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

Diff of /code/trunk/ChangeLog

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

revision 221 by ph10, Fri Aug 17 09:25:08 2007 UTC revision 1504 by ph10, Fri Sep 26 09:21:46 2014 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.3 17-Aug-07  Version 8.36 26-September-2014
5    ------------------------------
6    
7    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
8        -Wmissing-field-initializers and -Wunused-parameter.
9    
10    2.  The tests for quantifiers being too big (greater than 65535) were being
11        applied after reading the number, and stupidly assuming that integer
12        overflow would give a negative number. The tests are now applied as the
13        numbers are read.
14    
15    3.  Tidy code in pcre_exec.c where two branches that used to be different are
16        now the same.
17    
18    4.  The JIT compiler did not generate match limit checks for certain
19        bracketed expressions with quantifiers. This may lead to exponential
20        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
21        issue should be resolved now.
22    
23    5.  Fixed an issue, which occures when nested alternatives are optimized
24        with table jumps.
25    
26    6.  Inserted two casts and changed some ints to size_t in the light of some
27        reported 64-bit compiler warnings (Bugzilla 1477).
28    
29    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
30        an empty string, which sometimes were behaving incorrectly in the
31        interpreter (though correctly in the JIT matcher). This pcretest input is
32        an example:
33    
34          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
35          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
36    
37        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
38        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
39        for an empty string was breaking the inner loop and carrying on at a lower
40        level, when possessive repeated groups should always return to a higher
41        level as they have no backtrack points in them. The empty string test now
42        occurs at the outer level.
43    
44    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
45        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
46    
47    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
48        doesn't).
49    
50    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
51        to hold a pattern to be incorrectly computed (too small) when there were
52        named back references to duplicated names. This could cause "internal
53        error: code overflow" or "double free or corruption" or other memory
54        handling errors.
55    
56    11. When named subpatterns had the same prefixes, back references could be
57        confused. For example, in this pattern:
58    
59          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
60    
61        the reference to 'Name' was incorrectly treated as a reference to a
62        duplicate name.
63    
64    12. A pattern such as /^s?c/mi8 where the optional character has more than
65        one "other case" was incorrectly compiled such that it would only try to
66        match starting at "c".
67    
68    13. When a pattern starting with \s was studied, VT was not included in the
69        list of possible starting characters; this should have been part of the
70        8.34/18 patch.
71    
72    14. If a character class started [\Qx]... where x is any character, the class
73        was incorrectly terminated at the ].
74    
75    15. If a pattern that started with a caseless match for a character with more
76        than one "other case" was studied, PCRE did not set up the starting code
77        unit bit map for the list of possible characters. Now it does. This is an
78        optimization improvement, not a bug fix.
79    
80    16. The Unicode data tables have been updated to Unicode 7.0.0.
81    
82    17. Fixed a number of memory leaks in pcregrep.
83    
84    18. Avoid a compiler warning (from some compilers) for a function call with
85        a cast that removes "const" from an lvalue by using an intermediate
86        variable (to which the compiler does not object).
87    
88    19. Incorrect code was compiled if a group that contained an internal recursive
89        back reference was optional (had quantifier with a minimum of zero). This
90        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
91        caused segmentation faults because of stack overflows at compile time.
92    
93    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
94        group that is quantified with an indefinite repeat, caused a compile-time
95        loop which used up all the system stack and provoked a segmentation fault.
96        This was not the same bug as 19 above.
97    
98    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
99        Patch by Mike Frysinger.
100    
101    
102    Version 8.35 04-April-2014
103    --------------------------
104    
105    1.  A new flag is set, when property checks are present in an XCLASS.
106        When this flag is not set, PCRE can perform certain optimizations
107        such as studying these XCLASS-es.
108    
109    2.  The auto-possessification of character sets were improved: a normal
110        and an extended character set can be compared now. Furthermore
111        the JIT compiler optimizes more character set checks.
112    
113    3.  Got rid of some compiler warnings for potentially uninitialized variables
114        that show up only when compiled with -O2.
115    
116    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
117        of a match later then the end of the match. The pcretest program was not
118        handling the case sensibly - it was outputting from the start to the next
119        binary zero. It now reports this situation in a message, and outputs the
120        text from the end to the start.
121    
122    5.  Fast forward search is improved in JIT. Instead of the first three
123        characters, any three characters with fixed position can be searched.
124        Search order: first, last, middle.
125    
126    6.  Improve character range checks in JIT. Characters are read by an inprecise
127        function now, which returns with an unknown value if the character code is
128        above a certain threshold (e.g: 256). The only limitation is that the value
129        must be bigger than the threshold as well. This function is useful when
130        the characters above the threshold are handled in the same way.
131    
132    7.  The macros whose names start with RAWUCHAR are placeholders for a future
133        mode in which only the bottom 21 bits of 32-bit data items are used. To
134        make this more memorable for those maintaining the code, the names have
135        been changed to start with UCHAR21, and an extensive comment has been added
136        to their definition.
137    
138    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
139        to the export list in Makefile.am (they were accidentally omitted from the
140        8.34 tarball).
141    
142    9.  The informational output from pcretest used the phrase "starting byte set"
143        which is inappropriate for the 16-bit and 32-bit libraries. As the output
144        for "first char" and "need char" really means "non-UTF-char", I've changed
145        "byte" to "char", and slightly reworded the output. The documentation about
146        these values has also been (I hope) clarified.
147    
148    10. Another JIT related optimization: use table jumps for selecting the correct
149        backtracking path, when more than four alternatives are present inside a
150        bracket.
151    
152    11. Empty match is not possible, when the minimum length is greater than zero,
153        and there is no \K in the pattern. JIT should avoid empty match checks in
154        such cases.
155    
156    12. In a caseless character class with UCP support, when a character with more
157        than one alternative case was not the first character of a range, not all
158        the alternative cases were added to the class. For example, s and \x{17f}
159        are both alternative cases for S: the class [RST] was handled correctly,
160        but [R-T] was not.
161    
162    13. The configure.ac file always checked for pthread support when JIT was
163        enabled. This is not used in Windows, so I have put this test inside a
164        check for the presence of windows.h (which was already tested for).
165    
166    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
167        The algorithm provides a way to skip certain starting offsets, and usually
168        faster than linear prefix searches.
169    
170    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
171        as for 'fr_FR' and 'french'. For some reason, however, it then used the
172        Windows-specific input and output files, which have 'french' screwed in.
173        So this could never have worked. One of the problems with locales is that
174        they aren't always the same. I have now updated RunTest so that it checks
175        the output of the locale test (test 3) against three different output
176        files, and it allows the test to pass if any one of them matches. With luck
177        this should make the test pass on some versions of Solaris where it was
178        failing. Because of the uncertainty, the script did not used to stop if
179        test 3 failed; it now does. If further versions of a French locale ever
180        come to light, they can now easily be added.
181    
182    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
183        there was a message during ./configure, but it did not stop. This now
184        provokes an error. The invalid example in README has been corrected.
185        If a value less than the minimum is given, the minimum value has always
186        been used, but now a warning is given.
187    
188    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
189        was a bug in the test system, which is now fixed. Also, the list of various
190        configurations that are tested for each release did not have one with both
191        16/32 bits and --enable-bar-anycrlf. It now does.
192    
193    18. pcretest was missing "-C bsr" for displaying the \R default setting.
194    
195    19. Little endian PowerPC systems are supported now by the JIT compiler.
196    
197    20. The fast forward newline mechanism could enter to an infinite loop on
198        certain invalid UTF-8 input. Although we don't support these cases
199        this issue can be fixed by a performance optimization.
200    
201    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
202        not take account if existing stack usage. There is now a new global
203        variable called pcre_stack_guard that can be set to point to an external
204        function to check stack availability. It is called at the start of
205        processing every parenthesized group.
206    
207    22. A typo in the code meant that in ungreedy mode the max/min qualifier
208        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
209        not match "ab".
210    
211    23. When UTF was disabled, the JIT program reported some incorrect compile
212        errors. These messages are silenced now.
213    
214    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
215        compiler.
216    
217    25. Change all the temporary files used in RunGrepTest to be different to those
218        used by RunTest so that the tests can be run simultaneously, for example by
219        "make -j check".
220    
221    
222    Version 8.34 15-December-2013
223    -----------------------------
224    
225    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
226        executable memory. Patch inspired by Carsten Klein.
227    
228    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
229        this macro is never tested and has no effect, because the work to support
230        coverage involves only compiling and linking options and special targets in
231        the Makefile. The comment in config.h implied that defining the macro would
232        enable coverage support, which is totally false. There was also support for
233        setting this macro in the CMake files (my fault, I just copied it from
234        configure). SUPPORT_GCOV has now been removed.
235    
236    3.  Make a small performance improvement in strlen16() and strlen32() in
237        pcretest.
238    
239    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
240        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
241        these cases). There was also one in pcretest.c.
242    
243    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
244    
245    6.  In UTF mode, the code for checking whether a group could match an empty
246        string (which is used for indefinitely repeated groups to allow for
247        breaking an infinite loop) was broken when the group contained a repeated
248        negated single-character class with a character that occupied more than one
249        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
250        UTF-8 mode). The effect was undefined: the group might or might not be
251        deemed as matching an empty string, or the program might have crashed.
252    
253    7.  The code for checking whether a group could match an empty string was not
254        recognizing that \h, \H, \v, \V, and \R must match a character.
255    
256    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
257        an empty string. If it can, pcretest shows this in its information output.
258    
259    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
260        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
261        matched by pcre_exec() without using JIT:
262    
263        (a) If the rest of the pattern did not match after a maximal run of
264            grapheme clusters, the code for backing up to try with fewer of them
265            did not always back up over a full grapheme when characters that do not
266            have the modifier quality were involved, e.g. Hangul syllables.
267    
268        (b) If the match point in a subject started with modifier character, and
269            there was no match, the code could incorrectly back up beyond the match
270            point, and potentially beyond the first character in the subject,
271            leading to a segfault or an incorrect match result.
272    
273    10. A conditional group with an assertion condition could lead to PCRE
274        recording an incorrect first data item for a match if no other first data
275        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
276        first data item, and therefore matched "ca" after "c" instead of at the
277        start.
278    
279    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
280        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
281    
282    12. The source of pcregrep now includes z/OS-specific code so that it can be
283        compiled for z/OS as part of the special z/OS distribution.
284    
285    13. Added the -T and -TM options to pcretest.
286    
287    14. The code in pcre_compile.c for creating the table of named capturing groups
288        has been refactored. Instead of creating the table dynamically during the
289        actual compiling pass, the information is remembered during the pre-compile
290        pass (on the stack unless there are more than 20 named groups, in which
291        case malloc() is used) and the whole table is created before the actual
292        compile happens. This has simplified the code (it is now nearly 150 lines
293        shorter) and prepared the way for better handling of references to groups
294        with duplicate names.
295    
296    15. A back reference to a named subpattern when there is more than one of the
297        same name now checks them in the order in which they appear in the pattern.
298        The first one that is set is used for the reference. Previously only the
299        first one was inspected. This change makes PCRE more compatible with Perl.
300    
301    16. Unicode character properties were updated from Unicode 6.3.0.
302    
303    17. The compile-time code for auto-possessification has been refactored, based
304        on a patch by Zoltan Herczeg. It now happens after instead of during
305        compilation. The code is cleaner, and more cases are handled. The option
306        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
307        options in pcretest are provided to set it. It can also be set by
308        (*NO_AUTO_POSSESS) at the start of a pattern.
309    
310    18. The character VT has been added to the default ("C" locale) set of
311        characters that match \s and are generally treated as white space,
312        following this same change in Perl 5.18. There is now no difference between
313        "Perl space" and "POSIX space". Whether VT is treated as white space in
314        other locales depends on the locale.
315    
316    19. The code for checking named groups as conditions, either for being set or
317        for being recursed, has been refactored (this is related to 14 and 15
318        above). Processing unduplicated named groups should now be as fast at
319        numerical groups, and processing duplicated groups should be faster than
320        before.
321    
322    20. Two patches to the CMake build system, by Alexander Barkov:
323    
324          (1) Replace the "source" command by "." in CMakeLists.txt because
325              "source" is a bash-ism.
326    
327          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
328              without these the CMake build does not work on Solaris.
329    
330    21. Perl has changed its handling of \8 and \9. If there is no previously
331        encountered capturing group of those numbers, they are treated as the
332        literal characters 8 and 9 instead of a binary zero followed by the
333        literals. PCRE now does the same.
334    
335    22. Following Perl, added \o{} to specify codepoints in octal, making it
336        possible to specify values greater than 0777 and also making them
337        unambiguous.
338    
339    23. Perl now gives an error for missing closing braces after \x{... instead of
340        treating the string as literal. PCRE now does the same.
341    
342    24. RunTest used to grumble if an inappropriate test was selected explicitly,
343        but just skip it when running all tests. This make it awkward to run ranges
344        of tests when one of them was inappropriate. Now it just skips any
345        inappropriate tests, as it always did when running all tests.
346    
347    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
348        character types such as \d or \w, too many callouts were inserted, and the
349        data that they returned was rubbish.
350    
351    26. In UCP mode, \s was not matching two of the characters that Perl matches,
352        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
353        were matched by \h. The code has now been refactored so that the lists of
354        the horizontal and vertical whitespace characters used for \h and \v (which
355        are defined only in one place) are now also used for \s.
356    
357    27. Add JIT support for the 64 bit TileGX architecture.
358        Patch by Jiong Wang (Tilera Corporation).
359    
360    28. Possessive quantifiers for classes (both explicit and automatically
361        generated) now use special opcodes instead of wrapping in ONCE brackets.
362    
363    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
364        (because it's meaningless), this was not happening when PCRE_CASELESS was
365        set. Not wrong, but inefficient.
366    
367    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
368        properties for \w, \d, etc) is present in a test regex. Otherwise if the
369        test contains no characters greater than 255, Perl doesn't realise it
370        should be using Unicode semantics.
371    
372    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
373        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
374        does in Unicode mode.
375    
376    32. Added the "forbid" facility to pcretest so that putting tests into the
377        wrong test files can sometimes be quickly detected.
378    
379    33. There is now a limit (default 250) on the depth of nesting of parentheses.
380        This limit is imposed to control the amount of system stack used at compile
381        time. It can be changed at build time by --with-parens-nest-limit=xxx or
382        the equivalent in CMake.
383    
384    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
385        errors. Perl warns for these when in warning mode, but PCRE has no facility
386        for giving warnings.
387    
388    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
389        However, this was not working for (?!) because it is optimized to (*FAIL),
390        for which PCRE does not allow quantifiers. The optimization is now disabled
391        when a quantifier follows (?!). I can't see any use for this, but it makes
392        things uniform.
393    
394    36. Perl no longer allows group names to start with digits, so I have made this
395        change also in PCRE. It simplifies the code a bit.
396    
397    37. In extended mode, Perl ignores spaces before a + that indicates a
398        possessive quantifier. PCRE allowed a space before the quantifier, but not
399        before the possessive +. It now does.
400    
401    38. The use of \K (reset reported match start) within a repeated possessive
402        group such as (a\Kb)*+ was not working.
403    
404    40. Document that the same character tables must be used at compile time and
405        run time, and that the facility to pass tables to pcre_exec() and
406        pcre_dfa_exec() is for use only with saved/restored patterns.
407    
408    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
409        features for Builds with MSVC:
410    
411        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
412           builds). The .rc files can be used to set FileDescription and many other
413           attributes.
414    
415        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
416           This allows higher-level build scripts which want .pdb files to avoid
417           hard-coding the exact files needed."
418    
419    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
420        mean "start of word" and "end of word", respectively, as a transition aid.
421    
422    43. A minimizing repeat of a class containing codepoints greater than 255 in
423        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
424        compiled to use the heap for recursion.
425    
426    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
427        is configured.
428    
429    
430    Version 8.33 28-May-2013
431    ------------------------
432    
433    1.  Added 'U' to some constants that are compared to unsigned integers, to
434        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
435        variables that are added to signed variables, to ensure the result is
436        signed and can be negated.
437    
438    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
439        CMake config files.
440    
441    3.  Revise the creation of config.h.generic so that all boolean macros are
442        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
443        overriding via -D on the command line possible.
444    
445    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
446        to unsigned int is reported to make a quite noticeable speed difference in
447        a specific Windows environment. Testing on Linux did also appear to show
448        some benefit (and it is clearly not harmful). Also fixed the definition of
449        Xop which should be unsigned.
450    
451    5.  Related to (4), changing the definition of the intermediate variable cc
452        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
453        improvements.
454    
455    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
456        unnecessary spaces.
457    
458    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
459        and later.
460    
461    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
462    
463    9.  Optimizing fast_forward_start_bits in JIT.
464    
465    10. Adding support for callouts in JIT, and fixing some issues revealed
466        during this work. Namely:
467    
468        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
469    
470        (b) Minimum length was not checked before the matching is started.
471    
472    11. The value of capture_last that is passed to callouts was incorrect in some
473        cases when there was a capture on one path that was subsequently abandoned
474        after a backtrack. Also, the capture_last value is now reset after a
475        recursion, since all captures are also reset in this case.
476    
477    12. The interpreter no longer returns the "too many substrings" error in the
478        case when an overflowing capture is in a branch that is subsequently
479        abandoned after a backtrack.
480    
481    13. In the pathological case when an offset vector of size 2 is used, pcretest
482        now prints out the matched string after a yield of 0 or 1.
483    
484    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
485        Only supported by the JIT compiler at the moment.
486    
487    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
488    
489    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
490        offset of the starting point of the matching process, provided the offsets
491        vector is large enough.
492    
493    17. The \A escape now records a lookbehind value of 1, though its execution
494        does not actually inspect the previous character. This is to ensure that,
495        in partial multi-segment matching, at least one character from the old
496        segment is retained when a new segment is processed. Otherwise, if there
497        are no lookbehinds in the pattern, \A might match incorrectly at the start
498        of a new segment.
499    
500    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
501    
502    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
503        gives some modest performance improvement in 8-bit mode.
504    
505    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
506        be expressed in certain programming languages using Universal Character
507        Names.
508    
509    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
510        which points out that "non characters" are not "characters that may not
511        appear in Unicode strings" but rather "characters that are reserved for
512        internal use and have only local meaning".
513    
514    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
515        there was a conditional group that depended on an assertion, if the
516        assertion was false, the callout that immediately followed the alternation
517        in the condition was skipped when pcre_exec() was used for matching.
518    
519    23. Allow an explicit callout to be inserted before an assertion that is the
520        condition for a conditional group, for compatibility with automatic
521        callouts, which always insert a callout at this point.
522    
523    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
524        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
525    
526    25. (*PRUNE) is now supported by the JIT compiler.
527    
528    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
529    
530    27. Fix the case where there are two or more SKIPs with arguments that may be
531        ignored.
532    
533    28. (*SKIP) is now supported by the JIT compiler.
534    
535    29. (*THEN) is now supported by the JIT compiler.
536    
537    30. Update RunTest with additional test selector options.
538    
539    31. The way PCRE handles backtracking verbs has been changed in two ways.
540    
541        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
542        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
543        backtracking. In some cases this makes it more Perl-compatible, but Perl's
544        rather obscure rules do not always do the same thing.
545    
546        (2) Previously, backtracking verbs were confined within assertions. This is
547        no longer the case for positive assertions, except for (*ACCEPT). Again,
548        this sometimes improves Perl compatibility, and sometimes does not.
549    
550    32. A number of tests that were in test 2 because Perl did things differently
551        have been moved to test 1, because either Perl or PCRE has changed, and
552        these tests are now compatible.
553    
554    32. Backtracking control verbs are now handled in the same way in JIT and
555        interpreter.
556    
557    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
558        contained a forward subroutine reference caused a compile error.
559    
560    34. Auto-detect and optimize limited repetitions in JIT.
561    
562    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
563        blocking (*UTF) etc.
564    
565    36. In the interpreter, maximizing pattern repetitions for characters and
566        character types now use tail recursion, which reduces stack usage.
567    
568    37. The value of the max lookbehind was not correctly preserved if a compiled
569        and saved regex was reloaded on a host of different endianness.
570    
571    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
572        of the compiled pattern block, expand the flags field from 16 to 32 bits
573        because it was almost full.
574    
575    39. Try madvise first before posix_madvise.
576    
577    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
578        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
579        original change locked out all matches of empty strings. This has been
580        changed so that one match of an empty string per line is recognized.
581        Subsequent searches on the same line (for colouring or for --only-matching,
582        for example) do not recognize empty strings.
583    
584    41. Applied a user patch to fix a number of spelling mistakes in comments.
585    
586    42. Data lines longer than 65536 caused pcretest to crash.
587    
588    43. Clarified the data type for length and startoffset arguments for pcre_exec
589        and pcre_dfa_exec in the function-specific man pages, where they were
590        explicitly stated to be in bytes, never having been updated. I also added
591        some clarification to the pcreapi man page.
592    
593    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
594        a segmentation fault.
595    
596    
597    Version 8.32 30-November-2012
598    -----------------------------
599    
600    1.  Improved JIT compiler optimizations for first character search and single
601        character iterators.
602    
603    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
604        Patch by Daniel Richard G.
605    
606    3.  Single character iterator optimizations in the JIT compiler.
607    
608    4.  Improved JIT compiler optimizations for character ranges.
609    
610    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
611        Reported by Giuseppe D'Angelo.
612    
613    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
614        of a line, was being set incorrectly in cases where .* appeared inside
615        atomic brackets at the start of a pattern, or where there was a subsequent
616        *PRUNE or *SKIP.
617    
618    7.  Improved instruction cache flush for POWER/PowerPC.
619        Patch by Daniel Richard G.
620    
621    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
622        grep:
623    
624        (a) There is now no limit to the number of patterns to be matched.
625    
626        (b) An error is given if a pattern is too long.
627    
628        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
629            are now supported.
630    
631        (d) --exclude-from and --include-from (multiple use) have been added.
632    
633        (e) Exclusions and inclusions now apply to all files and directories, not
634            just to those obtained from scanning a directory recursively.
635    
636        (f) Multiple uses of -f and --file-list are now supported.
637    
638        (g) In a Windows environment, the default for -d has been changed from
639            "read" (the GNU grep default) to "skip", because otherwise the presence
640            of a directory in the file list provokes an error.
641    
642        (h) The documentation has been revised and clarified in places.
643    
644    9.  Improve the matching speed of capturing brackets.
645    
646    10. Changed the meaning of \X so that it now matches a Unicode extended
647        grapheme cluster.
648    
649    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
650        out POSIX threads when JIT support is configured.
651    
652    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
653    
654    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
655        structure could go wrong in environments where size_t is not the same size
656        as int.
657    
658    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
659        set.
660    
661    15. The EBCDIC support had decayed; later updates to the code had included
662        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
663        general tidy up of EBCDIC-related issues, and the documentation was also
664        not quite right. There is now a test that can be run on ASCII systems to
665        check some of the EBCDIC-related things (but is it not a full test).
666    
667    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
668        in a small tidy to the code.
669    
670    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
671    
672    18. If the --only-matching (-o) option in pcregrep is specified multiple
673        times, each one causes appropriate output. For example, -o1 -o2 outputs the
674        substrings matched by the 1st and 2nd capturing parentheses. A separating
675        string can be specified by --om-separator (default empty).
676    
677    19. Improving the first n character searches.
678    
679    20. Turn case lists for horizontal and vertical white space into macros so that
680        they are defined only once.
681    
682    21. This set of changes together give more compatible Unicode case-folding
683        behaviour for characters that have more than one other case when UCP
684        support is available.
685    
686        (a) The Unicode property table now has offsets into a new table of sets of
687            three or more characters that are case-equivalent. The MultiStage2.py
688            script that generates these tables (the pcre_ucd.c file) now scans
689            CaseFolding.txt instead of UnicodeData.txt for character case
690            information.
691    
692        (b) The code for adding characters or ranges of characters to a character
693            class has been abstracted into a generalized function that also handles
694            case-independence. In UTF-mode with UCP support, this uses the new data
695            to handle characters with more than one other case.
696    
697        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
698            whose other case is greater than 256 are now correctly matched
699            caselessly. Previously, the high codepoint matched the low one, but not
700            vice versa.
701    
702        (d) The processing of \h, \H, \v, and \ in character classes now makes use
703            of the new class addition function, using character lists defined as
704            macros alongside the case definitions of 20 above.
705    
706        (e) Caseless back references now work with characters that have more than
707            one other case.
708    
709        (f) General caseless matching of characters with more than one other case
710            is supported.
711    
712    22. Unicode character properties were updated from Unicode 6.2.0
713    
714    23. Improved CMake support under Windows. Patch by Daniel Richard G.
715    
716    24. Add support for 32-bit character strings, and UTF-32
717    
718    25. Major JIT compiler update (code refactoring and bugfixing).
719        Experimental Sparc 32 support is added.
720    
721    26. Applied a modified version of Daniel Richard G's patch to create
722        pcre.h.generic and config.h.generic by "make" instead of in the
723        PrepareRelease script.
724    
725    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
726        pcre_compile.c when checking for a zero character.
727    
728    28. Introducing a native interface for JIT. Through this interface, the compiled
729        machine code can be directly executed. The purpose of this interface is to
730        provide fast pattern matching, so several sanity checks are not performed.
731        However, feature tests are still performed. The new interface provides
732        1.4x speedup compared to the old one.
733    
734    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
735        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
736        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
737    
738    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
739        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
740        the "old" RFC 2279). Instead, it ended up passing a negative length to
741        pcre_exec().
742    
743    31. Add support for GCC's visibility feature to hide internal functions.
744    
745    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
746        "unknown -C option" after outputting 0 or 1.
747    
748    33. There is now support for generating a code coverage report for the test
749        suite in environments where gcc is the compiler and lcov is installed. This
750        is mainly for the benefit of the developers.
751    
752    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
753        unaddressable using valgrind annotations, allowing valgrind to detect
754        invalid memory accesses. This is mainly for the benefit of the developers.
755    
756    25. (*UTF) can now be used to start a pattern in any of the three libraries.
757    
758    26. Give configure error if --enable-cpp but no C++ compiler found.
759    
760    
761    Version 8.31 06-July-2012
762    -------------------------
763    
764    1.  Fixing a wrong JIT test case and some compiler warnings.
765    
766    2.  Removed a bashism from the RunTest script.
767    
768    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
769        to unsigned type, result still unsigned" that was given by an MS compiler
770        on encountering the code "-sizeof(xxx)".
771    
772    4.  Partial matching support is added to the JIT compiler.
773    
774    5.  Fixed several bugs concerned with partial matching of items that consist
775        of more than one character:
776    
777        (a) /^(..)\1/ did not partially match "aba" because checking references was
778            done on an "all or nothing" basis. This also applied to repeated
779            references.
780    
781        (b) \R did not give a hard partial match if \r was found at the end of the
782            subject.
783    
784        (c) \X did not give a hard partial match after matching one or more
785            characters at the end of the subject.
786    
787        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
788            a partial match for the string "\r".
789    
790        (e) When newline was set to CRLF, the metacharacter "." did not recognize
791            a partial match for a CR character at the end of the subject string.
792    
793    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
794        running pcretest, the text "(JIT)" added to the output whenever JIT is
795        actually used to run the match.
796    
797    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
798        or /S+[+] with a digit between 1 and 7.
799    
800    8.  OP_NOT now supports any UTF character not just single-byte ones.
801    
802    9.  (*MARK) control verb is now supported by the JIT compiler.
803    
804    10. The command "./RunTest list" lists the available tests without actually
805        running any of them. (Because I keep forgetting what they all are.)
806    
807    11. Add PCRE_INFO_MAXLOOKBEHIND.
808    
809    12. Applied a (slightly modified) user-supplied patch that improves performance
810        when the heap is used for recursion (compiled with --disable-stack-for-
811        recursion). Instead of malloc and free for each heap frame each time a
812        logical recursion happens, frames are retained on a chain and re-used where
813        possible. This sometimes gives as much as 30% improvement.
814    
815    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
816        call.
817    
818    14. As documented, (*COMMIT) is now confined to within a positive assertion.
819    
820    15. It is now possible to link pcretest with libedit as an alternative to
821        libreadline.
822    
823    16. (*COMMIT) control verb is now supported by the JIT compiler.
824    
825    17. The Unicode data tables have been updated to Unicode 6.1.0.
826    
827    18. Added --file-list option to pcregrep.
828    
829    19. Added binary file support to pcregrep, including the -a, --binary-files,
830        -I, and --text options.
831    
832    20. The madvise function is renamed for posix_madvise for QNX compatibility
833        reasons. Fixed by Giuseppe D'Angelo.
834    
835    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
836    
837    22. Changed the option for creating symbolic links for 16-bit man pages from
838        -s to -sf so that re-installing does not cause issues.
839    
840    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
841    
842    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
843        matches in certain environments (the workspace was not being correctly
844        retained). Also added to pcre_dfa_exec() a simple plausibility check on
845        some of the workspace data at the beginning of a restart.
846    
847    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
848        was not doing so when it should - probably a typo introduced by SVN 528
849        (change 8.10/14).
850    
851    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
852        \w+ when the character tables indicated that \x{c4} was a word character.
853        There were several related cases, all because the tests for doing a table
854        lookup were testing for characters less than 127 instead of 255.
855    
856    27. If a pattern contains capturing parentheses that are not used in a match,
857        their slots in the ovector are set to -1. For those that are higher than
858        any matched groups, this happens at the end of processing. In the case when
859        there were back references that the ovector was too small to contain
860        (causing temporary malloc'd memory to be used during matching), and the
861        highest capturing number was not used, memory off the end of the ovector
862        was incorrectly being set to -1. (It was using the size of the temporary
863        memory instead of the true size.)
864    
865    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
866        ovector size, it uses memory at the end of the block that it has got.
867    
868    29. Check for an overlong MARK name and give an error at compile time. The
869        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
870    
871    30. JIT compiler update.
872    
873    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
874        Rill for the patch.
875    
876    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
877    
878    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
879    
880    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
881        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
882        was enabled.
883    
884    35. Improve JIT code generation for greedy plus quantifier.
885    
886    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
887        1 to "aa" instead of to an empty string. The bug affected repeated groups
888        that could potentially match an empty string.
889    
890    37. Optimizing single character iterators in JIT.
891    
892    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
893        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
894        codepoints that are too big for the mode are faulted, and in a UTF mode,
895        disallowed codepoints are also faulted.
896    
897    39. If PCRE was compiled with UTF support, in three places in the DFA
898        matcher there was code that should only have been obeyed in UTF mode, but
899        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
900        processing when bytes with values greater than 127 were present. In 16-bit
901        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
902        both cases the values are those that cannot be the first data item in a UTF
903        character. The three items that might have provoked this were recursions,
904        possessively repeated groups, and atomic groups.
905    
906    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
907        and pcregrep, because some OS require shared objects to be explicitly
908        passed to ld, causing the link step to fail if they are not.
909    
910    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
911        mode, patterns that started with \h* or \R* might be incorrectly matched.
912    
913    
914    Version 8.30 04-February-2012
915    -----------------------------
916    
917    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
918        name is defined in ctype.h.
919    
920    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
921        only in quite long subpatterns.
922    
923    3.  Removed the function pcre_info(), which has been obsolete and deprecated
924        since it was replaced by pcre_fullinfo() in February 2000.
925    
926    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
927        match a (*MARK), and the match failed at the start of the subject, a
928        reference to memory before the start of the subject could occur. This bug
929        was introduced by fix 17 of release 8.21.
930    
931    5.  A reference to an unset group with zero minimum repetition was giving
932        totally wrong answers (in non-JavaScript-compatibility mode). For example,
933        /(another)?(\1?)test/ matched against "hello world test". This bug was
934        introduced in release 8.13.
935    
936    6.  Add support for 16-bit character strings (a large amount of work involving
937        many changes and refactorings).
938    
939    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
940        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
941        from a file.
942    
943    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
944        rounding is not applied in this particular case).
945    
946    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
947        if they appear, or are escaped, in patterns.
948    
949    10. Get rid of a number of -Wunused-but-set-variable warnings.
950    
951    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
952        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
953        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
954        also returns the mark "x". This bug applied to capturing parentheses,
955        non-capturing parentheses, and atomic parentheses. It also applied to some
956        assertions.
957    
958    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
959        information out of configure.ac instead of relying on pcre.h.generic, which
960        is not stored in the repository.
961    
962    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
963        -lreadline.
964    
965    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
966    
967    15. Applied Graycode's patch to put the top-level frame on the stack rather
968        than the heap when not using the stack for recursion. This gives a
969        performance improvement in many cases when recursion is not deep.
970    
971    16. Experimental code added to "pcretest -C" to output the stack frame size.
972    
973    
974    Version 8.21 12-Dec-2011
975    ------------------------
976    
977    1.  Updating the JIT compiler.
978    
979    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
980        are added as well.
981    
982    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
983        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
984        calling _pcre_jit_exec. Some extra comments are added.
985    
986    4.  (*MARK) settings inside atomic groups that do not contain any capturing
987        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
988        was introduced by change 18 for 8.20.
989    
990    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
991        ECMA-262 standard.
992    
993    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
994        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
995        This bug was probably introduced by change 9 of 8.13.
996    
997    7.  While fixing 6 above, I noticed that a number of other items were being
998        incorrectly rejected as "not fixed length". This arose partly because newer
999        opcodes had not been added to the fixed-length checking code. I have (a)
1000        corrected the bug and added tests for these items, and (b) arranged for an
1001        error to occur if an unknown opcode is encountered while checking for fixed
1002        length instead of just assuming "not fixed length". The items that were
1003        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1004        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1005        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1006    
1007    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1008        being incorrectly compiled and would have given unpredicatble results.
1009    
1010    9.  A possessively repeated subpattern with minimum repeat count greater than
1011        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1012        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1013        the first (A) could occur when it should not.
1014    
1015    10. Add a cast and remove a redundant test from the code.
1016    
1017    11. JIT should use pcre_malloc/pcre_free for allocation.
1018    
1019    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1020        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1021        is anything other than /usr, -L is still shown).
1022    
1023    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1024    
1025    14. Perl does not support \N without a following name in a [] class; PCRE now
1026        also gives an error.
1027    
1028    15. If a forward reference was repeated with an upper limit of around 2000,
1029        it caused the error "internal error: overran compiling workspace". The
1030        maximum number of forward references (including repeats) was limited by the
1031        internal workspace, and dependent on the LINK_SIZE. The code has been
1032        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1033        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1034        of around 200,000 forward references. While doing this, I also speeded up
1035        the filling in of repeated forward references.
1036    
1037    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1038        incorrectly expecting the subject to contain another "a" after the start.
1039    
1040    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1041        in the match, the SKIP should be ignored. This was not happening; instead
1042        the SKIP was being treated as NOMATCH. For patterns such as
1043        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1044        tested.
1045    
1046    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1047        now much more compatible with Perl, in particular in cases where the result
1048        is a non-match for a non-anchored pattern. For example, if
1049        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1050        "m", where previously it did not return a name. A side effect of this
1051        change is that for partial matches, the last encountered mark name is
1052        returned, as for non matches. A number of tests that were previously not
1053        Perl-compatible have been moved into the Perl-compatible test files. The
1054        refactoring has had the pleasing side effect of removing one argument from
1055        the match() function, thus reducing its stack requirements.
1056    
1057    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1058        subsequent uses of /S (without +) incorrectly behaved like /S+.
1059    
1060    21. Retrieve executable code size support for the JIT compiler and fixing
1061        some warnings.
1062    
1063    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1064        not work when the shorter character appeared right at the end of the
1065        subject string.
1066    
1067    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1068        systems.
1069    
1070    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1071        output it when the /M option is used in pcretest.
1072    
1073    25. The CheckMan script was not being included in the distribution. Also, added
1074        an explicit "perl" to run Perl scripts from the PrepareRelease script
1075        because this is reportedly needed in Windows.
1076    
1077    26. If study data was being save in a file and studying had not found a set of
1078        "starts with" bytes for the pattern, the data written to the file (though
1079        never used) was taken from uninitialized memory and so caused valgrind to
1080        complain.
1081    
1082    27. Updated RunTest.bat as provided by Sheri Pierce.
1083    
1084    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1085    
1086    29. Computation of memory usage for the table of capturing group names was
1087        giving an unnecessarily large value.
1088    
1089    
1090    Version 8.20 21-Oct-2011
1091    ------------------------
1092    
1093    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1094        a POSIX class. After further experiments with Perl, which convinced me that
1095        Perl has bugs and confusions, a closing square bracket is no longer allowed
1096        in a POSIX name. This bug also affected patterns with classes that started
1097        with full stops.
1098    
1099    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1100        captured substring, but while checking the failing first alternative,
1101        substring 1 is temporarily captured. If the output vector supplied to
1102        pcre_exec() was not big enough for this capture, the yield of the function
1103        was still zero ("insufficient space for captured substrings"). This cannot
1104        be totally fixed without adding another stack variable, which seems a lot
1105        of expense for a edge case. However, I have improved the situation in cases
1106        such as /(a)(b)x|abc/ matched against "abc", where the return code
1107        indicates that fewer than the maximum number of slots in the ovector have
1108        been set.
1109    
1110    3.  Related to (2) above: when there are more back references in a pattern than
1111        slots in the output vector, pcre_exec() uses temporary memory during
1112        matching, and copies in the captures as far as possible afterwards. It was
1113        using the entire output vector, but this conflicts with the specification
1114        that only 2/3 is used for passing back captured substrings. Now it uses
1115        only the first 2/3, for compatibility. This is, of course, another edge
1116        case.
1117    
1118    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1119        main code base, and can be used by building with --enable-jit. When this is
1120        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1121        runtime --no-jit option is given.
1122    
1123    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1124        ovector, the return from the function was zero, implying that there were
1125        other matches that did not fit. The correct "exactly full" value is now
1126        returned.
1127    
1128    6.  If a subpattern that was called recursively or as a subroutine contained
1129        (*PRUNE) or any other control that caused it to give a non-standard return,
1130        invalid errors such as "Error -26 (nested recursion at the same subject
1131        position)" or even infinite loops could occur.
1132    
1133    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1134        computing the minimum length on reaching *ACCEPT, and so ended up with the
1135        wrong value of 1 rather than 0. Further investigation indicates that
1136        computing a minimum subject length in the presence of *ACCEPT is difficult
1137        (think back references, subroutine calls), and so I have changed the code
1138        so that no minimum is registered for a pattern that contains *ACCEPT.
1139    
1140    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1141        it was not handled as intended. [But see 16 below.]
1142    
1143    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1144        Sheri Pierce.
1145    
1146    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1147        the first byte in a match must be "a".
1148    
1149    11. Change 17 for 8.13 increased the recursion depth for patterns like
1150        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1151        pattern contains any instances of (*THEN). If it does not, the old
1152        optimizations are restored. It would be nice to do this on a per-group
1153        basis, but at the moment that is not feasible.
1154    
1155    12. In some environments, the output of pcretest -C is CRLF terminated. This
1156        broke RunTest's code that checks for the link size. A single white space
1157        character after the value is now allowed for.
1158    
1159    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1160        For "fr", it uses the Windows-specific input and output files.
1161    
1162    14. If (*THEN) appeared in a group that was called recursively or as a
1163        subroutine, it did not work as intended. [But see next item.]
1164    
1165    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1166        pattern fragments (but not containing any | characters). If A and B are
1167        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1168        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1169        D. In other words, Perl considers parentheses that do not contain any |
1170        characters to be part of a surrounding alternative, whereas PCRE was
1171        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1172        differently. PCRE now behaves in the same way as Perl, except in the case
1173        of subroutine/recursion calls such as (?1) which have in any case always
1174        been different (but PCRE had them first :-).
1175    
1176    16. Related to 15 above: Perl does not treat the | in a conditional group as
1177        creating alternatives. Such a group is treated in the same way as an
1178        ordinary group without any | characters when processing (*THEN). PCRE has
1179        been changed to match Perl's behaviour.
1180    
1181    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1182        RunGrepTest script failed.
1183    
1184    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1185        inevitable for groups that contain captures, but it can lead to a lot of
1186        stack use in large patterns. The old behaviour has been restored for atomic
1187        groups that do not contain any capturing parentheses.
1188    
1189    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1190        suppress the check for a minimum subject length at run time. (If it was
1191        given to pcre_exec() or pcre_dfa_exec() it did work.)
1192    
1193    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1194        fail to work when decoding hex characters in data strings in EBCDIC
1195        environments.
1196    
1197    21. It appears that in at least one Mac OS environment, the isxdigit() function
1198        is implemented as a macro that evaluates to its argument more than once,
1199        contravening the C 90 Standard (I haven't checked a later standard). There
1200        was an instance in pcretest which caused it to go wrong when processing
1201        \x{...} escapes in subject strings. The has been rewritten to avoid using
1202        things like p++ in the argument of isxdigit().
1203    
1204    
1205    Version 8.13 16-Aug-2011
1206    ------------------------
1207    
1208    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1209    
1210    2.  Two minor typos in pcre_internal.h have been fixed.
1211    
1212    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1213        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1214        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1215    
1216    4.  There were a number of related bugs in the code for matching backrefences
1217        caselessly in UTF-8 mode when codes for the characters concerned were
1218        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1219        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1220        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1221        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1222        2-byte code at the end of the subject (it thought there wasn't enough data
1223        left).
1224    
1225    5.  Comprehensive information about what went wrong is now returned by
1226        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1227        as the output vector has at least 2 elements. The offset of the start of
1228        the failing character and a reason code are placed in the vector.
1229    
1230    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1231        now returned is for the first byte of the failing character, instead of the
1232        last byte inspected. This is an incompatible change, but I hope it is small
1233        enough not to be a problem. It makes the returned offset consistent with
1234        pcre_exec() and pcre_dfa_exec().
1235    
1236    7.  pcretest now gives a text phrase as well as the error number when
1237        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1238        failure, the offset and reason code are output.
1239    
1240    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1241        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1242        back over a single character (\n). This seems wrong (because it treated the
1243        two characters as a single entity when going forwards), conflicts with the
1244        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1245        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1246        has been changed.
1247    
1248    9.  Some internal refactoring has changed the processing so that the handling
1249        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1250        time (the PCRE_DOTALL option was changed this way some time ago: version
1251        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1252        which was always a bit of a fudge. It also means that there is one less
1253        argument for the match() function, which reduces its stack requirements
1254        slightly. This change also fixes an incompatibility with Perl: the pattern
1255        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1256    
1257    10. More internal refactoring has drastically reduced the number of recursive
1258        calls to match() for possessively repeated groups such as (abc)++ when
1259        using pcre_exec().
1260    
1261    11. While implementing 10, a number of bugs in the handling of groups were
1262        discovered and fixed:
1263    
1264        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1265        (a|)*(?1) gave a compile-time internal error.
1266        ((a|)+)+  did not notice that the outer group could match an empty string.
1267        (^a|^)+   was not marked as anchored.
1268        (.*a|.*)+ was not marked as matching at start or after a newline.
1269    
1270    12. Yet more internal refactoring has removed another argument from the match()
1271        function. Special calls to this function are now indicated by setting a
1272        value in a variable in the "match data" data block.
1273    
1274    13. Be more explicit in pcre_study() instead of relying on "default" for
1275        opcodes that mean there is no starting character; this means that when new
1276        ones are added and accidentally left out of pcre_study(), testing should
1277        pick them up.
1278    
1279    14. The -s option of pcretest has been documented for ages as being an old
1280        synonym of -m (show memory usage). I have changed it to mean "force study
1281        for every regex", that is, assume /S for every regex. This is similar to -i
1282        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1283        using it. It makes it easier to run collections of tests with and without
1284        study enabled, and thereby test pcre_study() more easily. All the standard
1285        tests are now run with and without -s (but some patterns can be marked as
1286        "never study" - see 20 below).
1287    
1288    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1289        restoration of the capturing data to the outer values was not happening
1290        correctly.
1291    
1292    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1293        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1294        pattern had matched an empty string, and so incorrectly returned a no
1295        match.
1296    
1297    17. There was optimizing code for the last branch of non-capturing parentheses,
1298        and also for the obeyed branch of a conditional subexpression, which used
1299        tail recursion to cut down on stack usage. Unfortunately, now that there is
1300        the possibility of (*THEN) occurring in these branches, tail recursion is
1301        no longer possible because the return has to be checked for (*THEN). These
1302        two optimizations have therefore been removed. [But see 8.20/11 above.]
1303    
1304    18. If a pattern containing \R was studied, it was assumed that \R always
1305        matched two bytes, thus causing the minimum subject length to be
1306        incorrectly computed because \R can also match just one byte.
1307    
1308    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1309        was incorrectly computed.
1310    
1311    20. If /S is present twice on a test pattern in pcretest input, it now
1312        *disables* studying, thereby overriding the use of -s on the command line
1313        (see 14 above). This is necessary for one or two tests to keep the output
1314        identical in both cases.
1315    
1316    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1317        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1318    
1319    22. When an atomic group that contained a capturing parenthesis was
1320        successfully matched, but the branch in which it appeared failed, the
1321        capturing was not being forgotten if a higher numbered group was later
1322        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1323        group 1 to "a", when in fact it should be unset. This applied to multi-
1324        branched capturing and non-capturing groups, repeated or not, and also to
1325        positive assertions (capturing in negative assertions does not happen
1326        in PCRE) and also to nested atomic groups.
1327    
1328    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1329        subject after a captured substring, to make it easier to tell which of a
1330        number of identical substrings has been captured.
1331    
1332    24. The way atomic groups are processed by pcre_exec() has been changed so that
1333        if they are repeated, backtracking one repetition now resets captured
1334        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1335        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1336        "aaa". Previously, it would have been "a". As part of this code
1337        refactoring, the way recursive calls are handled has also been changed.
1338    
1339    25. If an assertion condition captured any substrings, they were not passed
1340        back unless some other capturing happened later. For example, if
1341        (?(?=(a))a) was matched against "a", no capturing was returned.
1342    
1343    26. When studying a pattern that contained subroutine calls or assertions,
1344        the code for finding the minimum length of a possible match was handling
1345        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1346        group 1 called group 2 while simultaneously a separate group 2 called group
1347        1). A stack overflow occurred in this case. I have fixed this by limiting
1348        the recursion depth to 10.
1349    
1350    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1351        Fortmann. This supports explicit test numbers on the command line, and has
1352        argument validation and error reporting.
1353    
1354    28. An instance of \X with an unlimited repeat could fail if at any point the
1355        first character it looked at was a mark character.
1356    
1357    29. Some minor code refactoring concerning Unicode properties and scripts
1358        should reduce the stack requirement of match() slightly.
1359    
1360    30. Added the '=' option to pcretest to check the setting of unused capturing
1361        slots at the end of the pattern, which are documented as being -1, but are
1362        not included in the return count.
1363    
1364    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1365        compiled something random. Now it gives a compile-time error (as does
1366        Perl).
1367    
1368    32. A *MARK encountered during the processing of a positive assertion is now
1369        recorded and passed back (compatible with Perl).
1370    
1371    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1372        had alternative anchored branches, the search for a second match in a line
1373        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1374        matched the line "0102" twice. The same bug affected patterns that started
1375        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1376        twice.
1377    
1378    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1379        does, and because of capturing effects, quantifying parenthesized
1380        assertions may at times be useful. Quantifiers are now allowed for
1381        parenthesized assertions.
1382    
1383    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1384    
1385    36. \g was being checked for fancy things in a character class, when it should
1386        just be a literal "g".
1387    
1388    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1389        appearance of a nested POSIX class supersedes an apparent external class.
1390        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1391        unescaped square brackets may also appear as part of class names. For
1392        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1393        more like Perl. (But see 8.20/1 above.)
1394    
1395    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1396        was because it thought it was \N{name}, which is not supported).
1397    
1398    39. Add minix to OS list not supporting the -S option in pcretest.
1399    
1400    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1401        cannot analyze patterns in sufficient detail to catch mutual recursions
1402        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1403        subgroup is called recursively as a subpattern for a second time at the
1404        same position in the subject string. In previous releases this might have
1405        been caught by the recursion limit, or it might have run out of stack.
1406    
1407    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1408        happen only once. PCRE was, however incorrectly giving a compile time error
1409        "recursive call could loop indefinitely" because it cannot analyze the
1410        pattern in sufficient detail. The compile time test no longer happens when
1411        PCRE is compiling a conditional subpattern, but actual runaway loops are
1412        now caught at runtime (see 40 above).
1413    
1414    42. It seems that Perl allows any characters other than a closing parenthesis
1415        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1416        has been changed to be the same.
1417    
1418    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1419        as not to get warnings when autogen.sh is called. Also changed
1420        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1421    
1422    44. To help people who use pcregrep to scan files containing exceedingly long
1423        lines, the following changes have been made:
1424    
1425        (a) The default value of the buffer size parameter has been increased from
1426            8K to 20K. (The actual buffer used is three times this size.)
1427    
1428        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1429            PCRE is built.
1430    
1431        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1432            to be set at run time.
1433    
1434        (d) Numerical values in pcregrep options can be followed by K or M, for
1435            example --buffer-size=50K.
1436    
1437        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1438            given and the return code is set to 2.
1439    
1440    45. Add a pointer to the latest mark to the callout data block.
1441    
1442    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1443        partial match of an empty string instead of no match. This was specific to
1444        the use of ".".
1445    
1446    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1447        complete match instead of a partial match. This bug was dependent on both
1448        the PCRE_UTF8 and PCRE_DOTALL options being set.
1449    
1450    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1451        starting byte set, because \b was not being ignored.
1452    
1453    
1454    Version 8.12 15-Jan-2011
1455    ------------------------
1456    
1457    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1458        checks for such things as part of the documentation building process.
1459    
1460    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1461        --match-limit and --recursion-limit options (added for 8.11). In
1462        particular, this made one of the standard tests fail. (The integer value
1463        went into the wrong half of a long int.)
1464    
1465    3.  If the --colour option was given to pcregrep with -v (invert match), it
1466        did strange things, either producing crazy output, or crashing. It should,
1467        of course, ignore a request for colour when reporting lines that do not
1468        match.
1469    
1470    4.  Another pcregrep bug caused similar problems if --colour was specified with
1471        -M (multiline) and the pattern match finished with a line ending.
1472    
1473    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1474        matched in multiline mode, the following line was shown as part of the
1475        match. This seems wrong, so I have changed it.
1476    
1477    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1478        the check for further matches in the same line (so they could be coloured)
1479        to overrun the end of the current line. If another match was found, it was
1480        incorrectly shown (and then shown again when found in the next line).
1481    
1482    7.  If pcregrep was compiled under Windows, there was a reference to the
1483        function pcregrep_exit() before it was defined. I am assuming this was
1484        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1485        reported by a user. I've moved the definition above the reference.
1486    
1487    
1488    Version 8.11 10-Dec-2010
1489    ------------------------
1490    
1491    1.  (*THEN) was not working properly if there were untried alternatives prior
1492        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1493        backtracked to try for "b" instead of moving to the next alternative branch
1494        at the same level (in this case, to look for "c"). The Perl documentation
1495        is clear that when (*THEN) is backtracked onto, it goes to the "next
1496        alternative in the innermost enclosing group".
1497    
1498    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1499        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1500        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1501        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1502        (*THEN).
1503    
1504    3.  If \s appeared in a character class, it removed the VT character from
1505        the class, even if it had been included by some previous item, for example
1506        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1507        of \s, but is part of the POSIX "space" class.)
1508    
1509    4.  A partial match never returns an empty string (because you can always
1510        match an empty string at the end of the subject); however the checking for
1511        an empty string was starting at the "start of match" point. This has been
1512        changed to the "earliest inspected character" point, because the returned
1513        data for a partial match starts at this character. This means that, for
1514        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1515        (previously it gave "no match").
1516    
1517    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1518        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1519        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1520        has an implication that the given string is incomplete (because a partial
1521        match is preferred over a full match). For this reason, these items now
1522        give a partial match in this situation. [Aside: previously, the one case
1523        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1524        match rather than a full match, which was wrong by the old rules, but is
1525        now correct.]
1526    
1527    6.  There was a bug in the handling of #-introduced comments, recognized when
1528        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1529        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1530        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1531        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1532        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1533        places in pcre_compile().
1534    
1535    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1536        comments when looking ahead for named forward references to subpatterns,
1537        the only newline sequence it recognized was NL. It now handles newlines
1538        according to the set newline convention.
1539    
1540    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1541        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1542        cater for a lack of strerror(). These oversights have been fixed.
1543    
1544    9.  Added --match-limit and --recursion-limit to pcregrep.
1545    
1546    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1547    
1548    11. When the -o option was used, pcregrep was setting a return code of 1, even
1549        when matches were found, and --line-buffered was not being honoured.
1550    
1551    12. Added an optional parentheses number to the -o and --only-matching options
1552        of pcregrep.
1553    
1554    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1555        can match an empty string. The code to do it in pcretest and pcredemo
1556        needed fixing:
1557    
1558        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1559            only one byte after an empty string match just before CRLF (this case
1560            just got forgotten; "any" and "anycrlf" were OK).
1561    
1562        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1563            mode when an empty string match preceded an ASCII character followed by
1564            a non-ASCII character. (The code for advancing by one character rather
1565            than one byte was nonsense.)
1566    
1567        (c) The pcredemo.c sample program did not have any code at all to handle
1568            the cases when CRLF is a valid newline sequence.
1569    
1570    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1571        as a starting offset was within the subject string. There is now a new
1572        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1573        negative or greater than the length of the string. In order to test this,
1574        pcretest is extended to allow the setting of negative starting offsets.
1575    
1576    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1577        starting offset points to the beginning of a UTF-8 character was
1578        unnecessarily clumsy. I tidied it up.
1579    
1580    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1581        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1582    
1583    17. Nobody had reported that the --include_dir option, which was added in
1584        release 7.7 should have been called --include-dir (hyphen, not underscore)
1585        for compatibility with GNU grep. I have changed it to --include-dir, but
1586        left --include_dir as an undocumented synonym, and the same for
1587        --exclude-dir, though that is not available in GNU grep, at least as of
1588        release 2.5.4.
1589    
1590    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1591        characters from a string of bytes) have been redefined so as not to use
1592        loops, in order to improve performance in some environments. At the same
1593        time, I abstracted some of the common code into auxiliary macros to save
1594        repetition (this should not affect the compiled code).
1595    
1596    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1597        compile-time error is now given if \c is not followed by an ASCII
1598        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1599        different, and any byte value is allowed.)
1600    
1601    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1602        START_OPTIMIZE option, which is now allowed at compile time - but just
1603        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1604        to pcregrep and other applications that have no direct access to PCRE
1605        options. The new /Y option in pcretest sets this option when calling
1606        pcre_compile().
1607    
1608    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1609        back references. Groups containing recursive back references were forced to
1610        be atomic by that change, but in the case of named groups, the amount of
1611        memory required was incorrectly computed, leading to "Failed: internal
1612        error: code overflow". This has been fixed.
1613    
1614    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1615        pcretest.c, to avoid build problems in some Borland environments.
1616    
1617    
1618    Version 8.10 25-Jun-2010
1619    ------------------------
1620    
1621    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1622        THEN.
1623    
1624    2.  (*ACCEPT) was not working when inside an atomic group.
1625    
1626    3.  Inside a character class, \B is treated as a literal by default, but
1627        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1628        causes the error). The code is unchanged, but I tidied the documentation.
1629    
1630    4.  Inside a character class, PCRE always treated \R and \X as literals,
1631        whereas Perl faults them if its -w option is set. I have changed PCRE so
1632        that it faults them when PCRE_EXTRA is set.
1633    
1634    5.  Added support for \N, which always matches any character other than
1635        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1636    
1637    6.  When compiling pcregrep with newer versions of gcc which may have
1638        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1639        declared with attribute warn_unused_result" were given. Just casting the
1640        result to (void) does not stop the warnings; a more elaborate fudge is
1641        needed. I've used a macro to implement this.
1642    
1643    7.  Minor change to pcretest.c to avoid a compiler warning.
1644    
1645    8.  Added four artifical Unicode properties to help with an option to make
1646        \s etc use properties (see next item). The new properties are: Xan
1647        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1648    
1649    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1650        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1651        this option. Modified pcretest to add /W to test this facility. Added
1652        REG_UCP to make it available via the POSIX interface.
1653    
1654    10. Added --line-buffered to pcregrep.
1655    
1656    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1657        studied, and the match started with a letter with a code point greater than
1658        127 whose first byte was different to the first byte of the other case of
1659        the letter, the other case of this starting letter was not recognized
1660        (#976).
1661    
1662    12. If a pattern that was studied started with a repeated Unicode property
1663        test, for example, \p{Nd}+, there was the theoretical possibility of
1664        setting up an incorrect bitmap of starting bytes, but fortunately it could
1665        not have actually happened in practice until change 8 above was made (it
1666        added property types that matched character-matching opcodes).
1667    
1668    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1669        possible starting bytes for non-anchored patterns.
1670    
1671    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1672        \R, and also a number of cases that involve Unicode properties, both
1673        explicit and implicit when PCRE_UCP is set.
1674    
1675    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1676        input, it could crash or give wrong results if characters with values
1677        greater than 0xc0 were present in the subject string. (Detail: it assumed
1678        UTF-8 input when processing these items.)
1679    
1680    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1681        size_t is 64-bit (#991).
1682    
1683    17. Added a check for running out of memory when PCRE is compiled with
1684        --disable-stack-for-recursion (#990).
1685    
1686    18. If the last data line in a file for pcretest does not have a newline on
1687        the end, a newline was missing in the output.
1688    
1689    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1690        less than 128) in its various bitmaps. However, there is a facility for
1691        generating tables according to the current locale when PCRE is compiled. It
1692        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1693        characters, are recognized by isspace() and therefore were getting set in
1694        these tables, and indeed these tables seem to approximate to ISO 8859. This
1695        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1696        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1697        which of course cannot start UTF-8 characters. I have changed the code so
1698        that only real ASCII characters (less than 128) and the correct starting
1699        bytes for UTF-8 encodings are set for characters greater than 127 when in
1700        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1701        altogether.)
1702    
1703    20. Added the /T option to pcretest so as to be able to run tests with non-
1704        standard character tables, thus making it possible to include the tests
1705        used for 19 above in the standard set of tests.
1706    
1707    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1708        reference to a subpattern the other side of a comment that contains an
1709        opening parenthesis caused either an internal compiling error, or a
1710        reference to the wrong subpattern.
1711    
1712    
1713    Version 8.02 19-Mar-2010
1714    ------------------------
1715    
1716    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1717    
1718    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1719        configured.
1720    
1721    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1722        original author of that file, following a query about its status.
1723    
1724    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1725        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1726    
1727    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1728        quantifier applied to a forward-referencing subroutine call, could compile
1729        incorrect code or give the error "internal error: previously-checked
1730        referenced subpattern not found".
1731    
1732    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1733        variables to point to external functions. For these systems, therefore,
1734        pcre_malloc etc. are now initialized to local functions that call the
1735        relevant global functions.
1736    
1737    7.  There were two entries missing in the vectors called coptable and poptable
1738        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1739        I've fixed the data, and added a kludgy way of testing at compile time that
1740        the lengths are correct (equal to the number of opcodes).
1741    
1742    8.  Following on from 7, I added a similar kludge to check the length of the
1743        eint vector in pcreposix.c.
1744    
1745    9.  Error texts for pcre_compile() are held as one long string to avoid too
1746        much relocation at load time. To find a text, the string is searched,
1747        counting zeros. There was no check for running off the end of the string,
1748        which could happen if a new error number was added without updating the
1749        string.
1750    
1751    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1752    
1753    11. \K was not working if it appeared in an atomic group or in a group that
1754        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1755        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1756        the assertion is positive, but not if it is negative.
1757    
1758    12. Change 11 fortuitously reduced the size of the stack frame used in the
1759        "match()" function of pcre_exec.c by one pointer. Forthcoming
1760        implementation of support for (*MARK) will need an extra pointer on the
1761        stack; I have reserved it now, so that the stack frame size does not
1762        decrease.
1763    
1764    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1765        item in branch that calls a recursion is a subroutine call - as in the
1766        second branch in the above example - was incorrectly given the compile-
1767        time error "recursive call could loop indefinitely" because pcre_compile()
1768        was not correctly checking the subroutine for matching a non-empty string.
1769    
1770    14. The checks for overrunning compiling workspace could trigger after an
1771        overrun had occurred. This is a "should never occur" error, but it can be
1772        triggered by pathological patterns such as hundreds of nested parentheses.
1773        The checks now trigger 100 bytes before the end of the workspace.
1774    
1775    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1776    
1777    
1778    Version 8.01 19-Jan-2010
1779    ------------------------
1780    
1781    1.  If a pattern contained a conditional subpattern with only one branch (in
1782        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1783        computed the wrong minimum data length (which is of course zero for such
1784        subpatterns). This could cause incorrect "no match" results.
1785    
1786    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1787        the pattern is reset in the first branch, pcre_compile() failed with
1788        "internal error: code overflow at offset...". This happened only when
1789        the reset was to the original external option setting. (An optimization
1790        abstracts leading options settings into an external setting, which was the
1791        cause of this.)
1792    
1793    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1794        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1795        assertion pattern did not match (meaning that the assertion was true), it
1796        was incorrectly treated as false if the SKIP had been reached during the
1797        matching. This also applied to assertions used as conditions.
1798    
1799    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1800        assertion subpattern, including such a pattern used as a condition,
1801        unpredictable results occurred, instead of the error return
1802        PCRE_ERROR_DFA_UITEM.
1803    
1804    5.  The C++ GlobalReplace function was not working like Perl for the special
1805        situation when an empty string is matched. It now does the fancy magic
1806        stuff that is necessary.
1807    
1808    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1809        removed. (These were left over from very, very early versions of PCRE.)
1810    
1811    7.  Some cosmetic changes to the code to make life easier when compiling it
1812        as part of something else:
1813    
1814        (a) Change DEBUG to PCRE_DEBUG.
1815    
1816        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1817            called "current" as "current_branch", to prevent a collision with the
1818            Linux macro when compiled as a kernel module.
1819    
1820        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1821            prevent a collision with the Linux macro when compiled as a kernel
1822            module.
1823    
1824    8.  In pcre_compile() there are some checks for integer overflows that used to
1825        cast potentially large values to (double). This has been changed to that
1826        when building, a check for int64_t is made, and if it is found, it is used
1827        instead, thus avoiding the use of floating point arithmetic. (There is no
1828        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1829        double.
1830    
1831    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1832        2005 (difference between two addresses compared to an unsigned value).
1833    
1834    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1835        custom one, because of the following reported problem in Windows:
1836    
1837          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1838              under Win32.
1839          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1840              therefore missing the function definition.
1841          - The compiler thus generates a "C" signature for the test function.
1842          - The linker fails to find the "C" function.
1843          - PCRE fails to configure if asked to do so against libbz2.
1844    
1845    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1846        messages were output:
1847    
1848          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1849          rerunning libtoolize, to keep the correct libtool macros in-tree.
1850          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1851    
1852        I have done both of these things.
1853    
1854    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1855        most of the time, it *can* run out if it is given a pattern that contains a
1856        runaway infinite recursion. I updated the discussion in the pcrestack man
1857        page.
1858    
1859    13. Now that we have gone to the x.xx style of version numbers, the minor
1860        version may start with zero. Using 08 or 09 is a bad idea because users
1861        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1862        interpreted as invalid octal numbers. I've updated the previous comment in
1863        configure.ac, and also added a check that gives an error if 08 or 09 are
1864        used.
1865    
1866    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1867        causing partial matching to fail when the end of the subject matched \W
1868        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1869    
1870    15. There were some discrepancies between the declarations in pcre_internal.h
1871        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1872        their definitions. The declarations used "const uschar *" and the
1873        definitions used USPTR. Even though USPTR is normally defined as "const
1874        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1875        reported that: "This difference in casting confuses some C++ compilers, for
1876        example, SunCC recognizes above declarations as different functions and
1877        generates broken code for hbpcre." I have changed the declarations to use
1878        USPTR.
1879    
1880    16. GNU libtool is named differently on some systems. The autogen.sh script now
1881        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1882        (FreeBSD).
1883    
1884    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1885        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1886        comment: "Figure out how to create a longlong from a string: strtoll and
1887        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1888        instance, but it only takes 2 args instead of 3!"
1889    
1890    18. A subtle bug concerned with back references has been fixed by a change of
1891        specification, with a corresponding code fix. A pattern such as
1892        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1893        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1894        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1895        same bug. Such groups have to be quantified to be useful, or contained
1896        inside another quantified group. (If there's no repetition, the reference
1897        can never match.) The problem arises because, having left the group and
1898        moved on to the rest of the pattern, a later failure that backtracks into
1899        the group uses the captured value from the final iteration of the group
1900        rather than the correct earlier one. I have fixed this in PCRE by forcing
1901        any group that contains a reference to itself to be an atomic group; that
1902        is, there cannot be any backtracking into it once it has completed. This is
1903        similar to recursive and subroutine calls.
1904    
1905    
1906    Version 8.00 19-Oct-09
1907    ----------------------
1908    
1909    1.  The table for translating pcre_compile() error codes into POSIX error codes
1910        was out-of-date, and there was no check on the pcre_compile() error code
1911        being within the table. This could lead to an OK return being given in
1912        error.
1913    
1914    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1915        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1916        in a Windows environment.
1917    
1918    3.  The pcregrep --count option prints the count for each file even when it is
1919        zero, as does GNU grep. However, pcregrep was also printing all files when
1920        --files-with-matches was added. Now, when both options are given, it prints
1921        counts only for those files that have at least one match. (GNU grep just
1922        prints the file name in this circumstance, but including the count seems
1923        more useful - otherwise, why use --count?) Also ensured that the
1924        combination -clh just lists non-zero counts, with no names.
1925    
1926    4.  The long form of the pcregrep -F option was incorrectly implemented as
1927        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1928        but it seems right to fix it, and I didn't think it was worth preserving
1929        the old behaviour.
1930    
1931    5.  The command line items --regex=pattern and --regexp=pattern were not
1932        recognized by pcregrep, which required --regex pattern or --regexp pattern
1933        (with a space rather than an '='). The man page documented the '=' forms,
1934        which are compatible with GNU grep; these now work.
1935    
1936    6.  No libpcreposix.pc file was created for pkg-config; there was just
1937        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1938    
1939    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1940        when UCP support is not needed, by modifying the Python script that
1941        generates it from Unicode data files. This should not matter if the module
1942        is correctly used as a library, but I received one complaint about 50K of
1943        unwanted data. My guess is that the person linked everything into his
1944        program rather than using a library. Anyway, it does no harm.
1945    
1946    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1947        was a minimum greater than 1 for a wide character in a possessive
1948        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1949        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1950        character. Chaos in the form of incorrect output or a compiling loop could
1951        result.
1952    
1953    9.  The restrictions on what a pattern can contain when partial matching is
1954        requested for pcre_exec() have been removed. All patterns can now be
1955        partially matched by this function. In addition, if there are at least two
1956        slots in the offset vector, the offset of the earliest inspected character
1957        for the match and the offset of the end of the subject are set in them when
1958        PCRE_ERROR_PARTIAL is returned.
1959    
1960    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1961        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1962        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1963        and may be more useful for multi-segment matching.
1964    
1965    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1966        used to be given if ever the end of the subject was reached; now it is
1967        given only if matching could not proceed because another character was
1968        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1969        string "Z", which now yields "no match" instead of "partial match". In the
1970        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1971        final character ended with (*FAIL).
1972    
1973    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1974        if the pattern had a "must contain" character that was already found in the
1975        earlier partial match, unless partial matching was again requested. For
1976        example, with the pattern /dog.(body)?/, the "must contain" character is
1977        "g". If the first part-match was for the string "dog", restarting with
1978        "sbody" failed. This bug has been fixed.
1979    
1980    13. The string returned by pcre_dfa_exec() after a partial match has been
1981        changed so that it starts at the first inspected character rather than the
1982        first character of the match. This makes a difference only if the pattern
1983        starts with a lookbehind assertion or \b or \B (\K is not supported by
1984        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1985        matching functions compatible, and I think it's the right thing to do.
1986    
1987    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1988        so that the demonstration program is easily available in environments where
1989        PCRE has not been installed from source.
1990    
1991    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1992        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1993        library.
1994    
1995    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1996        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1997        is not the first non-POSIX option to be added. Clearly some people find
1998        these options useful.
1999    
2000    17. If a caller to the POSIX matching function regexec() passes a non-zero
2001        value for nmatch with a NULL value for pmatch, the value of
2002        nmatch is forced to zero.
2003    
2004    18. RunGrepTest did not have a test for the availability of the -u option of
2005        the diff command, as RunTest does. It now checks in the same way as
2006        RunTest, and also checks for the -b option.
2007    
2008    19. If an odd number of negated classes containing just a single character
2009        interposed, within parentheses, between a forward reference to a named
2010        subpattern and the definition of the subpattern, compilation crashed with
2011        an internal error, complaining that it could not find the referenced
2012        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
2013        [The bug was that it was starting one character too far in when skipping
2014        over the character class, thus treating the ] as data rather than
2015        terminating the class. This meant it could skip too much.]
2016    
2017    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
2018        /g option in pcretest when the pattern contains \K, which makes it possible
2019        to have an empty string match not at the start, even when the pattern is
2020        anchored. Updated pcretest and pcredemo to use this option.
2021    
2022    21. If the maximum number of capturing subpatterns in a recursion was greater
2023        than the maximum at the outer level, the higher number was returned, but
2024        with unset values at the outer level. The correct (outer level) value is
2025        now given.
2026    
2027    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
2028        PCRE did not set those parentheses (unlike Perl). I have now found a way to
2029        make it do so. The string so far is captured, making this feature
2030        compatible with Perl.
2031    
2032    23. The tests have been re-organized, adding tests 11 and 12, to make it
2033        possible to check the Perl 5.10 features against Perl 5.10.
2034    
2035    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
2036        pattern matches a fixed length string. PCRE did not allow this; now it
2037        does. Neither allows recursion.
2038    
2039    25. I finally figured out how to implement a request to provide the minimum
2040        length of subject string that was needed in order to match a given pattern.
2041        (It was back references and recursion that I had previously got hung up
2042        on.) This code has now been added to pcre_study(); it finds a lower bound
2043        to the length of subject needed. It is not necessarily the greatest lower
2044        bound, but using it to avoid searching strings that are too short does give
2045        some useful speed-ups. The value is available to calling programs via
2046        pcre_fullinfo().
2047    
2048    26. While implementing 25, I discovered to my embarrassment that pcretest had
2049        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
2050        study optimizations had never been tested with that matching function.
2051        Oops. What is worse, even when it was passed study data, there was a bug in
2052        pcre_dfa_exec() that meant it never actually used it. Double oops. There
2053        were also very few tests of studied patterns with pcre_dfa_exec().
2054    
2055    27. If (?| is used to create subpatterns with duplicate numbers, they are now
2056        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
2057        on the other side of the coin, they are no longer allowed to have different
2058        names, because these cannot be distinguished in PCRE, and this has caused
2059        confusion. (This is a difference from Perl.)
2060    
2061    28. When duplicate subpattern names are present (necessarily with different
2062        numbers, as required by 27 above), and a test is made by name in a
2063        conditional pattern, either for a subpattern having been matched, or for
2064        recursion in such a pattern, all the associated numbered subpatterns are
2065        tested, and the overall condition is true if the condition is true for any
2066        one of them. This is the way Perl works, and is also more like the way
2067        testing by number works.
2068    
2069    
2070    Version 7.9 11-Apr-09
2071    ---------------------
2072    
2073    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
2074        (pcretest), all targets were linked against these libraries. This included
2075        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
2076        libraries. This caused unwanted dependencies to be created. This problem
2077        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
2078        pcretest is linked with readline.
2079    
2080    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
2081        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
2082        moved outside it again, because FALSE and TRUE are already defined in AIX,
2083        but BOOL is not.
2084    
2085    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
2086        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
2087    
2088    4.  The pcregrep documentation said spaces were inserted as well as colons (or
2089        hyphens) following file names and line numbers when outputting matching
2090        lines. This is not true; no spaces are inserted. I have also clarified the
2091        wording for the --colour (or --color) option.
2092    
2093    5.  In pcregrep, when --colour was used with -o, the list of matching strings
2094        was not coloured; this is different to GNU grep, so I have changed it to be
2095        the same.
2096    
2097    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
2098        each matching line was coloured. Now it goes on to look for further matches
2099        of any of the test patterns, which is the same behaviour as GNU grep.
2100    
2101    7.  A pattern that could match an empty string could cause pcregrep to loop; it
2102        doesn't make sense to accept an empty string match in pcregrep, so I have
2103        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2104        seems to be how GNU grep behaves. [But see later change 40 for release
2105        8.33.]
2106    
2107    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2108        start or after a newline", because the conditional assertion was not being
2109        correctly handled. The rule now is that both the assertion and what follows
2110        in the first alternative must satisfy the test.
2111    
2112    9.  If auto-callout was enabled in a pattern with a conditional group whose
2113        condition was an assertion, PCRE could crash during matching, both with
2114        pcre_exec() and pcre_dfa_exec().
2115    
2116    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2117        used for matching.
2118    
2119    11. Unicode property support in character classes was not working for
2120        characters (bytes) greater than 127 when not in UTF-8 mode.
2121    
2122    12. Added the -M command line option to pcretest.
2123    
2124    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2125    
2126    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2127    
2128    16. Added comments and documentation about mis-use of no_arg in the C++
2129        wrapper.
2130    
2131    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2132        from Martin Jerabek that uses macro names for all relevant character and
2133        string constants.
2134    
2135    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2136        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2137        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2138        these, but not everybody uses configure.
2139    
2140    19. A conditional group that had only one branch was not being correctly
2141        recognized as an item that could match an empty string. This meant that an
2142        enclosing group might also not be so recognized, causing infinite looping
2143        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2144        with the subject "ab", where knowledge that the repeated group can match
2145        nothing is needed in order to break the loop.
2146    
2147    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2148        exec(), but without supplying a callout function, matching went wrong.
2149    
2150    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2151        leak if the size of the offset vector was greater than 30. When the vector
2152        is smaller, the saved offsets during recursion go onto a local stack
2153        vector, but for larger vectors malloc() is used. It was failing to free
2154        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2155        error, in fact).
2156    
2157    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2158        heapframe that is used only when UTF-8 support is enabled. This caused no
2159        problem, but was untidy.
2160    
2161    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2162        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2163        included within another project.
2164    
2165    24. Steven Van Ingelgem's patches to add more options to the CMake support,
2166        slightly modified by me:
2167    
2168          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2169              not building pcregrep.
2170    
2171          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2172              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2173    
2174    25. Forward references, both numeric and by name, in patterns that made use of
2175        duplicate group numbers, could behave incorrectly or give incorrect errors,
2176        because when scanning forward to find the reference group, PCRE was not
2177        taking into account the duplicate group numbers. A pattern such as
2178        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2179    
2180    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2181        the feature of a custom pointer more persuasive (as requested by a user).
2182    
2183    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2184        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2185        pre-defined.
2186    
2187    28. Added support for (*UTF8) at the start of a pattern.
2188    
2189    29. Arrange for flags added by the "release type" setting in CMake to be shown
2190        in the configuration summary.
2191    
2192    
2193    Version 7.8 05-Sep-08
2194    ---------------------
2195    
2196    1.  Replaced UCP searching code with optimized version as implemented for Ad
2197        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2198        stage table and inline lookup instead of a function, giving speed ups of 2
2199        to 5 times on some simple patterns that I tested. Permission was given to
2200        distribute the MultiStage2.py script that generates the tables (it's not in
2201        the tarball, but is in the Subversion repository).
2202    
2203    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2204        scripts.
2205    
2206    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2207        a group with a zero qualifier. The result of the study could be incorrect,
2208        or the function might crash, depending on the pattern.
2209    
2210    4.  Caseless matching was not working for non-ASCII characters in back
2211        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2212        It now works when Unicode Property Support is available.
2213    
2214    5.  In pcretest, an escape such as \x{de} in the data was always generating
2215        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2216        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2217        truncation.
2218    
2219    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2220    
2221    7.  Added two (int) casts to pcregrep when printing the difference of two
2222        pointers, in case they are 64-bit values.
2223    
2224    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2225        test 2 if it fails.
2226    
2227    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2228        and a #define of that name to empty if it is not externally set. This is to
2229        allow users of MSVC to set it if necessary.
2230    
2231    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2232        the convenience functions in the pcre_get.c source file.
2233    
2234    11. An option change at the start of a pattern that had top-level alternatives
2235        could cause overwriting and/or a crash. This command provoked a crash in
2236        some environments:
2237    
2238          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2239    
2240        This potential security problem was recorded as CVE-2008-2371.
2241    
2242    12. For a pattern where the match had to start at the beginning or immediately
2243        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2244        pcre_dfa_exec() could read past the end of the passed subject if there was
2245        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2246        pcretest so that it places the subject at the end of its malloc-ed buffer.
2247    
2248    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2249        exec() might read past the end of the data buffer in UTF-8 mode.
2250    
2251    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2252        the data contained the byte 0x85 as part of a UTF-8 character within its
2253        first line. This applied both to normal and DFA matching.
2254    
2255    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2256        /^[^d]*?$/8 failed to match "abc".
2257    
2258    16. Added a missing copyright notice to pcrecpp_internal.h.
2259    
2260    17. Make it more clear in the documentation that values returned from
2261        pcre_exec() in ovector are byte offsets, not character counts.
2262    
2263    18. Tidied a few places to stop certain compilers from issuing warnings.
2264    
2265    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2266        supplied by Stefan Weber. I made a further small update for 7.8 because
2267        there is a change of source arrangements: the pcre_searchfuncs.c module is
2268        replaced by pcre_ucd.c.
2269    
2270    
2271    Version 7.7 07-May-08
2272    ---------------------
2273    
2274    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2275        a string to a long long, pretend we don't even have a long long." This is
2276        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2277    
2278    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2279        pre-7.6 versions, which defined a global no_arg variable instead of putting
2280        it in the RE class. (See also #8 below.)
2281    
2282    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2283        Lopes.
2284    
2285    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2286    
2287        (1) The include/exclude patterns were being applied to the whole pathnames
2288            of files, instead of just to the final components.
2289    
2290        (2) If there was more than one level of directory, the subdirectories were
2291            skipped unless they satisfied the include/exclude conditions. This is
2292            inconsistent with GNU grep (and could even be seen as contrary to the
2293            pcregrep specification - which I improved to make it absolutely clear).
2294            The action now is always to scan all levels of directory, and just
2295            apply the include/exclude patterns to regular files.
2296    
2297    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2298        --exclude_dir in the tests to avoid scanning .svn directories.
2299    
2300    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2301        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2302        doesn't support NULs in patterns.
2303    
2304    7.  Added some missing "const"s to declarations of static tables in
2305        pcre_compile.c and pcre_dfa_exec.c.
2306    
2307    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2308        caused by fix #2  above. (Subsequently also a second patch to fix the
2309        first patch. And a third patch - this was a messy problem.)
2310    
2311    9.  Applied Craig's patch to remove the use of push_back().
2312    
2313    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2314        matching function regexec().
2315    
2316    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2317        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2318        references. PCRE supports relative numbers with this syntax (I don't think
2319        Oniguruma does).
2320    
2321    12. Previously, a group with a zero repeat such as (...){0} was completely
2322        omitted from the compiled regex. However, this means that if the group
2323        was called as a subroutine from elsewhere in the pattern, things went wrong
2324        (an internal error was given). Such groups are now left in the compiled
2325        pattern, with a new opcode that causes them to be skipped at execution
2326        time.
2327    
2328    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2329        to the way PCRE behaves:
2330    
2331        (a) A lone ] character is dis-allowed (Perl treats it as data).
2332    
2333        (b) A back reference to an unmatched subpattern matches an empty string
2334            (Perl fails the current match path).
2335    
2336        (c) A data ] in a character class must be notated as \] because if the
2337            first data character in a class is ], it defines an empty class. (In
2338            Perl it is not possible to have an empty class.) The empty class []
2339            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2340            The negative empty class [^] matches any one character, independently
2341            of the DOTALL setting.
2342    
2343    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2344        non-existent subpattern following a character class starting with ']' and
2345        containing () gave an internal compiling error instead of "reference to
2346        non-existent subpattern". Fortunately, when the pattern did exist, the
2347        compiled code was correct. (When scanning forwards to check for the
2348        existence of the subpattern, it was treating the data ']' as terminating
2349        the class, so got the count wrong. When actually compiling, the reference
2350        was subsequently set up correctly.)
2351    
2352    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2353        it was being rejected as not supported by pcre_dfa_exec(), even though
2354        other assertions are supported. I have made pcre_dfa_exec() support
2355        (*FAIL).
2356    
2357    16. The implementation of 13c above involved the invention of a new opcode,
2358        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2359        cannot be changed at match time, I realized I could make a small
2360        improvement to matching performance by compiling OP_ALLANY instead of
2361        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2362        on the OP_ANY path.
2363    
2364    17. Compiling pcretest on Windows with readline support failed without the
2365        following two fixes: (1) Make the unistd.h include conditional on
2366        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2367    
2368    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2369        ncurses library to be included for pcretest when ReadLine support is
2370        requested, but also to allow for it to be overridden. This patch came from
2371        Daniel Bergström.
2372    
2373    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2374        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2375        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2376        spotting this.
2377    
2378    
2379    Version 7.6 28-Jan-08
2380    ---------------------
2381    
2382    1.  A character class containing a very large number of characters with
2383        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2384        overflow.
2385    
2386    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2387        HAVE_LONG_LONG is not defined.
2388    
2389    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2390        bring them up to date and include new features. This patch includes:
2391    
2392        - Fixed PH's badly added libz and libbz2 support.
2393        - Fixed a problem with static linking.
2394        - Added pcredemo. [But later removed - see 7 below.]
2395        - Fixed dftables problem and added an option.
2396        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2397            HAVE_LONG_LONG.
2398        - Added readline support for pcretest.
2399        - Added an listing of the option settings after cmake has run.
2400    
2401    4.  A user submitted a patch to Makefile that makes it easy to create
2402        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2403        Makefile.am that cause it to include this special target, without
2404        affecting anything else. Note that the same mingw target plus all
2405        the other distribution libraries and programs are now supported
2406        when configuring with CMake (see 6 below) instead of with
2407        Configure/Make.
2408    
2409    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2410        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2411        exported in the Windows port". It has not yet been confirmed that the patch
2412        solves the problem, but it does no harm.
2413    
2414    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2415        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2416        with CMake, and also correct the comment about stack recursion.
2417    
2418    7.  Remove the automatic building of pcredemo from the ./configure system and
2419        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2420        of a program that users should build themselves after PCRE is installed, so
2421        building it automatically is not really right. What is more, it gave
2422        trouble in some build environments.
2423    
2424    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2425    
2426    
2427    Version 7.5 10-Jan-08
2428    ---------------------
2429    
2430    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2431        values in parens when parsing an RE using the C++ wrapper."
2432    
2433    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2434        Characters greater than 255 were excluded from the class instead of being
2435        included.
2436    
2437    3.  The same bug as (2) above applied to negated POSIX classes such as
2438        [:^space:].
2439    
2440    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2441        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2442        I have changed it.
2443    
2444    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2445        first named subpattern) and a construct such as (?&a) would reference the
2446        first named subpattern whose name started with "a" (in other words, the
2447        length check was missing). Both these problems are fixed. "Subpattern name
2448        expected" is now given for (?&) (a zero-length name), and this patch also
2449        makes it give the same error for \k'' (previously it complained that that
2450        was a reference to a non-existent subpattern).
2451    
2452    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2453        this is right because (?- can be followed by option settings as well as by
2454        digits. I have, however, made the messages clearer.
2455    
2456    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2457        than the number used in the conditional) now cause a compile-time error.
2458        This is actually not compatible with Perl, which accepts such patterns, but
2459        treats the conditional as always being FALSE (as PCRE used to), but it
2460        seems to me that giving a diagnostic is better.
2461    
2462    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2463        and messages.
2464    
2465    9.  Fix two occurrences of "backslash" in comments that should have been
2466        "backspace".
2467    
2468    10. Remove two redundant lines of code that can never be obeyed (their function
2469        was moved elsewhere).
2470    
2471    11. The program that makes PCRE's Unicode character property table had a bug
2472        which caused it to generate incorrect table entries for sequences of
2473        characters that have the same character type, but are in different scripts.
2474        It amalgamated them into a single range, with the script of the first of
2475        them. In other words, some characters were in the wrong script. There were
2476        thirteen such cases, affecting characters in the following ranges:
2477    
2478          U+002b0 - U+002c1
2479          U+0060c - U+0060d
2480          U+0061e - U+00612
2481          U+0064b - U+0065e
2482          U+0074d - U+0076d
2483          U+01800 - U+01805
2484          U+01d00 - U+01d77
2485          U+01d9b - U+01dbf
2486          U+0200b - U+0200f
2487          U+030fc - U+030fe
2488          U+03260 - U+0327f
2489          U+0fb46 - U+0fbb1
2490          U+10450 - U+1049d
2491    
2492    12. The -o option (show only the matching part of a line) for pcregrep was not
2493        compatible with GNU grep in that, if there was more than one match in a
2494        line, it showed only the first of them. It now behaves in the same way as
2495        GNU grep.
2496    
2497    13. If the -o and -v options were combined for pcregrep, it printed a blank
2498        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2499        does the same. The return code can be used to tell if there were any
2500        non-matching lines.
2501    
2502    14. Added --file-offsets and --line-offsets to pcregrep.
2503    
2504    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2505        infinitely looping recursion. The bug was that positive lookaheads were not
2506        being skipped when checking for a possible empty match (negative lookaheads
2507        and both kinds of lookbehind were skipped).
2508    
2509    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2510        inclusion of <windows.h> to before rather than after the definition of
2511        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2512    
2513    17. Specifying a possessive quantifier with a specific limit for a Unicode
2514        character property caused pcre_compile() to compile bad code, which led at
2515        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2516        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2517        caused the error; without that there was no problem.
2518    
2519    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2520    
2521    19. Added --enable-pcretest-libreadline.
2522    
2523    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2524        RE::GlobalReplace(). As a result, the number of replacements returned was
2525        double what it should be. I removed one of the increments, but Craig sent a
2526        later patch that removed the other one (the right fix) and added unit tests
2527        that check the return values (which was not done before).
2528    
2529    21. Several CMake things:
2530    
2531        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2532            the names libpcre and libpcreposix, not just pcre and pcreposix.
2533    
2534        (2) The above change means that pcretest and pcregrep are now correctly
2535            linked with the newly-built libraries, not previously installed ones.
2536    
2537        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2538    
2539    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2540        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2541        UTF-8 newline character). The key issue is that the pattern starts .*;
2542        this means that the match must be either at the beginning, or after a
2543        newline. The bug was in the code for advancing after a failed match and
2544        checking that the new position followed a newline. It was not taking
2545        account of UTF-8 characters correctly.
2546    
2547    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2548        character classes. PCRE was not treating the sequence [:...:] as a
2549        character class unless the ... were all letters. Perl, however, seems to
2550        allow any characters between [: and :], though of course it rejects as
2551        unknown any "names" that contain non-letters, because all the known class
2552        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2553        for example, whereas PCRE did not - it did not recognize a POSIX character
2554        class. This seemed a bit dangerous, so the code has been changed to be
2555        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2556        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2557        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2558        Perl does, and where it didn't before.
2559    
2560    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2561        Windows environments %n is disabled by default.
2562    
2563    
2564    Version 7.4 21-Sep-07
2565    ---------------------
2566    
2567    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2568        means that a class such as [\s] counted as "explicit reference to CR or
2569        LF". That isn't really right - the whole point of the change was to try to
2570        help when there was an actual mention of one of the two characters. So now
2571        the change happens only if \r or \n (or a literal CR or LF) character is
2572        encountered.
2573    
2574    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2575        of both had grown to the point where there were only 3 bits left.
2576        Fortunately, there was spare space in the data structure, and so I have
2577        moved the internal flags into a new 16-bit field to free up more option
2578        bits.
2579    
2580    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2581        but did not set the internal JCHANGED flag - either of these is enough to
2582        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2583        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2584        start sets both bits.
2585    
2586    4.  Added options (at build time, compile time, exec time) to change \R from
2587        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2588    
2589    5.  doc/pcresyntax.html was missing from the distribution.
2590    
2591    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2592        compatibility, even though it is no longer used.
2593    
2594    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2595        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2596        windows.h file is present (where different names are used). [This was
2597        reversed later after testing - see 16 below.]
2598    
2599    8.  Changed all #include <config.h> to #include "config.h". There were also
2600        some further <pcre.h> cases that I changed to "pcre.h".
2601    
2602    9.  When pcregrep was used with the --colour option, it missed the line ending
2603        sequence off the lines that it output.
2604    
2605    10. It was pointed out to me that arrays of string pointers cause lots of
2606        relocations when a shared library is dynamically loaded. A technique of
2607        using a single long string with a table of offsets can drastically reduce
2608        these. I have refactored PCRE in four places to do this. The result is
2609        dramatic:
2610    
2611          Originally:                          290
2612          After changing UCP table:            187
2613          After changing error message table:   43
2614          After changing table of "verbs"       36
2615          After changing table of Posix names   22
2616    
2617        Thanks to the folks working on Gregex for glib for this insight.
2618    
2619    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2620        unicode-properties was also set.
2621    
2622    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2623    
2624    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2625        checked only for CRLF.
2626    
2627    14. Added casts to pcretest.c to avoid compiler warnings.
2628    
2629    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2630    
2631    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2632        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2633        entirely. This removes changes made in 7 above.
2634    
2635    17. The CMake files have been updated, and there is now more information about
2636        building with CMake in the NON-UNIX-USE document.
2637    
2638    
2639    Version 7.3 28-Aug-07
2640  ---------------------  ---------------------
2641    
2642   1. In the rejigging of the build system that eventually resulted in 7.1, the   1. In the rejigging of the build system that eventually resulted in 7.1, the
# Line 143  Version 7.3 17-Aug-07 Line 2778  Version 7.3 17-Aug-07
2778  24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte  24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2779      characters were involved (for example /\S{2}/8g with "A\x{a3}BC").      characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2780    
2781    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2782    
2783    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2784        character were causing crashes (broken optimization).
2785    
2786    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2787        \p or \P) caused a compile-time loop.
2788    
2789    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2790        break. For example, the unstudied pattern [\r\n]A does not match the string
2791        "\r\nA" because change 7.0/46 below moves the current point on by two
2792        characters after failing to match at the start. However, the pattern \nA
2793        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2794        the same is true. There doesn't seem any very clean way out of this, but
2795        what I have chosen to do makes the common cases work: PCRE now takes note
2796        of whether there can be an explicit match for \r or \n anywhere in the
2797        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2798        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2799        pattern has explicit CR or LF references.
2800    
2801    29. Added (*CR) etc for changing newline setting at start of pattern.
2802    
2803    
2804  Version 7.2 19-Jun-07  Version 7.2 19-Jun-07
2805  ---------------------  ---------------------

Legend:
Removed from v.221  
changed lines
  Added in v.1504

  ViewVC Help
Powered by ViewVC 1.1.5