude_dir in the tests to avoid scanning .svn directories.
2132    
2133    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2134        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2135        doesn't support NULs in patterns.
2136    
2137    7.  Added some missing "const"s to declarations of static tables in
2138        pcre_compile.c and pcre_dfa_exec.c.
2139    
2140    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2141        caused by fix #2  above. (Subsequently also a second patch to fix the
2142        first patch. And a third patch - this was a messy problem.)
2143    
2144    9.  Applied Craig's patch to remove the use of push_back().
2145    
2146    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2147        matching function regexec().
2148    
2149    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2150        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2151        references. PCRE supports relative numbers with this syntax (I don't think
2152        Oniguruma does).
2153    
2154    12. Previously, a group with a zero repeat such as (...){0} was completely
2155        omitted from the compiled regex. However, this means that if the group
2156        was called as a subroutine from elsewhere in the pattern, things went wrong
2157        (an internal error was given). Such groups are now left in the compiled
2158        pattern, with a new opcode that causes them to be skipped at execution
2159        time.
2160    
2161    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2162        to the way PCRE behaves:
2163    
2164        (a) A lone ] character is dis-allowed (Perl treats it as data).
2165    
2166        (b) A back reference to an unmatched subpattern matches an empty string
2167            (Perl fails the current match path).
2168    
2169        (c) A data ] in a character class must be notated as \] because if the
2170            first data character in a class is ], it defines an empty class. (In
2171            Perl it is not possible to have an empty class.) The empty class []
2172            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2173            The negative empty class [^] matches any one character, independently
2174            of the DOTALL setting.
2175    
2176    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2177        non-existent subpattern following a character class starting with ']' and
2178        containing () gave an internal compiling error instead of "reference to
2179        non-existent subpattern". Fortunately, when the pattern did exist, the
2180        compiled code was correct. (When scanning forwards to check for the
2181        existence of the subpattern, it was treating the data ']' as terminating
2182        the class, so got the count wrong. When actually compiling, the reference
2183        was subsequently set up correctly.)
2184    
2185    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2186        it was being rejected as not supported by pcre_dfa_exec(), even though
2187        other assertions are supported. I have made pcre_dfa_exec() support
2188        (*FAIL).
2189    
2190    16. The implementation of 13c above involved the invention of a new opcode,
2191        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2192        cannot be changed at match time, I realized I could make a small
2193        improvement to matching performance by compiling OP_ALLANY instead of
2194        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2195        on the OP_ANY path.
2196    
2197    17. Compiling pcretest on Windows with readline support failed without the
2198        following two fixes: (1) Make the unistd.h include conditional on
2199        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2200    
2201    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2202        ncurses library to be included for pcretest when ReadLine support is
2203        requested, but also to allow for it to be overridden. This patch came from
2204        Daniel Bergström.
2205    
2206    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2207        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2208        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2209        spotting this.
2210    
2211    
2212  Version 7.6 28-Jan-08  Version 7.6 28-Jan-08
2213  ---------------------  ---------------------
2214    

Legend:
Removed from v.319  
changed lines
  Added in v.1434

  ViewVC Help
Powered by ViewVC 1.1.5