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

Diff of /code/trunk/ChangeLog

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

revision 431 by ph10, Wed Sep 2 09:23:21 2009 UTC revision 1534 by ph10, Tue Mar 24 10:33:21 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 8.00 ??-???-??  Version 8.37 xx-xxx-2015
5    ------------------------
6    
7    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
8        for those parentheses to be closed with whatever has been captured so far.
9        However, it was failing to mark any other groups between the hightest
10        capture so far and the currrent group as "unset". Thus, the ovector for
11        those groups contained whatever was previously there. An example is the
12        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
13    
14    2.  If an assertion condition was quantified with a minimum of zero (an odd
15        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
16    
17    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
18        unrecognized modifier, a crash could occur.
19    
20    4.  An attempt to do global matching in pcretest with a zero-length ovector
21        caused a crash.
22    
23    5.  Fixed a memory leak during matching that could occur for a subpattern
24        subroutine call (recursive or otherwise) if the number of captured groups
25        that had to be saved was greater than ten.
26    
27    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
28        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
29        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
30    
31    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
32        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
33    
34    8. When a pattern is compiled, it remembers the highest back reference so that
35       when matching, if the ovector is too small, extra memory can be obtained to
36       use instead. A conditional subpattern whose condition is a check on a
37       capture having happened, such as, for example in the pattern
38       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
39       setting the highest backreference number. This mattered only if pcre_exec()
40       was called with an ovector that was too small to hold the capture, and there
41       was no other kind of back reference (a situation which is probably quite
42       rare). The effect of the bug was that the condition was always treated as
43       FALSE when the capture could not be consulted, leading to a incorrect
44       behaviour by pcre2_match(). This bug has been fixed.
45    
46    9. A reference to a duplicated named group (either a back reference or a test
47       for being set in a conditional) that occurred in a part of the pattern where
48       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
49       to be incorrectly calculated, leading to overwriting.
50    
51    10. A mutually recursive set of back references such as (\2)(\1) caused a
52        segfault at study time (while trying to find the minimum matching length).
53        The infinite loop is now broken (with the minimum length unset, that is,
54        zero).
55    
56    11. If an assertion that was used as a condition was quantified with a minimum
57        of zero, matching went wrong. In particular, if the whole group had
58        unlimited repetition and could match an empty string, a segfault was
59        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
60        assertions to be quantified, but not if they are being used as conditions,
61        so the above pattern is faulted by Perl. PCRE has now been changed so that
62        it also rejects such patterns.
63    
64    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
65        failed to allow the zero-repeat case if pcre2_exec() was called with an
66        ovector too small to capture the group.
67    
68    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
69        Red Hat Product Security:
70    
71        (a) A crash if /K and /F were both set with the option to save the compiled
72        pattern.
73    
74        (b) Another crash if the option to print captured substrings in a callout
75        was combined with setting a null ovector, for example \O\C+ as a subject
76        string.
77    
78    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
79        forward reference repeated a large (but limited) number of times within a
80        repeated outer group that has a zero minimum quantifier, caused incorrect
81        code to be compiled, leading to the error "internal error:
82        previously-checked referenced subpattern not found" when an incorrect
83        memory address was read. This bug was reported as "heap overflow",
84        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
85        CVE-2015-2325.
86    
87    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
88        call within a group that also contained a recursive back reference caused
89        incorrect code to be compiled. This bug was reported as "heap overflow",
90        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
91        number CVE-2015-2326.
92    
93    24. Computing the size of the JIT read-only data in advance has been a source
94        of various issues, and new ones are still appear unfortunately. To fix
95        existing and future issues, size computation is eliminated from the code,
96        and replaced by on-demand memory allocation.
97    
98    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
99        adjacent to the end of the range, and the range contained characters with
100        more than one other case, caused incorrect behaviour when compiled in UTF
101        mode. In that example, the range a-j was left out of the class.
102    
103    26. Fix JIT compilation of conditional blocks, which assertion
104        is converted to (*FAIL). E.g: /(?(?!))/.
105    
106    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
107        discovered by the LLVM fuzzer.
108    
109    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
110        when this assertion was used as a condition, for example (?(?!)a|b). In
111        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
112        error about an unsupported item.
113    
114    
115    Version 8.36 26-September-2014
116    ------------------------------
117    
118    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
119        -Wmissing-field-initializers and -Wunused-parameter.
120    
121    2.  The tests for quantifiers being too big (greater than 65535) were being
122        applied after reading the number, and stupidly assuming that integer
123        overflow would give a negative number. The tests are now applied as the
124        numbers are read.
125    
126    3.  Tidy code in pcre_exec.c where two branches that used to be different are
127        now the same.
128    
129    4.  The JIT compiler did not generate match limit checks for certain
130        bracketed expressions with quantifiers. This may lead to exponential
131        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
132        issue should be resolved now.
133    
134    5.  Fixed an issue, which occures when nested alternatives are optimized
135        with table jumps.
136    
137    6.  Inserted two casts and changed some ints to size_t in the light of some
138        reported 64-bit compiler warnings (Bugzilla 1477).
139    
140    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
141        an empty string, which sometimes were behaving incorrectly in the
142        interpreter (though correctly in the JIT matcher). This pcretest input is
143        an example:
144    
145          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
146          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
147    
148        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
149        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
150        for an empty string was breaking the inner loop and carrying on at a lower
151        level, when possessive repeated groups should always return to a higher
152        level as they have no backtrack points in them. The empty string test now
153        occurs at the outer level.
154    
155    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
156        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
157    
158    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
159        doesn't).
160    
161    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
162        to hold a pattern to be incorrectly computed (too small) when there were
163        named back references to duplicated names. This could cause "internal
164        error: code overflow" or "double free or corruption" or other memory
165        handling errors.
166    
167    11. When named subpatterns had the same prefixes, back references could be
168        confused. For example, in this pattern:
169    
170          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
171    
172        the reference to 'Name' was incorrectly treated as a reference to a
173        duplicate name.
174    
175    12. A pattern such as /^s?c/mi8 where the optional character has more than
176        one "other case" was incorrectly compiled such that it would only try to
177        match starting at "c".
178    
179    13. When a pattern starting with \s was studied, VT was not included in the
180        list of possible starting characters; this should have been part of the
181        8.34/18 patch.
182    
183    14. If a character class started [\Qx]... where x is any character, the class
184        was incorrectly terminated at the ].
185    
186    15. If a pattern that started with a caseless match for a character with more
187        than one "other case" was studied, PCRE did not set up the starting code
188        unit bit map for the list of possible characters. Now it does. This is an
189        optimization improvement, not a bug fix.
190    
191    16. The Unicode data tables have been updated to Unicode 7.0.0.
192    
193    17. Fixed a number of memory leaks in pcregrep.
194    
195    18. Avoid a compiler warning (from some compilers) for a function call with
196        a cast that removes "const" from an lvalue by using an intermediate
197        variable (to which the compiler does not object).
198    
199    19. Incorrect code was compiled if a group that contained an internal recursive
200        back reference was optional (had quantifier with a minimum of zero). This
201        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
202        caused segmentation faults because of stack overflows at compile time.
203    
204    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
205        group that is quantified with an indefinite repeat, caused a compile-time
206        loop which used up all the system stack and provoked a segmentation fault.
207        This was not the same bug as 19 above.
208    
209    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
210        Patch by Mike Frysinger.
211    
212    
213    Version 8.35 04-April-2014
214    --------------------------
215    
216    1.  A new flag is set, when property checks are present in an XCLASS.
217        When this flag is not set, PCRE can perform certain optimizations
218        such as studying these XCLASS-es.
219    
220    2.  The auto-possessification of character sets were improved: a normal
221        and an extended character set can be compared now. Furthermore
222        the JIT compiler optimizes more character set checks.
223    
224    3.  Got rid of some compiler warnings for potentially uninitialized variables
225        that show up only when compiled with -O2.
226    
227    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
228        of a match later then the end of the match. The pcretest program was not
229        handling the case sensibly - it was outputting from the start to the next
230        binary zero. It now reports this situation in a message, and outputs the
231        text from the end to the start.
232    
233    5.  Fast forward search is improved in JIT. Instead of the first three
234        characters, any three characters with fixed position can be searched.
235        Search order: first, last, middle.
236    
237    6.  Improve character range checks in JIT. Characters are read by an inprecise
238        function now, which returns with an unknown value if the character code is
239        above a certain threshold (e.g: 256). The only limitation is that the value
240        must be bigger than the threshold as well. This function is useful when
241        the characters above the threshold are handled in the same way.
242    
243    7.  The macros whose names start with RAWUCHAR are placeholders for a future
244        mode in which only the bottom 21 bits of 32-bit data items are used. To
245        make this more memorable for those maintaining the code, the names have
246        been changed to start with UCHAR21, and an extensive comment has been added
247        to their definition.
248    
249    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
250        to the export list in Makefile.am (they were accidentally omitted from the
251        8.34 tarball).
252    
253    9.  The informational output from pcretest used the phrase "starting byte set"
254        which is inappropriate for the 16-bit and 32-bit libraries. As the output
255        for "first char" and "need char" really means "non-UTF-char", I've changed
256        "byte" to "char", and slightly reworded the output. The documentation about
257        these values has also been (I hope) clarified.
258    
259    10. Another JIT related optimization: use table jumps for selecting the correct
260        backtracking path, when more than four alternatives are present inside a
261        bracket.
262    
263    11. Empty match is not possible, when the minimum length is greater than zero,
264        and there is no \K in the pattern. JIT should avoid empty match checks in
265        such cases.
266    
267    12. In a caseless character class with UCP support, when a character with more
268        than one alternative case was not the first character of a range, not all
269        the alternative cases were added to the class. For example, s and \x{17f}
270        are both alternative cases for S: the class [RST] was handled correctly,
271        but [R-T] was not.
272    
273    13. The configure.ac file always checked for pthread support when JIT was
274        enabled. This is not used in Windows, so I have put this test inside a
275        check for the presence of windows.h (which was already tested for).
276    
277    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
278        The algorithm provides a way to skip certain starting offsets, and usually
279        faster than linear prefix searches.
280    
281    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
282        as for 'fr_FR' and 'french'. For some reason, however, it then used the
283        Windows-specific input and output files, which have 'french' screwed in.
284        So this could never have worked. One of the problems with locales is that
285        they aren't always the same. I have now updated RunTest so that it checks
286        the output of the locale test (test 3) against three different output
287        files, and it allows the test to pass if any one of them matches. With luck
288        this should make the test pass on some versions of Solaris where it was
289        failing. Because of the uncertainty, the script did not used to stop if
290        test 3 failed; it now does. If further versions of a French locale ever
291        come to light, they can now easily be added.
292    
293    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
294        there was a message during ./configure, but it did not stop. This now
295        provokes an error. The invalid example in README has been corrected.
296        If a value less than the minimum is given, the minimum value has always
297        been used, but now a warning is given.
298    
299    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
300        was a bug in the test system, which is now fixed. Also, the list of various
301        configurations that are tested for each release did not have one with both
302        16/32 bits and --enable-bar-anycrlf. It now does.
303    
304    18. pcretest was missing "-C bsr" for displaying the \R default setting.
305    
306    19. Little endian PowerPC systems are supported now by the JIT compiler.
307    
308    20. The fast forward newline mechanism could enter to an infinite loop on
309        certain invalid UTF-8 input. Although we don't support these cases
310        this issue can be fixed by a performance optimization.
311    
312    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
313        not take account if existing stack usage. There is now a new global
314        variable called pcre_stack_guard that can be set to point to an external
315        function to check stack availability. It is called at the start of
316        processing every parenthesized group.
317    
318    22. A typo in the code meant that in ungreedy mode the max/min qualifier
319        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
320        not match "ab".
321    
322    23. When UTF was disabled, the JIT program reported some incorrect compile
323        errors. These messages are silenced now.
324    
325    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
326        compiler.
327    
328    25. Change all the temporary files used in RunGrepTest to be different to those
329        used by RunTest so that the tests can be run simultaneously, for example by
330        "make -j check".
331    
332    
333    Version 8.34 15-December-2013
334    -----------------------------
335    
336    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
337        executable memory. Patch inspired by Carsten Klein.
338    
339    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
340        this macro is never tested and has no effect, because the work to support
341        coverage involves only compiling and linking options and special targets in
342        the Makefile. The comment in config.h implied that defining the macro would
343        enable coverage support, which is totally false. There was also support for
344        setting this macro in the CMake files (my fault, I just copied it from
345        configure). SUPPORT_GCOV has now been removed.
346    
347    3.  Make a small performance improvement in strlen16() and strlen32() in
348        pcretest.
349    
350    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
351        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
352        these cases). There was also one in pcretest.c.
353    
354    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
355    
356    6.  In UTF mode, the code for checking whether a group could match an empty
357        string (which is used for indefinitely repeated groups to allow for
358        breaking an infinite loop) was broken when the group contained a repeated
359        negated single-character class with a character that occupied more than one
360        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
361        UTF-8 mode). The effect was undefined: the group might or might not be
362        deemed as matching an empty string, or the program might have crashed.
363    
364    7.  The code for checking whether a group could match an empty string was not
365        recognizing that \h, \H, \v, \V, and \R must match a character.
366    
367    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
368        an empty string. If it can, pcretest shows this in its information output.
369    
370    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
371        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
372        matched by pcre_exec() without using JIT:
373    
374        (a) If the rest of the pattern did not match after a maximal run of
375            grapheme clusters, the code for backing up to try with fewer of them
376            did not always back up over a full grapheme when characters that do not
377            have the modifier quality were involved, e.g. Hangul syllables.
378    
379        (b) If the match point in a subject started with modifier character, and
380            there was no match, the code could incorrectly back up beyond the match
381            point, and potentially beyond the first character in the subject,
382            leading to a segfault or an incorrect match result.
383    
384    10. A conditional group with an assertion condition could lead to PCRE
385        recording an incorrect first data item for a match if no other first data
386        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
387        first data item, and therefore matched "ca" after "c" instead of at the
388        start.
389    
390    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
391        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
392    
393    12. The source of pcregrep now includes z/OS-specific code so that it can be
394        compiled for z/OS as part of the special z/OS distribution.
395    
396    13. Added the -T and -TM options to pcretest.
397    
398    14. The code in pcre_compile.c for creating the table of named capturing groups
399        has been refactored. Instead of creating the table dynamically during the
400        actual compiling pass, the information is remembered during the pre-compile
401        pass (on the stack unless there are more than 20 named groups, in which
402        case malloc() is used) and the whole table is created before the actual
403        compile happens. This has simplified the code (it is now nearly 150 lines
404        shorter) and prepared the way for better handling of references to groups
405        with duplicate names.
406    
407    15. A back reference to a named subpattern when there is more than one of the
408        same name now checks them in the order in which they appear in the pattern.
409        The first one that is set is used for the reference. Previously only the
410        first one was inspected. This change makes PCRE more compatible with Perl.
411    
412    16. Unicode character properties were updated from Unicode 6.3.0.
413    
414    17. The compile-time code for auto-possessification has been refactored, based
415        on a patch by Zoltan Herczeg. It now happens after instead of during
416        compilation. The code is cleaner, and more cases are handled. The option
417        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
418        options in pcretest are provided to set it. It can also be set by
419        (*NO_AUTO_POSSESS) at the start of a pattern.
420    
421    18. The character VT has been added to the default ("C" locale) set of
422        characters that match \s and are generally treated as white space,
423        following this same change in Perl 5.18. There is now no difference between
424        "Perl space" and "POSIX space". Whether VT is treated as white space in
425        other locales depends on the locale.
426    
427    19. The code for checking named groups as conditions, either for being set or
428        for being recursed, has been refactored (this is related to 14 and 15
429        above). Processing unduplicated named groups should now be as fast at
430        numerical groups, and processing duplicated groups should be faster than
431        before.
432    
433    20. Two patches to the CMake build system, by Alexander Barkov:
434    
435          (1) Replace the "source" command by "." in CMakeLists.txt because
436              "source" is a bash-ism.
437    
438          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
439              without these the CMake build does not work on Solaris.
440    
441    21. Perl has changed its handling of \8 and \9. If there is no previously
442        encountered capturing group of those numbers, they are treated as the
443        literal characters 8 and 9 instead of a binary zero followed by the
444        literals. PCRE now does the same.
445    
446    22. Following Perl, added \o{} to specify codepoints in octal, making it
447        possible to specify values greater than 0777 and also making them
448        unambiguous.
449    
450    23. Perl now gives an error for missing closing braces after \x{... instead of
451        treating the string as literal. PCRE now does the same.
452    
453    24. RunTest used to grumble if an inappropriate test was selected explicitly,
454        but just skip it when running all tests. This make it awkward to run ranges
455        of tests when one of them was inappropriate. Now it just skips any
456        inappropriate tests, as it always did when running all tests.
457    
458    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
459        character types such as \d or \w, too many callouts were inserted, and the
460        data that they returned was rubbish.
461    
462    26. In UCP mode, \s was not matching two of the characters that Perl matches,
463        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
464        were matched by \h. The code has now been refactored so that the lists of
465        the horizontal and vertical whitespace characters used for \h and \v (which
466        are defined only in one place) are now also used for \s.
467    
468    27. Add JIT support for the 64 bit TileGX architecture.
469        Patch by Jiong Wang (Tilera Corporation).
470    
471    28. Possessive quantifiers for classes (both explicit and automatically
472        generated) now use special opcodes instead of wrapping in ONCE brackets.
473    
474    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
475        (because it's meaningless), this was not happening when PCRE_CASELESS was
476        set. Not wrong, but inefficient.
477    
478    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
479        properties for \w, \d, etc) is present in a test regex. Otherwise if the
480        test contains no characters greater than 255, Perl doesn't realise it
481        should be using Unicode semantics.
482    
483    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
484        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
485        does in Unicode mode.
486    
487    32. Added the "forbid" facility to pcretest so that putting tests into the
488        wrong test files can sometimes be quickly detected.
489    
490    33. There is now a limit (default 250) on the depth of nesting of parentheses.
491        This limit is imposed to control the amount of system stack used at compile
492        time. It can be changed at build time by --with-parens-nest-limit=xxx or
493        the equivalent in CMake.
494    
495    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
496        errors. Perl warns for these when in warning mode, but PCRE has no facility
497        for giving warnings.
498    
499    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
500        However, this was not working for (?!) because it is optimized to (*FAIL),
501        for which PCRE does not allow quantifiers. The optimization is now disabled
502        when a quantifier follows (?!). I can't see any use for this, but it makes
503        things uniform.
504    
505    36. Perl no longer allows group names to start with digits, so I have made this
506        change also in PCRE. It simplifies the code a bit.
507    
508    37. In extended mode, Perl ignores spaces before a + that indicates a
509        possessive quantifier. PCRE allowed a space before the quantifier, but not
510        before the possessive +. It now does.
511    
512    38. The use of \K (reset reported match start) within a repeated possessive
513        group such as (a\Kb)*+ was not working.
514    
515    40. Document that the same character tables must be used at compile time and
516        run time, and that the facility to pass tables to pcre_exec() and
517        pcre_dfa_exec() is for use only with saved/restored patterns.
518    
519    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
520        features for Builds with MSVC:
521    
522        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
523           builds). The .rc files can be used to set FileDescription and many other
524           attributes.
525    
526        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
527           This allows higher-level build scripts which want .pdb files to avoid
528           hard-coding the exact files needed."
529    
530    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
531        mean "start of word" and "end of word", respectively, as a transition aid.
532    
533    43. A minimizing repeat of a class containing codepoints greater than 255 in
534        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
535        compiled to use the heap for recursion.
536    
537    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
538        is configured.
539    
540    
541    Version 8.33 28-May-2013
542    ------------------------
543    
544    1.  Added 'U' to some constants that are compared to unsigned integers, to
545        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
546        variables that are added to signed variables, to ensure the result is
547        signed and can be negated.
548    
549    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
550        CMake config files.
551    
552    3.  Revise the creation of config.h.generic so that all boolean macros are
553        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
554        overriding via -D on the command line possible.
555    
556    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
557        to unsigned int is reported to make a quite noticeable speed difference in
558        a specific Windows environment. Testing on Linux did also appear to show
559        some benefit (and it is clearly not harmful). Also fixed the definition of
560        Xop which should be unsigned.
561    
562    5.  Related to (4), changing the definition of the intermediate variable cc
563        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
564        improvements.
565    
566    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
567        unnecessary spaces.
568    
569    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
570        and later.
571    
572    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
573    
574    9.  Optimizing fast_forward_start_bits in JIT.
575    
576    10. Adding support for callouts in JIT, and fixing some issues revealed
577        during this work. Namely:
578    
579        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
580    
581        (b) Minimum length was not checked before the matching is started.
582    
583    11. The value of capture_last that is passed to callouts was incorrect in some
584        cases when there was a capture on one path that was subsequently abandoned
585        after a backtrack. Also, the capture_last value is now reset after a
586        recursion, since all captures are also reset in this case.
587    
588    12. The interpreter no longer returns the "too many substrings" error in the
589        case when an overflowing capture is in a branch that is subsequently
590        abandoned after a backtrack.
591    
592    13. In the pathological case when an offset vector of size 2 is used, pcretest
593        now prints out the matched string after a yield of 0 or 1.
594    
595    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
596        Only supported by the JIT compiler at the moment.
597    
598    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
599    
600    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
601        offset of the starting point of the matching process, provided the offsets
602        vector is large enough.
603    
604    17. The \A escape now records a lookbehind value of 1, though its execution
605        does not actually inspect the previous character. This is to ensure that,
606        in partial multi-segment matching, at least one character from the old
607        segment is retained when a new segment is processed. Otherwise, if there
608        are no lookbehinds in the pattern, \A might match incorrectly at the start
609        of a new segment.
610    
611    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
612    
613    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
614        gives some modest performance improvement in 8-bit mode.
615    
616    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
617        be expressed in certain programming languages using Universal Character
618        Names.
619    
620    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
621        which points out that "non characters" are not "characters that may not
622        appear in Unicode strings" but rather "characters that are reserved for
623        internal use and have only local meaning".
624    
625    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
626        there was a conditional group that depended on an assertion, if the
627        assertion was false, the callout that immediately followed the alternation
628        in the condition was skipped when pcre_exec() was used for matching.
629    
630    23. Allow an explicit callout to be inserted before an assertion that is the
631        condition for a conditional group, for compatibility with automatic
632        callouts, which always insert a callout at this point.
633    
634    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
635        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
636    
637    25. (*PRUNE) is now supported by the JIT compiler.
638    
639    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
640    
641    27. Fix the case where there are two or more SKIPs with arguments that may be
642        ignored.
643    
644    28. (*SKIP) is now supported by the JIT compiler.
645    
646    29. (*THEN) is now supported by the JIT compiler.
647    
648    30. Update RunTest with additional test selector options.
649    
650    31. The way PCRE handles backtracking verbs has been changed in two ways.
651    
652        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
653        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
654        backtracking. In some cases this makes it more Perl-compatible, but Perl's
655        rather obscure rules do not always do the same thing.
656    
657        (2) Previously, backtracking verbs were confined within assertions. This is
658        no longer the case for positive assertions, except for (*ACCEPT). Again,
659        this sometimes improves Perl compatibility, and sometimes does not.
660    
661    32. A number of tests that were in test 2 because Perl did things differently
662        have been moved to test 1, because either Perl or PCRE has changed, and
663        these tests are now compatible.
664    
665    32. Backtracking control verbs are now handled in the same way in JIT and
666        interpreter.
667    
668    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
669        contained a forward subroutine reference caused a compile error.
670    
671    34. Auto-detect and optimize limited repetitions in JIT.
672    
673    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
674        blocking (*UTF) etc.
675    
676    36. In the interpreter, maximizing pattern repetitions for characters and
677        character types now use tail recursion, which reduces stack usage.
678    
679    37. The value of the max lookbehind was not correctly preserved if a compiled
680        and saved regex was reloaded on a host of different endianness.
681    
682    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
683        of the compiled pattern block, expand the flags field from 16 to 32 bits
684        because it was almost full.
685    
686    39. Try madvise first before posix_madvise.
687    
688    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
689        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
690        original change locked out all matches of empty strings. This has been
691        changed so that one match of an empty string per line is recognized.
692        Subsequent searches on the same line (for colouring or for --only-matching,
693        for example) do not recognize empty strings.
694    
695    41. Applied a user patch to fix a number of spelling mistakes in comments.
696    
697    42. Data lines longer than 65536 caused pcretest to crash.
698    
699    43. Clarified the data type for length and startoffset arguments for pcre_exec
700        and pcre_dfa_exec in the function-specific man pages, where they were
701        explicitly stated to be in bytes, never having been updated. I also added
702        some clarification to the pcreapi man page.
703    
704    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
705        a segmentation fault.
706    
707    
708    Version 8.32 30-November-2012
709    -----------------------------
710    
711    1.  Improved JIT compiler optimizations for first character search and single
712        character iterators.
713    
714    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
715        Patch by Daniel Richard G.
716    
717    3.  Single character iterator optimizations in the JIT compiler.
718    
719    4.  Improved JIT compiler optimizations for character ranges.
720    
721    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
722        Reported by Giuseppe D'Angelo.
723    
724    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
725        of a line, was being set incorrectly in cases where .* appeared inside
726        atomic brackets at the start of a pattern, or where there was a subsequent
727        *PRUNE or *SKIP.
728    
729    7.  Improved instruction cache flush for POWER/PowerPC.
730        Patch by Daniel Richard G.
731    
732    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
733        grep:
734    
735        (a) There is now no limit to the number of patterns to be matched.
736    
737        (b) An error is given if a pattern is too long.
738    
739        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
740            are now supported.
741    
742        (d) --exclude-from and --include-from (multiple use) have been added.
743    
744        (e) Exclusions and inclusions now apply to all files and directories, not
745            just to those obtained from scanning a directory recursively.
746    
747        (f) Multiple uses of -f and --file-list are now supported.
748    
749        (g) In a Windows environment, the default for -d has been changed from
750            "read" (the GNU grep default) to "skip", because otherwise the presence
751            of a directory in the file list provokes an error.
752    
753        (h) The documentation has been revised and clarified in places.
754    
755    9.  Improve the matching speed of capturing brackets.
756    
757    10. Changed the meaning of \X so that it now matches a Unicode extended
758        grapheme cluster.
759    
760    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
761        out POSIX threads when JIT support is configured.
762    
763    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
764    
765    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
766        structure could go wrong in environments where size_t is not the same size
767        as int.
768    
769    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
770        set.
771    
772    15. The EBCDIC support had decayed; later updates to the code had included
773        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
774        general tidy up of EBCDIC-related issues, and the documentation was also
775        not quite right. There is now a test that can be run on ASCII systems to
776        check some of the EBCDIC-related things (but is it not a full test).
777    
778    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
779        in a small tidy to the code.
780    
781    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
782    
783    18. If the --only-matching (-o) option in pcregrep is specified multiple
784        times, each one causes appropriate output. For example, -o1 -o2 outputs the
785        substrings matched by the 1st and 2nd capturing parentheses. A separating
786        string can be specified by --om-separator (default empty).
787    
788    19. Improving the first n character searches.
789    
790    20. Turn case lists for horizontal and vertical white space into macros so that
791        they are defined only once.
792    
793    21. This set of changes together give more compatible Unicode case-folding
794        behaviour for characters that have more than one other case when UCP
795        support is available.
796    
797        (a) The Unicode property table now has offsets into a new table of sets of
798            three or more characters that are case-equivalent. The MultiStage2.py
799            script that generates these tables (the pcre_ucd.c file) now scans
800            CaseFolding.txt instead of UnicodeData.txt for character case
801            information.
802    
803        (b) The code for adding characters or ranges of characters to a character
804            class has been abstracted into a generalized function that also handles
805            case-independence. In UTF-mode with UCP support, this uses the new data
806            to handle characters with more than one other case.
807    
808        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
809            whose other case is greater than 256 are now correctly matched
810            caselessly. Previously, the high codepoint matched the low one, but not
811            vice versa.
812    
813        (d) The processing of \h, \H, \v, and \ in character classes now makes use
814            of the new class addition function, using character lists defined as
815            macros alongside the case definitions of 20 above.
816    
817        (e) Caseless back references now work with characters that have more than
818            one other case.
819    
820        (f) General caseless matching of characters with more than one other case
821            is supported.
822    
823    22. Unicode character properties were updated from Unicode 6.2.0
824    
825    23. Improved CMake support under Windows. Patch by Daniel Richard G.
826    
827    24. Add support for 32-bit character strings, and UTF-32
828    
829    25. Major JIT compiler update (code refactoring and bugfixing).
830        Experimental Sparc 32 support is added.
831    
832    26. Applied a modified version of Daniel Richard G's patch to create
833        pcre.h.generic and config.h.generic by "make" instead of in the
834        PrepareRelease script.
835    
836    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
837        pcre_compile.c when checking for a zero character.
838    
839    28. Introducing a native interface for JIT. Through this interface, the compiled
840        machine code can be directly executed. The purpose of this interface is to
841        provide fast pattern matching, so several sanity checks are not performed.
842        However, feature tests are still performed. The new interface provides
843        1.4x speedup compared to the old one.
844    
845    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
846        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
847        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
848    
849    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
850        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
851        the "old" RFC 2279). Instead, it ended up passing a negative length to
852        pcre_exec().
853    
854    31. Add support for GCC's visibility feature to hide internal functions.
855    
856    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
857        "unknown -C option" after outputting 0 or 1.
858    
859    33. There is now support for generating a code coverage report for the test
860        suite in environments where gcc is the compiler and lcov is installed. This
861        is mainly for the benefit of the developers.
862    
863    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
864        unaddressable using valgrind annotations, allowing valgrind to detect
865        invalid memory accesses. This is mainly for the benefit of the developers.
866    
867    25. (*UTF) can now be used to start a pattern in any of the three libraries.
868    
869    26. Give configure error if --enable-cpp but no C++ compiler found.
870    
871    
872    Version 8.31 06-July-2012
873    -------------------------
874    
875    1.  Fixing a wrong JIT test case and some compiler warnings.
876    
877    2.  Removed a bashism from the RunTest script.
878    
879    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
880        to unsigned type, result still unsigned" that was given by an MS compiler
881        on encountering the code "-sizeof(xxx)".
882    
883    4.  Partial matching support is added to the JIT compiler.
884    
885    5.  Fixed several bugs concerned with partial matching of items that consist
886        of more than one character:
887    
888        (a) /^(..)\1/ did not partially match "aba" because checking references was
889            done on an "all or nothing" basis. This also applied to repeated
890            references.
891    
892        (b) \R did not give a hard partial match if \r was found at the end of the
893            subject.
894    
895        (c) \X did not give a hard partial match after matching one or more
896            characters at the end of the subject.
897    
898        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
899            a partial match for the string "\r".
900    
901        (e) When newline was set to CRLF, the metacharacter "." did not recognize
902            a partial match for a CR character at the end of the subject string.
903    
904    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
905        running pcretest, the text "(JIT)" added to the output whenever JIT is
906        actually used to run the match.
907    
908    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
909        or /S+[+] with a digit between 1 and 7.
910    
911    8.  OP_NOT now supports any UTF character not just single-byte ones.
912    
913    9.  (*MARK) control verb is now supported by the JIT compiler.
914    
915    10. The command "./RunTest list" lists the available tests without actually
916        running any of them. (Because I keep forgetting what they all are.)
917    
918    11. Add PCRE_INFO_MAXLOOKBEHIND.
919    
920    12. Applied a (slightly modified) user-supplied patch that improves performance
921        when the heap is used for recursion (compiled with --disable-stack-for-
922        recursion). Instead of malloc and free for each heap frame each time a
923        logical recursion happens, frames are retained on a chain and re-used where
924        possible. This sometimes gives as much as 30% improvement.
925    
926    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
927        call.
928    
929    14. As documented, (*COMMIT) is now confined to within a positive assertion.
930    
931    15. It is now possible to link pcretest with libedit as an alternative to
932        libreadline.
933    
934    16. (*COMMIT) control verb is now supported by the JIT compiler.
935    
936    17. The Unicode data tables have been updated to Unicode 6.1.0.
937    
938    18. Added --file-list option to pcregrep.
939    
940    19. Added binary file support to pcregrep, including the -a, --binary-files,
941        -I, and --text options.
942    
943    20. The madvise function is renamed for posix_madvise for QNX compatibility
944        reasons. Fixed by Giuseppe D'Angelo.
945    
946    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
947    
948    22. Changed the option for creating symbolic links for 16-bit man pages from
949        -s to -sf so that re-installing does not cause issues.
950    
951    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
952    
953    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
954        matches in certain environments (the workspace was not being correctly
955        retained). Also added to pcre_dfa_exec() a simple plausibility check on
956        some of the workspace data at the beginning of a restart.
957    
958    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
959        was not doing so when it should - probably a typo introduced by SVN 528
960        (change 8.10/14).
961    
962    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
963        \w+ when the character tables indicated that \x{c4} was a word character.
964        There were several related cases, all because the tests for doing a table
965        lookup were testing for characters less than 127 instead of 255.
966    
967    27. If a pattern contains capturing parentheses that are not used in a match,
968        their slots in the ovector are set to -1. For those that are higher than
969        any matched groups, this happens at the end of processing. In the case when
970        there were back references that the ovector was too small to contain
971        (causing temporary malloc'd memory to be used during matching), and the
972        highest capturing number was not used, memory off the end of the ovector
973        was incorrectly being set to -1. (It was using the size of the temporary
974        memory instead of the true size.)
975    
976    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
977        ovector size, it uses memory at the end of the block that it has got.
978    
979    29. Check for an overlong MARK name and give an error at compile time. The
980        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
981    
982    30. JIT compiler update.
983    
984    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
985        Rill for the patch.
986    
987    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
988    
989    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
990    
991    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
992        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
993        was enabled.
994    
995    35. Improve JIT code generation for greedy plus quantifier.
996    
997    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
998        1 to "aa" instead of to an empty string. The bug affected repeated groups
999        that could potentially match an empty string.
1000    
1001    37. Optimizing single character iterators in JIT.
1002    
1003    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1004        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1005        codepoints that are too big for the mode are faulted, and in a UTF mode,
1006        disallowed codepoints are also faulted.
1007    
1008    39. If PCRE was compiled with UTF support, in three places in the DFA
1009        matcher there was code that should only have been obeyed in UTF mode, but
1010        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1011        processing when bytes with values greater than 127 were present. In 16-bit
1012        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1013        both cases the values are those that cannot be the first data item in a UTF
1014        character. The three items that might have provoked this were recursions,
1015        possessively repeated groups, and atomic groups.
1016    
1017    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1018        and pcregrep, because some OS require shared objects to be explicitly
1019        passed to ld, causing the link step to fail if they are not.
1020    
1021    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1022        mode, patterns that started with \h* or \R* might be incorrectly matched.
1023    
1024    
1025    Version 8.30 04-February-2012
1026    -----------------------------
1027    
1028    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1029        name is defined in ctype.h.
1030    
1031    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1032        only in quite long subpatterns.
1033    
1034    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1035        since it was replaced by pcre_fullinfo() in February 2000.
1036    
1037    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1038        match a (*MARK), and the match failed at the start of the subject, a
1039        reference to memory before the start of the subject could occur. This bug
1040        was introduced by fix 17 of release 8.21.
1041    
1042    5.  A reference to an unset group with zero minimum repetition was giving
1043        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1044        /(another)?(\1?)test/ matched against "hello world test". This bug was
1045        introduced in release 8.13.
1046    
1047    6.  Add support for 16-bit character strings (a large amount of work involving
1048        many changes and refactorings).
1049    
1050    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1051        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1052        from a file.
1053    
1054    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1055        rounding is not applied in this particular case).
1056    
1057    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1058        if they appear, or are escaped, in patterns.
1059    
1060    10. Get rid of a number of -Wunused-but-set-variable warnings.
1061    
1062    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1063        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1064        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1065        also returns the mark "x". This bug applied to capturing parentheses,
1066        non-capturing parentheses, and atomic parentheses. It also applied to some
1067        assertions.
1068    
1069    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1070        information out of configure.ac instead of relying on pcre.h.generic, which
1071        is not stored in the repository.
1072    
1073    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1074        -lreadline.
1075    
1076    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1077    
1078    15. Applied Graycode's patch to put the top-level frame on the stack rather
1079        than the heap when not using the stack for recursion. This gives a
1080        performance improvement in many cases when recursion is not deep.
1081    
1082    16. Experimental code added to "pcretest -C" to output the stack frame size.
1083    
1084    
1085    Version 8.21 12-Dec-2011
1086    ------------------------
1087    
1088    1.  Updating the JIT compiler.
1089    
1090    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1091        are added as well.
1092    
1093    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1094        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1095        calling _pcre_jit_exec. Some extra comments are added.
1096    
1097    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1098        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1099        was introduced by change 18 for 8.20.
1100    
1101    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1102        ECMA-262 standard.
1103    
1104    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1105        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1106        This bug was probably introduced by change 9 of 8.13.
1107    
1108    7.  While fixing 6 above, I noticed that a number of other items were being
1109        incorrectly rejected as "not fixed length". This arose partly because newer
1110        opcodes had not been added to the fixed-length checking code. I have (a)
1111        corrected the bug and added tests for these items, and (b) arranged for an
1112        error to occur if an unknown opcode is encountered while checking for fixed
1113        length instead of just assuming "not fixed length". The items that were
1114        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1115        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1116        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1117    
1118    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1119        being incorrectly compiled and would have given unpredicatble results.
1120    
1121    9.  A possessively repeated subpattern with minimum repeat count greater than
1122        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1123        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1124        the first (A) could occur when it should not.
1125    
1126    10. Add a cast and remove a redundant test from the code.
1127    
1128    11. JIT should use pcre_malloc/pcre_free for allocation.
1129    
1130    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1131        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1132        is anything other than /usr, -L is still shown).
1133    
1134    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1135    
1136    14. Perl does not support \N without a following name in a [] class; PCRE now
1137        also gives an error.
1138    
1139    15. If a forward reference was repeated with an upper limit of around 2000,
1140        it caused the error "internal error: overran compiling workspace". The
1141        maximum number of forward references (including repeats) was limited by the
1142        internal workspace, and dependent on the LINK_SIZE. The code has been
1143        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1144        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1145        of around 200,000 forward references. While doing this, I also speeded up
1146        the filling in of repeated forward references.
1147    
1148    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1149        incorrectly expecting the subject to contain another "a" after the start.
1150    
1151    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1152        in the match, the SKIP should be ignored. This was not happening; instead
1153        the SKIP was being treated as NOMATCH. For patterns such as
1154        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1155        tested.
1156    
1157    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1158        now much more compatible with Perl, in particular in cases where the result
1159        is a non-match for a non-anchored pattern. For example, if
1160        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1161        "m", where previously it did not return a name. A side effect of this
1162        change is that for partial matches, the last encountered mark name is
1163        returned, as for non matches. A number of tests that were previously not
1164        Perl-compatible have been moved into the Perl-compatible test files. The
1165        refactoring has had the pleasing side effect of removing one argument from
1166        the match() function, thus reducing its stack requirements.
1167    
1168    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1169        subsequent uses of /S (without +) incorrectly behaved like /S+.
1170    
1171    21. Retrieve executable code size support for the JIT compiler and fixing
1172        some warnings.
1173    
1174    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1175        not work when the shorter character appeared right at the end of the
1176        subject string.
1177    
1178    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1179        systems.
1180    
1181    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1182        output it when the /M option is used in pcretest.
1183    
1184    25. The CheckMan script was not being included in the distribution. Also, added
1185        an explicit "perl" to run Perl scripts from the PrepareRelease script
1186        because this is reportedly needed in Windows.
1187    
1188    26. If study data was being save in a file and studying had not found a set of
1189        "starts with" bytes for the pattern, the data written to the file (though
1190        never used) was taken from uninitialized memory and so caused valgrind to
1191        complain.
1192    
1193    27. Updated RunTest.bat as provided by Sheri Pierce.
1194    
1195    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1196    
1197    29. Computation of memory usage for the table of capturing group names was
1198        giving an unnecessarily large value.
1199    
1200    
1201    Version 8.20 21-Oct-2011
1202    ------------------------
1203    
1204    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1205        a POSIX class. After further experiments with Perl, which convinced me that
1206        Perl has bugs and confusions, a closing square bracket is no longer allowed
1207        in a POSIX name. This bug also affected patterns with classes that started
1208        with full stops.
1209    
1210    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1211        captured substring, but while checking the failing first alternative,
1212        substring 1 is temporarily captured. If the output vector supplied to
1213        pcre_exec() was not big enough for this capture, the yield of the function
1214        was still zero ("insufficient space for captured substrings"). This cannot
1215        be totally fixed without adding another stack variable, which seems a lot
1216        of expense for a edge case. However, I have improved the situation in cases
1217        such as /(a)(b)x|abc/ matched against "abc", where the return code
1218        indicates that fewer than the maximum number of slots in the ovector have
1219        been set.
1220    
1221    3.  Related to (2) above: when there are more back references in a pattern than
1222        slots in the output vector, pcre_exec() uses temporary memory during
1223        matching, and copies in the captures as far as possible afterwards. It was
1224        using the entire output vector, but this conflicts with the specification
1225        that only 2/3 is used for passing back captured substrings. Now it uses
1226        only the first 2/3, for compatibility. This is, of course, another edge
1227        case.
1228    
1229    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1230        main code base, and can be used by building with --enable-jit. When this is
1231        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1232        runtime --no-jit option is given.
1233    
1234    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1235        ovector, the return from the function was zero, implying that there were
1236        other matches that did not fit. The correct "exactly full" value is now
1237        returned.
1238    
1239    6.  If a subpattern that was called recursively or as a subroutine contained
1240        (*PRUNE) or any other control that caused it to give a non-standard return,
1241        invalid errors such as "Error -26 (nested recursion at the same subject
1242        position)" or even infinite loops could occur.
1243    
1244    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1245        computing the minimum length on reaching *ACCEPT, and so ended up with the
1246        wrong value of 1 rather than 0. Further investigation indicates that
1247        computing a minimum subject length in the presence of *ACCEPT is difficult
1248        (think back references, subroutine calls), and so I have changed the code
1249        so that no minimum is registered for a pattern that contains *ACCEPT.
1250    
1251    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1252        it was not handled as intended. [But see 16 below.]
1253    
1254    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1255        Sheri Pierce.
1256    
1257    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1258        the first byte in a match must be "a".
1259    
1260    11. Change 17 for 8.13 increased the recursion depth for patterns like
1261        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1262        pattern contains any instances of (*THEN). If it does not, the old
1263        optimizations are restored. It would be nice to do this on a per-group
1264        basis, but at the moment that is not feasible.
1265    
1266    12. In some environments, the output of pcretest -C is CRLF terminated. This
1267        broke RunTest's code that checks for the link size. A single white space
1268        character after the value is now allowed for.
1269    
1270    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1271        For "fr", it uses the Windows-specific input and output files.
1272    
1273    14. If (*THEN) appeared in a group that was called recursively or as a
1274        subroutine, it did not work as intended. [But see next item.]
1275    
1276    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1277        pattern fragments (but not containing any | characters). If A and B are
1278        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1279        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1280        D. In other words, Perl considers parentheses that do not contain any |
1281        characters to be part of a surrounding alternative, whereas PCRE was
1282        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1283        differently. PCRE now behaves in the same way as Perl, except in the case
1284        of subroutine/recursion calls such as (?1) which have in any case always
1285        been different (but PCRE had them first :-).
1286    
1287    16. Related to 15 above: Perl does not treat the | in a conditional group as
1288        creating alternatives. Such a group is treated in the same way as an
1289        ordinary group without any | characters when processing (*THEN). PCRE has
1290        been changed to match Perl's behaviour.
1291    
1292    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1293        RunGrepTest script failed.
1294    
1295    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1296        inevitable for groups that contain captures, but it can lead to a lot of
1297        stack use in large patterns. The old behaviour has been restored for atomic
1298        groups that do not contain any capturing parentheses.
1299    
1300    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1301        suppress the check for a minimum subject length at run time. (If it was
1302        given to pcre_exec() or pcre_dfa_exec() it did work.)
1303    
1304    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1305        fail to work when decoding hex characters in data strings in EBCDIC
1306        environments.
1307    
1308    21. It appears that in at least one Mac OS environment, the isxdigit() function
1309        is implemented as a macro that evaluates to its argument more than once,
1310        contravening the C 90 Standard (I haven't checked a later standard). There
1311        was an instance in pcretest which caused it to go wrong when processing
1312        \x{...} escapes in subject strings. The has been rewritten to avoid using
1313        things like p++ in the argument of isxdigit().
1314    
1315    
1316    Version 8.13 16-Aug-2011
1317    ------------------------
1318    
1319    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1320    
1321    2.  Two minor typos in pcre_internal.h have been fixed.
1322    
1323    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1324        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1325        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1326    
1327    4.  There were a number of related bugs in the code for matching backrefences
1328        caselessly in UTF-8 mode when codes for the characters concerned were
1329        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1330        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1331        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1332        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1333        2-byte code at the end of the subject (it thought there wasn't enough data
1334        left).
1335    
1336    5.  Comprehensive information about what went wrong is now returned by
1337        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1338        as the output vector has at least 2 elements. The offset of the start of
1339        the failing character and a reason code are placed in the vector.
1340    
1341    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1342        now returned is for the first byte of the failing character, instead of the
1343        last byte inspected. This is an incompatible change, but I hope it is small
1344        enough not to be a problem. It makes the returned offset consistent with
1345        pcre_exec() and pcre_dfa_exec().
1346    
1347    7.  pcretest now gives a text phrase as well as the error number when
1348        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1349        failure, the offset and reason code are output.
1350    
1351    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1352        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1353        back over a single character (\n). This seems wrong (because it treated the
1354        two characters as a single entity when going forwards), conflicts with the
1355        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1356        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1357        has been changed.
1358    
1359    9.  Some internal refactoring has changed the processing so that the handling
1360        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1361        time (the PCRE_DOTALL option was changed this way some time ago: version
1362        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1363        which was always a bit of a fudge. It also means that there is one less
1364        argument for the match() function, which reduces its stack requirements
1365        slightly. This change also fixes an incompatibility with Perl: the pattern
1366        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1367    
1368    10. More internal refactoring has drastically reduced the number of recursive
1369        calls to match() for possessively repeated groups such as (abc)++ when
1370        using pcre_exec().
1371    
1372    11. While implementing 10, a number of bugs in the handling of groups were
1373        discovered and fixed:
1374    
1375        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1376        (a|)*(?1) gave a compile-time internal error.
1377        ((a|)+)+  did not notice that the outer group could match an empty string.
1378        (^a|^)+   was not marked as anchored.
1379        (.*a|.*)+ was not marked as matching at start or after a newline.
1380    
1381    12. Yet more internal refactoring has removed another argument from the match()
1382        function. Special calls to this function are now indicated by setting a
1383        value in a variable in the "match data" data block.
1384    
1385    13. Be more explicit in pcre_study() instead of relying on "default" for
1386        opcodes that mean there is no starting character; this means that when new
1387        ones are added and accidentally left out of pcre_study(), testing should
1388        pick them up.
1389    
1390    14. The -s option of pcretest has been documented for ages as being an old
1391        synonym of -m (show memory usage). I have changed it to mean "force study
1392        for every regex", that is, assume /S for every regex. This is similar to -i
1393        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1394        using it. It makes it easier to run collections of tests with and without
1395        study enabled, and thereby test pcre_study() more easily. All the standard
1396        tests are now run with and without -s (but some patterns can be marked as
1397        "never study" - see 20 below).
1398    
1399    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1400        restoration of the capturing data to the outer values was not happening
1401        correctly.
1402    
1403    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1404        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1405        pattern had matched an empty string, and so incorrectly returned a no
1406        match.
1407    
1408    17. There was optimizing code for the last branch of non-capturing parentheses,
1409        and also for the obeyed branch of a conditional subexpression, which used
1410        tail recursion to cut down on stack usage. Unfortunately, now that there is
1411        the possibility of (*THEN) occurring in these branches, tail recursion is
1412        no longer possible because the return has to be checked for (*THEN). These
1413        two optimizations have therefore been removed. [But see 8.20/11 above.]
1414    
1415    18. If a pattern containing \R was studied, it was assumed that \R always
1416        matched two bytes, thus causing the minimum subject length to be
1417        incorrectly computed because \R can also match just one byte.
1418    
1419    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1420        was incorrectly computed.
1421    
1422    20. If /S is present twice on a test pattern in pcretest input, it now
1423        *disables* studying, thereby overriding the use of -s on the command line
1424        (see 14 above). This is necessary for one or two tests to keep the output
1425        identical in both cases.
1426    
1427    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1428        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1429    
1430    22. When an atomic group that contained a capturing parenthesis was
1431        successfully matched, but the branch in which it appeared failed, the
1432        capturing was not being forgotten if a higher numbered group was later
1433        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1434        group 1 to "a", when in fact it should be unset. This applied to multi-
1435        branched capturing and non-capturing groups, repeated or not, and also to
1436        positive assertions (capturing in negative assertions does not happen
1437        in PCRE) and also to nested atomic groups.
1438    
1439    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1440        subject after a captured substring, to make it easier to tell which of a
1441        number of identical substrings has been captured.
1442    
1443    24. The way atomic groups are processed by pcre_exec() has been changed so that
1444        if they are repeated, backtracking one repetition now resets captured
1445        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1446        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1447        "aaa". Previously, it would have been "a". As part of this code
1448        refactoring, the way recursive calls are handled has also been changed.
1449    
1450    25. If an assertion condition captured any substrings, they were not passed
1451        back unless some other capturing happened later. For example, if
1452        (?(?=(a))a) was matched against "a", no capturing was returned.
1453    
1454    26. When studying a pattern that contained subroutine calls or assertions,
1455        the code for finding the minimum length of a possible match was handling
1456        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1457        group 1 called group 2 while simultaneously a separate group 2 called group
1458        1). A stack overflow occurred in this case. I have fixed this by limiting
1459        the recursion depth to 10.
1460    
1461    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1462        Fortmann. This supports explicit test numbers on the command line, and has
1463        argument validation and error reporting.
1464    
1465    28. An instance of \X with an unlimited repeat could fail if at any point the
1466        first character it looked at was a mark character.
1467    
1468    29. Some minor code refactoring concerning Unicode properties and scripts
1469        should reduce the stack requirement of match() slightly.
1470    
1471    30. Added the '=' option to pcretest to check the setting of unused capturing
1472        slots at the end of the pattern, which are documented as being -1, but are
1473        not included in the return count.
1474    
1475    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1476        compiled something random. Now it gives a compile-time error (as does
1477        Perl).
1478    
1479    32. A *MARK encountered during the processing of a positive assertion is now
1480        recorded and passed back (compatible with Perl).
1481    
1482    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1483        had alternative anchored branches, the search for a second match in a line
1484        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1485        matched the line "0102" twice. The same bug affected patterns that started
1486        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1487        twice.
1488    
1489    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1490        does, and because of capturing effects, quantifying parenthesized
1491        assertions may at times be useful. Quantifiers are now allowed for
1492        parenthesized assertions.
1493    
1494    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1495    
1496    36. \g was being checked for fancy things in a character class, when it should
1497        just be a literal "g".
1498    
1499    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1500        appearance of a nested POSIX class supersedes an apparent external class.
1501        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1502        unescaped square brackets may also appear as part of class names. For
1503        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1504        more like Perl. (But see 8.20/1 above.)
1505    
1506    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1507        was because it thought it was \N{name}, which is not supported).
1508    
1509    39. Add minix to OS list not supporting the -S option in pcretest.
1510    
1511    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1512        cannot analyze patterns in sufficient detail to catch mutual recursions
1513        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1514        subgroup is called recursively as a subpattern for a second time at the
1515        same position in the subject string. In previous releases this might have
1516        been caught by the recursion limit, or it might have run out of stack.
1517    
1518    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1519        happen only once. PCRE was, however incorrectly giving a compile time error
1520        "recursive call could loop indefinitely" because it cannot analyze the
1521        pattern in sufficient detail. The compile time test no longer happens when
1522        PCRE is compiling a conditional subpattern, but actual runaway loops are
1523        now caught at runtime (see 40 above).
1524    
1525    42. It seems that Perl allows any characters other than a closing parenthesis
1526        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1527        has been changed to be the same.
1528    
1529    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1530        as not to get warnings when autogen.sh is called. Also changed
1531        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1532    
1533    44. To help people who use pcregrep to scan files containing exceedingly long
1534        lines, the following changes have been made:
1535    
1536        (a) The default value of the buffer size parameter has been increased from
1537            8K to 20K. (The actual buffer used is three times this size.)
1538    
1539        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1540            PCRE is built.
1541    
1542        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1543            to be set at run time.
1544    
1545        (d) Numerical values in pcregrep options can be followed by K or M, for
1546            example --buffer-size=50K.
1547    
1548        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1549            given and the return code is set to 2.
1550    
1551    45. Add a pointer to the latest mark to the callout data block.
1552    
1553    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1554        partial match of an empty string instead of no match. This was specific to
1555        the use of ".".
1556    
1557    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1558        complete match instead of a partial match. This bug was dependent on both
1559        the PCRE_UTF8 and PCRE_DOTALL options being set.
1560    
1561    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1562        starting byte set, because \b was not being ignored.
1563    
1564    
1565    Version 8.12 15-Jan-2011
1566    ------------------------
1567    
1568    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1569        checks for such things as part of the documentation building process.
1570    
1571    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1572        --match-limit and --recursion-limit options (added for 8.11). In
1573        particular, this made one of the standard tests fail. (The integer value
1574        went into the wrong half of a long int.)
1575    
1576    3.  If the --colour option was given to pcregrep with -v (invert match), it
1577        did strange things, either producing crazy output, or crashing. It should,
1578        of course, ignore a request for colour when reporting lines that do not
1579        match.
1580    
1581    4.  Another pcregrep bug caused similar problems if --colour was specified with
1582        -M (multiline) and the pattern match finished with a line ending.
1583    
1584    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1585        matched in multiline mode, the following line was shown as part of the
1586        match. This seems wrong, so I have changed it.
1587    
1588    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1589        the check for further matches in the same line (so they could be coloured)
1590        to overrun the end of the current line. If another match was found, it was
1591        incorrectly shown (and then shown again when found in the next line).
1592    
1593    7.  If pcregrep was compiled under Windows, there was a reference to the
1594        function pcregrep_exit() before it was defined. I am assuming this was
1595        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1596        reported by a user. I've moved the definition above the reference.
1597    
1598    
1599    Version 8.11 10-Dec-2010
1600    ------------------------
1601    
1602    1.  (*THEN) was not working properly if there were untried alternatives prior
1603        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1604        backtracked to try for "b" instead of moving to the next alternative branch
1605        at the same level (in this case, to look for "c"). The Perl documentation
1606        is clear that when (*THEN) is backtracked onto, it goes to the "next
1607        alternative in the innermost enclosing group".
1608    
1609    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1610        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1611        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1612        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1613        (*THEN).
1614    
1615    3.  If \s appeared in a character class, it removed the VT character from
1616        the class, even if it had been included by some previous item, for example
1617        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1618        of \s, but is part of the POSIX "space" class.)
1619    
1620    4.  A partial match never returns an empty string (because you can always
1621        match an empty string at the end of the subject); however the checking for
1622        an empty string was starting at the "start of match" point. This has been
1623        changed to the "earliest inspected character" point, because the returned
1624        data for a partial match starts at this character. This means that, for
1625        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1626        (previously it gave "no match").
1627    
1628    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1629        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1630        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1631        has an implication that the given string is incomplete (because a partial
1632        match is preferred over a full match). For this reason, these items now
1633        give a partial match in this situation. [Aside: previously, the one case
1634        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1635        match rather than a full match, which was wrong by the old rules, but is
1636        now correct.]
1637    
1638    6.  There was a bug in the handling of #-introduced comments, recognized when
1639        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1640        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1641        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1642        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1643        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1644        places in pcre_compile().
1645    
1646    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1647        comments when looking ahead for named forward references to subpatterns,
1648        the only newline sequence it recognized was NL. It now handles newlines
1649        according to the set newline convention.
1650    
1651    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1652        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1653        cater for a lack of strerror(). These oversights have been fixed.
1654    
1655    9.  Added --match-limit and --recursion-limit to pcregrep.
1656    
1657    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1658    
1659    11. When the -o option was used, pcregrep was setting a return code of 1, even
1660        when matches were found, and --line-buffered was not being honoured.
1661    
1662    12. Added an optional parentheses number to the -o and --only-matching options
1663        of pcregrep.
1664    
1665    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1666        can match an empty string. The code to do it in pcretest and pcredemo
1667        needed fixing:
1668    
1669        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1670            only one byte after an empty string match just before CRLF (this case
1671            just got forgotten; "any" and "anycrlf" were OK).
1672    
1673        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1674            mode when an empty string match preceded an ASCII character followed by
1675            a non-ASCII character. (The code for advancing by one character rather
1676            than one byte was nonsense.)
1677    
1678        (c) The pcredemo.c sample program did not have any code at all to handle
1679            the cases when CRLF is a valid newline sequence.
1680    
1681    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1682        as a starting offset was within the subject string. There is now a new
1683        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1684        negative or greater than the length of the string. In order to test this,
1685        pcretest is extended to allow the setting of negative starting offsets.
1686    
1687    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1688        starting offset points to the beginning of a UTF-8 character was
1689        unnecessarily clumsy. I tidied it up.
1690    
1691    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1692        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1693    
1694    17. Nobody had reported that the --include_dir option, which was added in
1695        release 7.7 should have been called --include-dir (hyphen, not underscore)
1696        for compatibility with GNU grep. I have changed it to --include-dir, but
1697        left --include_dir as an undocumented synonym, and the same for
1698        --exclude-dir, though that is not available in GNU grep, at least as of
1699        release 2.5.4.
1700    
1701    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1702        characters from a string of bytes) have been redefined so as not to use
1703        loops, in order to improve performance in some environments. At the same
1704        time, I abstracted some of the common code into auxiliary macros to save
1705        repetition (this should not affect the compiled code).
1706    
1707    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1708        compile-time error is now given if \c is not followed by an ASCII
1709        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1710        different, and any byte value is allowed.)
1711    
1712    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1713        START_OPTIMIZE option, which is now allowed at compile time - but just
1714        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1715        to pcregrep and other applications that have no direct access to PCRE
1716        options. The new /Y option in pcretest sets this option when calling
1717        pcre_compile().
1718    
1719    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1720        back references. Groups containing recursive back references were forced to
1721        be atomic by that change, but in the case of named groups, the amount of
1722        memory required was incorrectly computed, leading to "Failed: internal
1723        error: code overflow". This has been fixed.
1724    
1725    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1726        pcretest.c, to avoid build problems in some Borland environments.
1727    
1728    
1729    Version 8.10 25-Jun-2010
1730    ------------------------
1731    
1732    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1733        THEN.
1734    
1735    2.  (*ACCEPT) was not working when inside an atomic group.
1736    
1737    3.  Inside a character class, \B is treated as a literal by default, but
1738        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1739        causes the error). The code is unchanged, but I tidied the documentation.
1740    
1741    4.  Inside a character class, PCRE always treated \R and \X as literals,
1742        whereas Perl faults them if its -w option is set. I have changed PCRE so
1743        that it faults them when PCRE_EXTRA is set.
1744    
1745    5.  Added support for \N, which always matches any character other than
1746        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1747    
1748    6.  When compiling pcregrep with newer versions of gcc which may have
1749        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1750        declared with attribute warn_unused_result" were given. Just casting the
1751        result to (void) does not stop the warnings; a more elaborate fudge is
1752        needed. I've used a macro to implement this.
1753    
1754    7.  Minor change to pcretest.c to avoid a compiler warning.
1755    
1756    8.  Added four artifical Unicode properties to help with an option to make
1757        \s etc use properties (see next item). The new properties are: Xan
1758        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1759    
1760    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1761        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1762        this option. Modified pcretest to add /W to test this facility. Added
1763        REG_UCP to make it available via the POSIX interface.
1764    
1765    10. Added --line-buffered to pcregrep.
1766    
1767    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1768        studied, and the match started with a letter with a code point greater than
1769        127 whose first byte was different to the first byte of the other case of
1770        the letter, the other case of this starting letter was not recognized
1771        (#976).
1772    
1773    12. If a pattern that was studied started with a repeated Unicode property
1774        test, for example, \p{Nd}+, there was the theoretical possibility of
1775        setting up an incorrect bitmap of starting bytes, but fortunately it could
1776        not have actually happened in practice until change 8 above was made (it
1777        added property types that matched character-matching opcodes).
1778    
1779    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1780        possible starting bytes for non-anchored patterns.
1781    
1782    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1783        \R, and also a number of cases that involve Unicode properties, both
1784        explicit and implicit when PCRE_UCP is set.
1785    
1786    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1787        input, it could crash or give wrong results if characters with values
1788        greater than 0xc0 were present in the subject string. (Detail: it assumed
1789        UTF-8 input when processing these items.)
1790    
1791    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1792        size_t is 64-bit (#991).
1793    
1794    17. Added a check for running out of memory when PCRE is compiled with
1795        --disable-stack-for-recursion (#990).
1796    
1797    18. If the last data line in a file for pcretest does not have a newline on
1798        the end, a newline was missing in the output.
1799    
1800    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1801        less than 128) in its various bitmaps. However, there is a facility for
1802        generating tables according to the current locale when PCRE is compiled. It
1803        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1804        characters, are recognized by isspace() and therefore were getting set in
1805        these tables, and indeed these tables seem to approximate to ISO 8859. This
1806        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1807        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1808        which of course cannot start UTF-8 characters. I have changed the code so
1809        that only real ASCII characters (less than 128) and the correct starting
1810        bytes for UTF-8 encodings are set for characters greater than 127 when in
1811        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1812        altogether.)
1813    
1814    20. Added the /T option to pcretest so as to be able to run tests with non-
1815        standard character tables, thus making it possible to include the tests
1816        used for 19 above in the standard set of tests.
1817    
1818    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1819        reference to a subpattern the other side of a comment that contains an
1820        opening parenthesis caused either an internal compiling error, or a
1821        reference to the wrong subpattern.
1822    
1823    
1824    Version 8.02 19-Mar-2010
1825    ------------------------
1826    
1827    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1828    
1829    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1830        configured.
1831    
1832    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1833        original author of that file, following a query about its status.
1834    
1835    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1836        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1837    
1838    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1839        quantifier applied to a forward-referencing subroutine call, could compile
1840        incorrect code or give the error "internal error: previously-checked
1841        referenced subpattern not found".
1842    
1843    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1844        variables to point to external functions. For these systems, therefore,
1845        pcre_malloc etc. are now initialized to local functions that call the
1846        relevant global functions.
1847    
1848    7.  There were two entries missing in the vectors called coptable and poptable
1849        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1850        I've fixed the data, and added a kludgy way of testing at compile time that
1851        the lengths are correct (equal to the number of opcodes).
1852    
1853    8.  Following on from 7, I added a similar kludge to check the length of the
1854        eint vector in pcreposix.c.
1855    
1856    9.  Error texts for pcre_compile() are held as one long string to avoid too
1857        much relocation at load time. To find a text, the string is searched,
1858        counting zeros. There was no check for running off the end of the string,
1859        which could happen if a new error number was added without updating the
1860        string.
1861    
1862    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1863    
1864    11. \K was not working if it appeared in an atomic group or in a group that
1865        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1866        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1867        the assertion is positive, but not if it is negative.
1868    
1869    12. Change 11 fortuitously reduced the size of the stack frame used in the
1870        "match()" function of pcre_exec.c by one pointer. Forthcoming
1871        implementation of support for (*MARK) will need an extra pointer on the
1872        stack; I have reserved it now, so that the stack frame size does not
1873        decrease.
1874    
1875    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1876        item in branch that calls a recursion is a subroutine call - as in the
1877        second branch in the above example - was incorrectly given the compile-
1878        time error "recursive call could loop indefinitely" because pcre_compile()
1879        was not correctly checking the subroutine for matching a non-empty string.
1880    
1881    14. The checks for overrunning compiling workspace could trigger after an
1882        overrun had occurred. This is a "should never occur" error, but it can be
1883        triggered by pathological patterns such as hundreds of nested parentheses.
1884        The checks now trigger 100 bytes before the end of the workspace.
1885    
1886    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1887    
1888    
1889    Version 8.01 19-Jan-2010
1890    ------------------------
1891    
1892    1.  If a pattern contained a conditional subpattern with only one branch (in
1893        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1894        computed the wrong minimum data length (which is of course zero for such
1895        subpatterns). This could cause incorrect "no match" results.
1896    
1897    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1898        the pattern is reset in the first branch, pcre_compile() failed with
1899        "internal error: code overflow at offset...". This happened only when
1900        the reset was to the original external option setting. (An optimization
1901        abstracts leading options settings into an external setting, which was the
1902        cause of this.)
1903    
1904    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1905        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1906        assertion pattern did not match (meaning that the assertion was true), it
1907        was incorrectly treated as false if the SKIP had been reached during the
1908        matching. This also applied to assertions used as conditions.
1909    
1910    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1911        assertion subpattern, including such a pattern used as a condition,
1912        unpredictable results occurred, instead of the error return
1913        PCRE_ERROR_DFA_UITEM.
1914    
1915    5.  The C++ GlobalReplace function was not working like Perl for the special
1916        situation when an empty string is matched. It now does the fancy magic
1917        stuff that is necessary.
1918    
1919    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1920        removed. (These were left over from very, very early versions of PCRE.)
1921    
1922    7.  Some cosmetic changes to the code to make life easier when compiling it
1923        as part of something else:
1924    
1925        (a) Change DEBUG to PCRE_DEBUG.
1926    
1927        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1928            called "current" as "current_branch", to prevent a collision with the
1929            Linux macro when compiled as a kernel module.
1930    
1931        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1932            prevent a collision with the Linux macro when compiled as a kernel
1933            module.
1934    
1935    8.  In pcre_compile() there are some checks for integer overflows that used to
1936        cast potentially large values to (double). This has been changed to that
1937        when building, a check for int64_t is made, and if it is found, it is used
1938        instead, thus avoiding the use of floating point arithmetic. (There is no
1939        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1940        double.
1941    
1942    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1943        2005 (difference between two addresses compared to an unsigned value).
1944    
1945    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1946        custom one, because of the following reported problem in Windows:
1947    
1948          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1949              under Win32.
1950          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1951              therefore missing the function definition.
1952          - The compiler thus generates a "C" signature for the test function.
1953          - The linker fails to find the "C" function.
1954          - PCRE fails to configure if asked to do so against libbz2.
1955    
1956    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1957        messages were output:
1958    
1959          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1960          rerunning libtoolize, to keep the correct libtool macros in-tree.
1961          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1962    
1963        I have done both of these things.
1964    
1965    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1966        most of the time, it *can* run out if it is given a pattern that contains a
1967        runaway infinite recursion. I updated the discussion in the pcrestack man
1968        page.
1969    
1970    13. Now that we have gone to the x.xx style of version numbers, the minor
1971        version may start with zero. Using 08 or 09 is a bad idea because users
1972        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1973        interpreted as invalid octal numbers. I've updated the previous comment in
1974        configure.ac, and also added a check that gives an error if 08 or 09 are
1975        used.
1976    
1977    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1978        causing partial matching to fail when the end of the subject matched \W
1979        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1980    
1981    15. There were some discrepancies between the declarations in pcre_internal.h
1982        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1983        their definitions. The declarations used "const uschar *" and the
1984        definitions used USPTR. Even though USPTR is normally defined as "const
1985        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1986        reported that: "This difference in casting confuses some C++ compilers, for
1987        example, SunCC recognizes above declarations as different functions and
1988        generates broken code for hbpcre." I have changed the declarations to use
1989        USPTR.
1990    
1991    16. GNU libtool is named differently on some systems. The autogen.sh script now
1992        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1993        (FreeBSD).
1994    
1995    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1996        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1997        comment: "Figure out how to create a longlong from a string: strtoll and
1998        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1999        instance, but it only takes 2 args instead of 3!"
2000    
2001    18. A subtle bug concerned with back references has been fixed by a change of
2002        specification, with a corresponding code fix. A pattern such as
2003        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
2004        refers, was giving matches when it shouldn't. For example, xa=xaaa would
2005        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
2006        same bug. Such groups have to be quantified to be useful, or contained
2007        inside another quantified group. (If there's no repetition, the reference
2008        can never match.) The problem arises because, having left the group and
2009        moved on to the rest of the pattern, a later failure that backtracks into
2010        the group uses the captured value from the final iteration of the group
2011        rather than the correct earlier one. I have fixed this in PCRE by forcing
2012        any group that contains a reference to itself to be an atomic group; that
2013        is, there cannot be any backtracking into it once it has completed. This is
2014        similar to recursive and subroutine calls.
2015    
2016    
2017    Version 8.00 19-Oct-09
2018  ----------------------  ----------------------
2019    
2020  1.  The table for translating pcre_compile() error codes into POSIX error codes  1.  The table for translating pcre_compile() error codes into POSIX error codes
2021      was out-of-date, and there was no check on the pcre_compile() error code      was out-of-date, and there was no check on the pcre_compile() error code
2022      being within the table. This could lead to an OK return being given in      being within the table. This could lead to an OK return being given in
2023      error.      error.
2024    
2025  2.  Changed the call to open a subject file in pcregrep from fopen(pathname,  2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
2026      "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests      "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
2027      in a Windows environment.      in a Windows environment.
2028    
2029  3.  The pcregrep --count option prints the count for each file even when it is  3.  The pcregrep --count option prints the count for each file even when it is
2030      zero, as does GNU grep. However, pcregrep was also printing all files when      zero, as does GNU grep. However, pcregrep was also printing all files when
2031      --files-with-matches was added. Now, when both options are given, it prints      --files-with-matches was added. Now, when both options are given, it prints
2032      counts only for those files that have at least one match. (GNU grep just      counts only for those files that have at least one match. (GNU grep just
2033      prints the file name in this circumstance, but including the count seems      prints the file name in this circumstance, but including the count seems
2034      more useful - otherwise, why use --count?) Also ensured that the      more useful - otherwise, why use --count?) Also ensured that the
2035      combination -clh just lists non-zero counts, with no names.      combination -clh just lists non-zero counts, with no names.
2036    
2037  4.  The long form of the pcregrep -F option was incorrectly implemented as  4.  The long form of the pcregrep -F option was incorrectly implemented as
2038      --fixed_strings instead of --fixed-strings. This is an incompatible change,      --fixed_strings instead of --fixed-strings. This is an incompatible change,
2039      but it seems right to fix it, and I didn't think it was worth preserving      but it seems right to fix it, and I didn't think it was worth preserving
2040      the old behaviour.      the old behaviour.
2041    
2042  5.  The command line items --regex=pattern and --regexp=pattern were not  5.  The command line items --regex=pattern and --regexp=pattern were not
2043      recognized by pcregrep, which required --regex pattern or --regexp pattern      recognized by pcregrep, which required --regex pattern or --regexp pattern
2044      (with a space rather than an '='). The man page documented the '=' forms,      (with a space rather than an '='). The man page documented the '=' forms,
2045      which are compatible with GNU grep; these now work.      which are compatible with GNU grep; these now work.
2046    
2047  6.  No libpcreposix.pc file was created for pkg-config; there was just  6.  No libpcreposix.pc file was created for pkg-config; there was just
2048      libpcre.pc and libpcrecpp.pc. The omission has been rectified.      libpcre.pc and libpcrecpp.pc. The omission has been rectified.
2049    
2050  7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size  7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
2051      when UCP support is not needed, by modifying the Python script that      when UCP support is not needed, by modifying the Python script that
2052      generates it from Unicode data files. This should not matter if the module      generates it from Unicode data files. This should not matter if the module
2053      is correctly used as a library, but I received one complaint about 50K of      is correctly used as a library, but I received one complaint about 50K of
2054      unwanted data. My guess is that the person linked everything into his      unwanted data. My guess is that the person linked everything into his
2055      program rather than using a library. Anyway, it does no harm.      program rather than using a library. Anyway, it does no harm.
2056    
2057  8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger  8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
2058      was a minimum greater than 1 for a wide character in a possessive      was a minimum greater than 1 for a wide character in a possessive
2059      repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8      repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
2060      which had an unlimited repeat of a nested, fixed maximum repeat of a wide      which had an unlimited repeat of a nested, fixed maximum repeat of a wide
2061      character. Chaos in the form of incorrect output or a compiling loop could      character. Chaos in the form of incorrect output or a compiling loop could
2062      result.      result.
2063    
2064  9.  The restrictions on what a pattern can contain when partial matching is  9.  The restrictions on what a pattern can contain when partial matching is
2065      requested for pcre_exec() have been removed. All patterns can now be      requested for pcre_exec() have been removed. All patterns can now be
2066      partially matched by this function. In addition, if there are at least two      partially matched by this function. In addition, if there are at least two
2067      slots in the offset vector, the offsets of the first-encountered partial      slots in the offset vector, the offset of the earliest inspected character
2068      match are set in them when PCRE_ERROR_PARTIAL is returned.      for the match and the offset of the end of the subject are set in them when
2069        PCRE_ERROR_PARTIAL is returned.
2070    
2071  10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is  10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
2072      synonymous with PCRE_PARTIAL, for backwards compatibility, and      synonymous with PCRE_PARTIAL, for backwards compatibility, and
2073      PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,      PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
2074      and may be more useful for multi-segment matching, especially with      and may be more useful for multi-segment matching.
2075      pcre_exec().  
2076    11. Partial matching with pcre_exec() is now more intuitive. A partial match
2077  11. Partial matching with pcre_exec() is now more intuitive. A partial match      used to be given if ever the end of the subject was reached; now it is
2078      used to be given if ever the end of the subject was reached; now it is      given only if matching could not proceed because another character was
2079      given only if matching could not proceed because another character was      needed. This makes a difference in some odd cases such as Z(*FAIL) with the
2080      needed. This makes a difference in some odd cases such as Z(*FAIL) with the      string "Z", which now yields "no match" instead of "partial match". In the
2081      string "Z", which now yields "no match" instead of "partial match". In the      case of pcre_dfa_exec(), "no match" is given if every matching path for the
2082      case of pcre_dfa_exec(), "no match" is given if every matching path for the      final character ended with (*FAIL).
2083      final character ended with (*FAIL).  
   
2084  12. Restarting a match using pcre_dfa_exec() after a partial match did not work  12. Restarting a match using pcre_dfa_exec() after a partial match did not work
2085      if the pattern had a "must contain" character that was already found in the      if the pattern had a "must contain" character that was already found in the
2086      earlier partial match, unless partial matching was again requested. For      earlier partial match, unless partial matching was again requested. For
2087      example, with the pattern /dog.(body)?/, the "must contain" character is      example, with the pattern /dog.(body)?/, the "must contain" character is
2088      "g". If the first part-match was for the string "dog", restarting with      "g". If the first part-match was for the string "dog", restarting with
2089      "sbody" failed.      "sbody" failed. This bug has been fixed.
2090    
2091  13. Added a pcredemo man page, created automatically from the pcredemo.c file,  13. The string returned by pcre_dfa_exec() after a partial match has been
2092      so that the demonstration program is easily available in environments where      changed so that it starts at the first inspected character rather than the
2093      PCRE has not been installed from source.      first character of the match. This makes a difference only if the pattern
2094        starts with a lookbehind assertion or \b or \B (\K is not supported by
2095  14. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,      pcre_dfa_exec()). It's an incompatible change, but it makes the two
2096        matching functions compatible, and I think it's the right thing to do.
2097    
2098    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
2099        so that the demonstration program is easily available in environments where
2100        PCRE has not been installed from source.
2101    
2102    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
2103      libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared      libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
2104      library.      library.
2105    
2106    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
2107        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
2108        is not the first non-POSIX option to be added. Clearly some people find
2109        these options useful.
2110    
2111    17. If a caller to the POSIX matching function regexec() passes a non-zero
2112        value for nmatch with a NULL value for pmatch, the value of
2113        nmatch is forced to zero.
2114    
2115    18. RunGrepTest did not have a test for the availability of the -u option of
2116        the diff command, as RunTest does. It now checks in the same way as
2117        RunTest, and also checks for the -b option.
2118    
2119    19. If an odd number of negated classes containing just a single character
2120        interposed, within parentheses, between a forward reference to a named
2121        subpattern and the definition of the subpattern, compilation crashed with
2122        an internal error, complaining that it could not find the referenced
2123        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
2124        [The bug was that it was starting one character too far in when skipping
2125        over the character class, thus treating the ] as data rather than
2126        terminating the class. This meant it could skip too much.]
2127    
2128    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
2129        /g option in pcretest when the pattern contains \K, which makes it possible
2130        to have an empty string match not at the start, even when the pattern is
2131        anchored. Updated pcretest and pcredemo to use this option.
2132    
2133    21. If the maximum number of capturing subpatterns in a recursion was greater
2134        than the maximum at the outer level, the higher number was returned, but
2135        with unset values at the outer level. The correct (outer level) value is
2136        now given.
2137    
2138    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
2139        PCRE did not set those parentheses (unlike Perl). I have now found a way to
2140        make it do so. The string so far is captured, making this feature
2141        compatible with Perl.
2142    
2143    23. The tests have been re-organized, adding tests 11 and 12, to make it
2144        possible to check the Perl 5.10 features against Perl 5.10.
2145    
2146    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
2147        pattern matches a fixed length string. PCRE did not allow this; now it
2148        does. Neither allows recursion.
2149    
2150    25. I finally figured out how to implement a request to provide the minimum
2151        length of subject string that was needed in order to match a given pattern.
2152        (It was back references and recursion that I had previously got hung up
2153        on.) This code has now been added to pcre_study(); it finds a lower bound
2154        to the length of subject needed. It is not necessarily the greatest lower
2155        bound, but using it to avoid searching strings that are too short does give
2156        some useful speed-ups. The value is available to calling programs via
2157        pcre_fullinfo().
2158    
2159    26. While implementing 25, I discovered to my embarrassment that pcretest had
2160        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
2161        study optimizations had never been tested with that matching function.
2162        Oops. What is worse, even when it was passed study data, there was a bug in
2163        pcre_dfa_exec() that meant it never actually used it. Double oops. There
2164        were also very few tests of studied patterns with pcre_dfa_exec().
2165    
2166    27. If (?| is used to create subpatterns with duplicate numbers, they are now
2167        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
2168        on the other side of the coin, they are no longer allowed to have different
2169        names, because these cannot be distinguished in PCRE, and this has caused
2170        confusion. (This is a difference from Perl.)
2171    
2172    28. When duplicate subpattern names are present (necessarily with different
2173        numbers, as required by 27 above), and a test is made by name in a
2174        conditional pattern, either for a subpattern having been matched, or for
2175        recursion in such a pattern, all the associated numbered subpatterns are
2176        tested, and the overall condition is true if the condition is true for any
2177        one of them. This is the way Perl works, and is also more like the way
2178        testing by number works.
2179    
2180    
2181  Version 7.9 11-Apr-09  Version 7.9 11-Apr-09
2182  ---------------------  ---------------------
# Line 118  Version 7.9 11-Apr-09 Line 2212  Version 7.9 11-Apr-09
2212  7.  A pattern that could match an empty string could cause pcregrep to loop; it  7.  A pattern that could match an empty string could cause pcregrep to loop; it
2213      doesn't make sense to accept an empty string match in pcregrep, so I have      doesn't make sense to accept an empty string match in pcregrep, so I have
2214      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2215      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2216        8.33.]
2217    
2218  8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at  8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2219      start or after a newline", because the conditional assertion was not being      start or after a newline", because the conditional assertion was not being
# Line 361  Version 7.7 07-May-08 Line 2456  Version 7.7 07-May-08
2456      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2457      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2458      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2459      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2460      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2461      was subsequently set up correctly.)      was subsequently set up correctly.)
2462    

Legend:
Removed from v.431  
changed lines
  Added in v.1534

  ViewVC Help
Powered by ViewVC 1.1.5