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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.341  
changed lines
  Added in v.1526

  ViewVC Help
Powered by ViewVC 1.1.5