ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 483 by ph10, Mon Jan 4 16:19:17 2010 UTC revision 1585 by ph10, Wed Aug 5 15:38:32 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4  Version 8.01 11-Dec-09  Note that the PCRE 8.xx series (PCRE1) is now in a bugfix-only state. All
5  ----------------------  development is happening in the PCRE2 10.xx series.
7    Version 8.38 xx-xxx-xxxx
8    ------------------------
10    1.  If a group that contained a recursive back reference also contained a
11        forward reference subroutine call followed by a non-forward-reference
12        subroutine call, for example /.((?2)(?R)\1)()/, pcre2_compile() failed to
13        compile correct code, leading to undefined behaviour or an internally
14        detected error. This bug was discovered by the LLVM fuzzer.
16    2.  Quantification of certain items (e.g. atomic back references) could cause
17        incorrect code to be compiled when recursive forward references were
18        involved. For example, in this pattern: /(?1)()((((((\1++))\x85)+)|))/.
19        This bug was discovered by the LLVM fuzzer.
21    3.  A repeated conditional group whose condition was a reference by name caused
22        a buffer overflow if there was more than one group with the given name.
23        This bug was discovered by the LLVM fuzzer.
25    4.  A recursive back reference by name within a group that had the same name as
26        another group caused a buffer overflow. For example:
27        /(?J)(?'d'(?'d'\g{d}))/. This bug was discovered by the LLVM fuzzer.
29    5.  A forward reference by name to a group whose number is the same as the
30        current group, for example in this pattern: /(?|(\k'Pm')|(?'Pm'))/, caused
31        a buffer overflow at compile time. This bug was discovered by the LLVM
32        fuzzer.
34    6.  A lookbehind assertion within a set of mutually recursive subpatterns could
35        provoke a buffer overflow. This bug was discovered by the LLVM fuzzer.
37    7.  Another buffer overflow bug involved duplicate named groups with a
38        reference between their definition, with a group that reset capture
39        numbers, for example: /(?J:(?|(?'R')(\k'R')|((?'R'))))/. This has been
40        fixed by always allowing for more memory, even if not needed. (A proper fix
41        is implemented in PCRE2, but it involves more refactoring.)
43    8.  There was no check for integer overflow in subroutine calls such as (?123).
45    9.  The table entry for \l in EBCDIC environments was incorrect, leading to its
46        being treated as a literal 'l' instead of causing an error.
48    10. There was a buffer overflow if pcre_exec() was called with an ovector of
49        size 1. This bug was found by american fuzzy lop.
51    11. If a non-capturing group containing a conditional group that could match
52        an empty string was repeated, it was not identified as matching an empty
53        string itself. For example: /^(?:(?(1)x|)+)+$()/.
55    12. In an EBCDIC environment, pcretest was mishandling the escape sequences
56        \a and \e in test subject lines.
58    13. In an EBCDIC environment, \a in a pattern was converted to the ASCII
59        instead of the EBCDIC value.
61    14. The handling of \c in an EBCDIC environment has been revised so that it is
62        now compatible with the specification in Perl's perlebcdic page.
64    15. The EBCDIC character 0x41 is a non-breaking space, equivalent to 0xa0 in
65        ASCII/Unicode. This has now been added to the list of characters that are
66        recognized as white space in EBCDIC.
68    16. When PCRE was compiled without UCP support, the use of \p and \P gave an
69        error (correctly) when used outside a class, but did not give an error
70        within a class.
72    17. \h within a class was incorrectly compiled in EBCDIC environments.
74    18. A pattern with an unmatched closing parenthesis that contained a backward
75        assertion which itself contained a forward reference caused buffer
76        overflow. And example pattern is: /(?=di(?<=(?1))|(?=(.))))/.
78    19. JIT should return with error when the compiled pattern requires more stack
79        space than the maximum.
81    20. A possessively repeated conditional group that could match an empty string,
82        for example, /(?(R))*+/, was incorrectly compiled.
84    21. Fix infinite recursion in the JIT compiler when certain patterns such as
85        /(?:|a|){100}x/ are analysed.
87    22. Some patterns with character classes involving [: and \\ were incorrectly
88        compiled and could cause reading from uninitialized memory or an incorrect
89        error diagnosis.
91    23. Pathological patterns containing many nested occurrences of [: caused
92        pcre_compile() to run for a very long time.
94    24. A conditional group with only one branch has an implicit empty alternative
95        branch and must therefore be treated as potentially matching an empty
96        string.
98    25. If (?R was followed by - or + incorrect behaviour happened instead of a
99        diagnostic.
101    26. Arrange to give up on finding the minimum matching length for overly
102        complex patterns.
104    27. Similar to (4) above: in a pattern with duplicated named groups and an
105        occurrence of (?| it is possible for an apparently non-recursive back
106        reference to become recursive if a later named group with the relevant
107        number is encountered. This could lead to a buffer overflow. Wen Guanxing
108        from Venustech ADLAB discovered this bug.
111    Version 8.37 28-April-2015
112    --------------------------
114    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
115        for those parentheses to be closed with whatever has been captured so far.
116        However, it was failing to mark any other groups between the hightest
117        capture so far and the currrent group as "unset". Thus, the ovector for
118        those groups contained whatever was previously there. An example is the
119        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
121    2.  If an assertion condition was quantified with a minimum of zero (an odd
122        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
124    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
125        unrecognized modifier, a crash could occur.
127    4.  An attempt to do global matching in pcretest with a zero-length ovector
128        caused a crash.
130    5.  Fixed a memory leak during matching that could occur for a subpattern
131        subroutine call (recursive or otherwise) if the number of captured groups
132        that had to be saved was greater than ten.
134    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
135        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
136        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
138    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
139        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
141    8. When a pattern is compiled, it remembers the highest back reference so that
142       when matching, if the ovector is too small, extra memory can be obtained to
143       use instead. A conditional subpattern whose condition is a check on a
144       capture having happened, such as, for example in the pattern
145       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
146       setting the highest backreference number. This mattered only if pcre_exec()
147       was called with an ovector that was too small to hold the capture, and there
148       was no other kind of back reference (a situation which is probably quite
149       rare). The effect of the bug was that the condition was always treated as
150       FALSE when the capture could not be consulted, leading to a incorrect
151       behaviour by pcre_exec(). This bug has been fixed.
153    9. A reference to a duplicated named group (either a back reference or a test
154       for being set in a conditional) that occurred in a part of the pattern where
155       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
156       to be incorrectly calculated, leading to overwriting.
158    10. A mutually recursive set of back references such as (\2)(\1) caused a
159        segfault at study time (while trying to find the minimum matching length).
160        The infinite loop is now broken (with the minimum length unset, that is,
161        zero).
163    11. If an assertion that was used as a condition was quantified with a minimum
164        of zero, matching went wrong. In particular, if the whole group had
165        unlimited repetition and could match an empty string, a segfault was
166        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
167        assertions to be quantified, but not if they are being used as conditions,
168        so the above pattern is faulted by Perl. PCRE has now been changed so that
169        it also rejects such patterns.
171    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
172        failed to allow the zero-repeat case if pcre2_exec() was called with an
173        ovector too small to capture the group.
175    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
176        Red Hat Product Security:
178        (a) A crash if /K and /F were both set with the option to save the compiled
179        pattern.
181        (b) Another crash if the option to print captured substrings in a callout
182        was combined with setting a null ovector, for example \O\C+ as a subject
183        string.
185    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
186        forward reference repeated a large (but limited) number of times within a
187        repeated outer group that has a zero minimum quantifier, caused incorrect
188        code to be compiled, leading to the error "internal error:
189        previously-checked referenced subpattern not found" when an incorrect
190        memory address was read. This bug was reported as "heap overflow",
191        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
192        CVE-2015-2325.
194    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
195        call within a group that also contained a recursive back reference caused
196        incorrect code to be compiled. This bug was reported as "heap overflow",
197        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
198        number CVE-2015-2326.
200    24. Computing the size of the JIT read-only data in advance has been a source
201        of various issues, and new ones are still appear unfortunately. To fix
202        existing and future issues, size computation is eliminated from the code,
203        and replaced by on-demand memory allocation.
205    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
206        adjacent to the end of the range, and the range contained characters with
207        more than one other case, caused incorrect behaviour when compiled in UTF
208        mode. In that example, the range a-j was left out of the class.
210    26. Fix JIT compilation of conditional blocks, which assertion
211        is converted to (*FAIL). E.g: /(?(?!))/.
213    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
214        discovered by the LLVM fuzzer.
216    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
217        when this assertion was used as a condition, for example (?(?!)a|b). In
218        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
219        error about an unsupported item.
221    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
222        possessification code could take exponential time to complete. A recursion
223        depth limit of 1000 has been imposed to limit the resources used by this
224        optimization.
226    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
227        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
228        because \S ensures they are all in the class. The code for doing this was
229        interacting badly with the code for computing the amount of space needed to
230        compile the pattern, leading to a buffer overflow. This bug was discovered
231        by the LLVM fuzzer.
233    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
234        other kinds of group caused stack overflow at compile time. This bug was
235        discovered by the LLVM fuzzer.
237    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
238        between a subroutine call and its quantifier was incorrectly compiled,
239        leading to buffer overflow or other errors. This bug was discovered by the
240        LLVM fuzzer.
242    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
243        assertion after (?(. The code was failing to check the character after
244        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
245        was discovered by the LLVM fuzzer.
247    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
248        a fixed maximum following a group that contains a subroutine reference was
249        incorrectly compiled and could trigger buffer overflow. This bug was
250        discovered by the LLVM fuzzer.
252    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
253        caused a stack overflow instead of the diagnosis of a non-fixed length
254        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
256    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
257        (e.g. /(?<=\Ka)/) could make pcregrep loop.
259    37. There was a similar problem to 36 in pcretest for global matches.
261    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
262        and a subsequent item in the pattern caused a non-match, backtracking over
263        the repeated \X did not stop, but carried on past the start of the subject,
264        causing reference to random memory and/or a segfault. There were also some
265        other cases where backtracking after \C could crash. This set of bugs was
266        discovered by the LLVM fuzzer.
268    39. The function for finding the minimum length of a matching string could take
269        a very long time if mutual recursion was present many times in a pattern,
270        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
271        method has been implemented. This infelicity was discovered by the LLVM
272        fuzzer.
274    40. Static linking against the PCRE library using the pkg-config module was
275        failing on missing pthread symbols.
278    Version 8.36 26-September-2014
279    ------------------------------
281    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
282        -Wmissing-field-initializers and -Wunused-parameter.
284    2.  The tests for quantifiers being too big (greater than 65535) were being
285        applied after reading the number, and stupidly assuming that integer
286        overflow would give a negative number. The tests are now applied as the
287        numbers are read.
289    3.  Tidy code in pcre_exec.c where two branches that used to be different are
290        now the same.
292    4.  The JIT compiler did not generate match limit checks for certain
293        bracketed expressions with quantifiers. This may lead to exponential
294        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
295        issue should be resolved now.
297    5.  Fixed an issue, which occures when nested alternatives are optimized
298        with table jumps.
300    6.  Inserted two casts and changed some ints to size_t in the light of some
301        reported 64-bit compiler warnings (Bugzilla 1477).
303    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
304        an empty string, which sometimes were behaving incorrectly in the
305        interpreter (though correctly in the JIT matcher). This pcretest input is
306        an example:
308          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
311        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
312        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
313        for an empty string was breaking the inner loop and carrying on at a lower
314        level, when possessive repeated groups should always return to a higher
315        level as they have no backtrack points in them. The empty string test now
316        occurs at the outer level.
318    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
319        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
321    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
322        doesn't).
324    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
325        to hold a pattern to be incorrectly computed (too small) when there were
326        named back references to duplicated names. This could cause "internal
327        error: code overflow" or "double free or corruption" or other memory
328        handling errors.
330    11. When named subpatterns had the same prefixes, back references could be
331        confused. For example, in this pattern:
333          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
335        the reference to 'Name' was incorrectly treated as a reference to a
336        duplicate name.
338    12. A pattern such as /^s?c/mi8 where the optional character has more than
339        one "other case" was incorrectly compiled such that it would only try to
340        match starting at "c".
342    13. When a pattern starting with \s was studied, VT was not included in the
343        list of possible starting characters; this should have been part of the
344        8.34/18 patch.
346    14. If a character class started [\Qx]... where x is any character, the class
347        was incorrectly terminated at the ].
349    15. If a pattern that started with a caseless match for a character with more
350        than one "other case" was studied, PCRE did not set up the starting code
351        unit bit map for the list of possible characters. Now it does. This is an
352        optimization improvement, not a bug fix.
354    16. The Unicode data tables have been updated to Unicode 7.0.0.
356    17. Fixed a number of memory leaks in pcregrep.
358    18. Avoid a compiler warning (from some compilers) for a function call with
359        a cast that removes "const" from an lvalue by using an intermediate
360        variable (to which the compiler does not object).
362    19. Incorrect code was compiled if a group that contained an internal recursive
363        back reference was optional (had quantifier with a minimum of zero). This
364        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
365        caused segmentation faults because of stack overflows at compile time.
367    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
368        group that is quantified with an indefinite repeat, caused a compile-time
369        loop which used up all the system stack and provoked a segmentation fault.
370        This was not the same bug as 19 above.
372    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
373        Patch by Mike Frysinger.
376    Version 8.35 04-April-2014
377    --------------------------
379    1.  A new flag is set, when property checks are present in an XCLASS.
380        When this flag is not set, PCRE can perform certain optimizations
381        such as studying these XCLASS-es.
383    2.  The auto-possessification of character sets were improved: a normal
384        and an extended character set can be compared now. Furthermore
385        the JIT compiler optimizes more character set checks.
387    3.  Got rid of some compiler warnings for potentially uninitialized variables
388        that show up only when compiled with -O2.
390    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
391        of a match later then the end of the match. The pcretest program was not
392        handling the case sensibly - it was outputting from the start to the next
393        binary zero. It now reports this situation in a message, and outputs the
394        text from the end to the start.
396    5.  Fast forward search is improved in JIT. Instead of the first three
397        characters, any three characters with fixed position can be searched.
398        Search order: first, last, middle.
400    6.  Improve character range checks in JIT. Characters are read by an inprecise
401        function now, which returns with an unknown value if the character code is
402        above a certain threshold (e.g: 256). The only limitation is that the value
403        must be bigger than the threshold as well. This function is useful when
404        the characters above the threshold are handled in the same way.
406    7.  The macros whose names start with RAWUCHAR are placeholders for a future
407        mode in which only the bottom 21 bits of 32-bit data items are used. To
408        make this more memorable for those maintaining the code, the names have
409        been changed to start with UCHAR21, and an extensive comment has been added
410        to their definition.
412    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
413        to the export list in Makefile.am (they were accidentally omitted from the
414        8.34 tarball).
416    9.  The informational output from pcretest used the phrase "starting byte set"
417        which is inappropriate for the 16-bit and 32-bit libraries. As the output
418        for "first char" and "need char" really means "non-UTF-char", I've changed
419        "byte" to "char", and slightly reworded the output. The documentation about
420        these values has also been (I hope) clarified.
422    10. Another JIT related optimization: use table jumps for selecting the correct
423        backtracking path, when more than four alternatives are present inside a
424        bracket.
426    11. Empty match is not possible, when the minimum length is greater than zero,
427        and there is no \K in the pattern. JIT should avoid empty match checks in
428        such cases.
430    12. In a caseless character class with UCP support, when a character with more
431        than one alternative case was not the first character of a range, not all
432        the alternative cases were added to the class. For example, s and \x{17f}
433        are both alternative cases for S: the class [RST] was handled correctly,
434        but [R-T] was not.
436    13. The configure.ac file always checked for pthread support when JIT was
437        enabled. This is not used in Windows, so I have put this test inside a
438        check for the presence of windows.h (which was already tested for).
440    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
441        The algorithm provides a way to skip certain starting offsets, and usually
442        faster than linear prefix searches.
444    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
445        as for 'fr_FR' and 'french'. For some reason, however, it then used the
446        Windows-specific input and output files, which have 'french' screwed in.
447        So this could never have worked. One of the problems with locales is that
448        they aren't always the same. I have now updated RunTest so that it checks
449        the output of the locale test (test 3) against three different output
450        files, and it allows the test to pass if any one of them matches. With luck
451        this should make the test pass on some versions of Solaris where it was
452        failing. Because of the uncertainty, the script did not used to stop if
453        test 3 failed; it now does. If further versions of a French locale ever
454        come to light, they can now easily be added.
456    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
457        there was a message during ./configure, but it did not stop. This now
458        provokes an error. The invalid example in README has been corrected.
459        If a value less than the minimum is given, the minimum value has always
460        been used, but now a warning is given.
462    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
463        was a bug in the test system, which is now fixed. Also, the list of various
464        configurations that are tested for each release did not have one with both
465        16/32 bits and --enable-bar-anycrlf. It now does.
467    18. pcretest was missing "-C bsr" for displaying the \R default setting.
469    19. Little endian PowerPC systems are supported now by the JIT compiler.
471    20. The fast forward newline mechanism could enter to an infinite loop on
472        certain invalid UTF-8 input. Although we don't support these cases
473        this issue can be fixed by a performance optimization.
475    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
476        not take account if existing stack usage. There is now a new global
477        variable called pcre_stack_guard that can be set to point to an external
478        function to check stack availability. It is called at the start of
479        processing every parenthesized group.
481    22. A typo in the code meant that in ungreedy mode the max/min qualifier
482        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
483        not match "ab".
485    23. When UTF was disabled, the JIT program reported some incorrect compile
486        errors. These messages are silenced now.
488    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
489        compiler.
491    25. Change all the temporary files used in RunGrepTest to be different to those
492        used by RunTest so that the tests can be run simultaneously, for example by
493        "make -j check".
496    Version 8.34 15-December-2013
497    -----------------------------
499    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
500        executable memory. Patch inspired by Carsten Klein.
502    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
503        this macro is never tested and has no effect, because the work to support
504        coverage involves only compiling and linking options and special targets in
505        the Makefile. The comment in config.h implied that defining the macro would
506        enable coverage support, which is totally false. There was also support for
507        setting this macro in the CMake files (my fault, I just copied it from
508        configure). SUPPORT_GCOV has now been removed.
510    3.  Make a small performance improvement in strlen16() and strlen32() in
511        pcretest.
513    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
514        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
515        these cases). There was also one in pcretest.c.
517    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
519    6.  In UTF mode, the code for checking whether a group could match an empty
520        string (which is used for indefinitely repeated groups to allow for
521        breaking an infinite loop) was broken when the group contained a repeated
522        negated single-character class with a character that occupied more than one
523        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
524        UTF-8 mode). The effect was undefined: the group might or might not be
525        deemed as matching an empty string, or the program might have crashed.
527    7.  The code for checking whether a group could match an empty string was not
528        recognizing that \h, \H, \v, \V, and \R must match a character.
530    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
531        an empty string. If it can, pcretest shows this in its information output.
533    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
534        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
535        matched by pcre_exec() without using JIT:
537        (a) If the rest of the pattern did not match after a maximal run of
538            grapheme clusters, the code for backing up to try with fewer of them
539            did not always back up over a full grapheme when characters that do not
540            have the modifier quality were involved, e.g. Hangul syllables.
542        (b) If the match point in a subject started with modifier character, and
543            there was no match, the code could incorrectly back up beyond the match
544            point, and potentially beyond the first character in the subject,
545            leading to a segfault or an incorrect match result.
547    10. A conditional group with an assertion condition could lead to PCRE
548        recording an incorrect first data item for a match if no other first data
549        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
550        first data item, and therefore matched "ca" after "c" instead of at the
551        start.
553    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
554        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
556    12. The source of pcregrep now includes z/OS-specific code so that it can be
557        compiled for z/OS as part of the special z/OS distribution.
559    13. Added the -T and -TM options to pcretest.
561    14. The code in pcre_compile.c for creating the table of named capturing groups
562        has been refactored. Instead of creating the table dynamically during the
563        actual compiling pass, the information is remembered during the pre-compile
564        pass (on the stack unless there are more than 20 named groups, in which
565        case malloc() is used) and the whole table is created before the actual
566        compile happens. This has simplified the code (it is now nearly 150 lines
567        shorter) and prepared the way for better handling of references to groups
568        with duplicate names.
570    15. A back reference to a named subpattern when there is more than one of the
571        same name now checks them in the order in which they appear in the pattern.
572        The first one that is set is used for the reference. Previously only the
573        first one was inspected. This change makes PCRE more compatible with Perl.
575    16. Unicode character properties were updated from Unicode 6.3.0.
577    17. The compile-time code for auto-possessification has been refactored, based
578        on a patch by Zoltan Herczeg. It now happens after instead of during
579        compilation. The code is cleaner, and more cases are handled. The option
580        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
581        options in pcretest are provided to set it. It can also be set by
582        (*NO_AUTO_POSSESS) at the start of a pattern.
584    18. The character VT has been added to the default ("C" locale) set of
585        characters that match \s and are generally treated as white space,
586        following this same change in Perl 5.18. There is now no difference between
587        "Perl space" and "POSIX space". Whether VT is treated as white space in
588        other locales depends on the locale.
590    19. The code for checking named groups as conditions, either for being set or
591        for being recursed, has been refactored (this is related to 14 and 15
592        above). Processing unduplicated named groups should now be as fast at
593        numerical groups, and processing duplicated groups should be faster than
594        before.
596    20. Two patches to the CMake build system, by Alexander Barkov:
598          (1) Replace the "source" command by "." in CMakeLists.txt because
599              "source" is a bash-ism.
601          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
602              without these the CMake build does not work on Solaris.
604    21. Perl has changed its handling of \8 and \9. If there is no previously
605        encountered capturing group of those numbers, they are treated as the
606        literal characters 8 and 9 instead of a binary zero followed by the
607        literals. PCRE now does the same.
609    22. Following Perl, added \o{} to specify codepoints in octal, making it
610        possible to specify values greater than 0777 and also making them
611        unambiguous.
613    23. Perl now gives an error for missing closing braces after \x{... instead of
614        treating the string as literal. PCRE now does the same.
616    24. RunTest used to grumble if an inappropriate test was selected explicitly,
617        but just skip it when running all tests. This make it awkward to run ranges
618        of tests when one of them was inappropriate. Now it just skips any
619        inappropriate tests, as it always did when running all tests.
621    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
622        character types such as \d or \w, too many callouts were inserted, and the
623        data that they returned was rubbish.
625    26. In UCP mode, \s was not matching two of the characters that Perl matches,
626        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
627        were matched by \h. The code has now been refactored so that the lists of
628        the horizontal and vertical whitespace characters used for \h and \v (which
629        are defined only in one place) are now also used for \s.
631    27. Add JIT support for the 64 bit TileGX architecture.
632        Patch by Jiong Wang (Tilera Corporation).
634    28. Possessive quantifiers for classes (both explicit and automatically
635        generated) now use special opcodes instead of wrapping in ONCE brackets.
637    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
638        (because it's meaningless), this was not happening when PCRE_CASELESS was
639        set. Not wrong, but inefficient.
641    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
642        properties for \w, \d, etc) is present in a test regex. Otherwise if the
643        test contains no characters greater than 255, Perl doesn't realise it
644        should be using Unicode semantics.
646    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
647        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
648        does in Unicode mode.
650    32. Added the "forbid" facility to pcretest so that putting tests into the
651        wrong test files can sometimes be quickly detected.
653    33. There is now a limit (default 250) on the depth of nesting of parentheses.
654        This limit is imposed to control the amount of system stack used at compile
655        time. It can be changed at build time by --with-parens-nest-limit=xxx or
656        the equivalent in CMake.
658    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
659        errors. Perl warns for these when in warning mode, but PCRE has no facility
660        for giving warnings.
662    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
663        However, this was not working for (?!) because it is optimized to (*FAIL),
664        for which PCRE does not allow quantifiers. The optimization is now disabled
665        when a quantifier follows (?!). I can't see any use for this, but it makes
666        things uniform.
668    36. Perl no longer allows group names to start with digits, so I have made this
669        change also in PCRE. It simplifies the code a bit.
671    37. In extended mode, Perl ignores spaces before a + that indicates a
672        possessive quantifier. PCRE allowed a space before the quantifier, but not
673        before the possessive +. It now does.
675    38. The use of \K (reset reported match start) within a repeated possessive
676        group such as (a\Kb)*+ was not working.
678    40. Document that the same character tables must be used at compile time and
679        run time, and that the facility to pass tables to pcre_exec() and
680        pcre_dfa_exec() is for use only with saved/restored patterns.
682    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
683        features for Builds with MSVC:
685        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
686           builds). The .rc files can be used to set FileDescription and many other
687           attributes.
689        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
690           This allows higher-level build scripts which want .pdb files to avoid
691           hard-coding the exact files needed."
693    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
694        mean "start of word" and "end of word", respectively, as a transition aid.
696    43. A minimizing repeat of a class containing codepoints greater than 255 in
697        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
698        compiled to use the heap for recursion.
700    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
701        is configured.
704    Version 8.33 28-May-2013
705    ------------------------
707    1.  Added 'U' to some constants that are compared to unsigned integers, to
708        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
709        variables that are added to signed variables, to ensure the result is
710        signed and can be negated.
712    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
713        CMake config files.
715    3.  Revise the creation of config.h.generic so that all boolean macros are
716        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
717        overriding via -D on the command line possible.
719    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
720        to unsigned int is reported to make a quite noticeable speed difference in
721        a specific Windows environment. Testing on Linux did also appear to show
722        some benefit (and it is clearly not harmful). Also fixed the definition of
723        Xop which should be unsigned.
725    5.  Related to (4), changing the definition of the intermediate variable cc
726        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
727        improvements.
729    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
730        unnecessary spaces.
732    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
733        and later.
735    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
737    9.  Optimizing fast_forward_start_bits in JIT.
739    10. Adding support for callouts in JIT, and fixing some issues revealed
740        during this work. Namely:
742        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
744        (b) Minimum length was not checked before the matching is started.
746    11. The value of capture_last that is passed to callouts was incorrect in some
747        cases when there was a capture on one path that was subsequently abandoned
748        after a backtrack. Also, the capture_last value is now reset after a
749        recursion, since all captures are also reset in this case.
751    12. The interpreter no longer returns the "too many substrings" error in the
752        case when an overflowing capture is in a branch that is subsequently
753        abandoned after a backtrack.
755    13. In the pathological case when an offset vector of size 2 is used, pcretest
756        now prints out the matched string after a yield of 0 or 1.
758    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
759        Only supported by the JIT compiler at the moment.
761    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
763    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
764        offset of the starting point of the matching process, provided the offsets
765        vector is large enough.
767    17. The \A escape now records a lookbehind value of 1, though its execution
768        does not actually inspect the previous character. This is to ensure that,
769        in partial multi-segment matching, at least one character from the old
770        segment is retained when a new segment is processed. Otherwise, if there
771        are no lookbehinds in the pattern, \A might match incorrectly at the start
772        of a new segment.
774    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
776    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
777        gives some modest performance improvement in 8-bit mode.
779    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
780        be expressed in certain programming languages using Universal Character
781        Names.
783    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
784        which points out that "non characters" are not "characters that may not
785        appear in Unicode strings" but rather "characters that are reserved for
786        internal use and have only local meaning".
788    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
789        there was a conditional group that depended on an assertion, if the
790        assertion was false, the callout that immediately followed the alternation
791        in the condition was skipped when pcre_exec() was used for matching.
793    23. Allow an explicit callout to be inserted before an assertion that is the
794        condition for a conditional group, for compatibility with automatic
795        callouts, which always insert a callout at this point.
797    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
798        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
800    25. (*PRUNE) is now supported by the JIT compiler.
802    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
804    27. Fix the case where there are two or more SKIPs with arguments that may be
805        ignored.
807    28. (*SKIP) is now supported by the JIT compiler.
809    29. (*THEN) is now supported by the JIT compiler.
811    30. Update RunTest with additional test selector options.
813    31. The way PCRE handles backtracking verbs has been changed in two ways.
815        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
816        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
817        backtracking. In some cases this makes it more Perl-compatible, but Perl's
818        rather obscure rules do not always do the same thing.
820        (2) Previously, backtracking verbs were confined within assertions. This is
821        no longer the case for positive assertions, except for (*ACCEPT). Again,
822        this sometimes improves Perl compatibility, and sometimes does not.
824    32. A number of tests that were in test 2 because Perl did things differently
825        have been moved to test 1, because either Perl or PCRE has changed, and
826        these tests are now compatible.
828    32. Backtracking control verbs are now handled in the same way in JIT and
829        interpreter.
831    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
832        contained a forward subroutine reference caused a compile error.
834    34. Auto-detect and optimize limited repetitions in JIT.
836    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
837        blocking (*UTF) etc.
839    36. In the interpreter, maximizing pattern repetitions for characters and
840        character types now use tail recursion, which reduces stack usage.
842    37. The value of the max lookbehind was not correctly preserved if a compiled
843        and saved regex was reloaded on a host of different endianness.
845    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
846        of the compiled pattern block, expand the flags field from 16 to 32 bits
847        because it was almost full.
849    39. Try madvise first before posix_madvise.
851    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
852        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
853        original change locked out all matches of empty strings. This has been
854        changed so that one match of an empty string per line is recognized.
855        Subsequent searches on the same line (for colouring or for --only-matching,
856        for example) do not recognize empty strings.
858    41. Applied a user patch to fix a number of spelling mistakes in comments.
860    42. Data lines longer than 65536 caused pcretest to crash.
862    43. Clarified the data type for length and startoffset arguments for pcre_exec
863        and pcre_dfa_exec in the function-specific man pages, where they were
864        explicitly stated to be in bytes, never having been updated. I also added
865        some clarification to the pcreapi man page.
867    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
868        a segmentation fault.
871    Version 8.32 30-November-2012
872    -----------------------------
874    1.  Improved JIT compiler optimizations for first character search and single
875        character iterators.
877    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
878        Patch by Daniel Richard G.
880    3.  Single character iterator optimizations in the JIT compiler.
882    4.  Improved JIT compiler optimizations for character ranges.
884    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
885        Reported by Giuseppe D'Angelo.
887    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
888        of a line, was being set incorrectly in cases where .* appeared inside
889        atomic brackets at the start of a pattern, or where there was a subsequent
890        *PRUNE or *SKIP.
892    7.  Improved instruction cache flush for POWER/PowerPC.
893        Patch by Daniel Richard G.
895    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
896        grep:
898        (a) There is now no limit to the number of patterns to be matched.
900        (b) An error is given if a pattern is too long.
902        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
903            are now supported.
905        (d) --exclude-from and --include-from (multiple use) have been added.
907        (e) Exclusions and inclusions now apply to all files and directories, not
908            just to those obtained from scanning a directory recursively.
910        (f) Multiple uses of -f and --file-list are now supported.
912        (g) In a Windows environment, the default for -d has been changed from
913            "read" (the GNU grep default) to "skip", because otherwise the presence
914            of a directory in the file list provokes an error.
916        (h) The documentation has been revised and clarified in places.
918    9.  Improve the matching speed of capturing brackets.
920    10. Changed the meaning of \X so that it now matches a Unicode extended
921        grapheme cluster.
923    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
924        out POSIX threads when JIT support is configured.
926    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
928    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
929        structure could go wrong in environments where size_t is not the same size
930        as int.
932    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
933        set.
935    15. The EBCDIC support had decayed; later updates to the code had included
936        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
937        general tidy up of EBCDIC-related issues, and the documentation was also
938        not quite right. There is now a test that can be run on ASCII systems to
939        check some of the EBCDIC-related things (but is it not a full test).
941    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
942        in a small tidy to the code.
944    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
946    18. If the --only-matching (-o) option in pcregrep is specified multiple
947        times, each one causes appropriate output. For example, -o1 -o2 outputs the
948        substrings matched by the 1st and 2nd capturing parentheses. A separating
949        string can be specified by --om-separator (default empty).
951    19. Improving the first n character searches.
953    20. Turn case lists for horizontal and vertical white space into macros so that
954        they are defined only once.
956    21. This set of changes together give more compatible Unicode case-folding
957        behaviour for characters that have more than one other case when UCP
958        support is available.
960        (a) The Unicode property table now has offsets into a new table of sets of
961            three or more characters that are case-equivalent. The MultiStage2.py
962            script that generates these tables (the pcre_ucd.c file) now scans
963            CaseFolding.txt instead of UnicodeData.txt for character case
964            information.
966        (b) The code for adding characters or ranges of characters to a character
967            class has been abstracted into a generalized function that also handles
968            case-independence. In UTF-mode with UCP support, this uses the new data
969            to handle characters with more than one other case.
971        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
972            whose other case is greater than 256 are now correctly matched
973            caselessly. Previously, the high codepoint matched the low one, but not
974            vice versa.
976        (d) The processing of \h, \H, \v, and \ in character classes now makes use
977            of the new class addition function, using character lists defined as
978            macros alongside the case definitions of 20 above.
980        (e) Caseless back references now work with characters that have more than
981            one other case.
983        (f) General caseless matching of characters with more than one other case
984            is supported.
986    22. Unicode character properties were updated from Unicode 6.2.0
988    23. Improved CMake support under Windows. Patch by Daniel Richard G.
990    24. Add support for 32-bit character strings, and UTF-32
992    25. Major JIT compiler update (code refactoring and bugfixing).
993        Experimental Sparc 32 support is added.
995    26. Applied a modified version of Daniel Richard G's patch to create
996        pcre.h.generic and config.h.generic by "make" instead of in the
997        PrepareRelease script.
999    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
1000        pcre_compile.c when checking for a zero character.
1002    28. Introducing a native interface for JIT. Through this interface, the compiled
1003        machine code can be directly executed. The purpose of this interface is to
1004        provide fast pattern matching, so several sanity checks are not performed.
1005        However, feature tests are still performed. The new interface provides
1006        1.4x speedup compared to the old one.
1008    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
1009        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
1010        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
1012    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
1013        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
1014        the "old" RFC 2279). Instead, it ended up passing a negative length to
1015        pcre_exec().
1017    31. Add support for GCC's visibility feature to hide internal functions.
1019    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
1020        "unknown -C option" after outputting 0 or 1.
1022    33. There is now support for generating a code coverage report for the test
1023        suite in environments where gcc is the compiler and lcov is installed. This
1024        is mainly for the benefit of the developers.
1026    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
1027        unaddressable using valgrind annotations, allowing valgrind to detect
1028        invalid memory accesses. This is mainly for the benefit of the developers.
1030    25. (*UTF) can now be used to start a pattern in any of the three libraries.
1032    26. Give configure error if --enable-cpp but no C++ compiler found.
1035    Version 8.31 06-July-2012
1036    -------------------------
1038    1.  Fixing a wrong JIT test case and some compiler warnings.
1040    2.  Removed a bashism from the RunTest script.
1042    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
1043        to unsigned type, result still unsigned" that was given by an MS compiler
1044        on encountering the code "-sizeof(xxx)".
1046    4.  Partial matching support is added to the JIT compiler.
1048    5.  Fixed several bugs concerned with partial matching of items that consist
1049        of more than one character:
1051        (a) /^(..)\1/ did not partially match "aba" because checking references was
1052            done on an "all or nothing" basis. This also applied to repeated
1053            references.
1055        (b) \R did not give a hard partial match if \r was found at the end of the
1056            subject.
1058        (c) \X did not give a hard partial match after matching one or more
1059            characters at the end of the subject.
1061        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1062            a partial match for the string "\r".
1064        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1065            a partial match for a CR character at the end of the subject string.
1067    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1068        running pcretest, the text "(JIT)" added to the output whenever JIT is
1069        actually used to run the match.
1071    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1072        or /S+[+] with a digit between 1 and 7.
1074    8.  OP_NOT now supports any UTF character not just single-byte ones.
1076    9.  (*MARK) control verb is now supported by the JIT compiler.
1078    10. The command "./RunTest list" lists the available tests without actually
1079        running any of them. (Because I keep forgetting what they all are.)
1083    12. Applied a (slightly modified) user-supplied patch that improves performance
1084        when the heap is used for recursion (compiled with --disable-stack-for-
1085        recursion). Instead of malloc and free for each heap frame each time a
1086        logical recursion happens, frames are retained on a chain and re-used where
1087        possible. This sometimes gives as much as 30% improvement.
1089    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1090        call.
1092    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1094    15. It is now possible to link pcretest with libedit as an alternative to
1095        libreadline.
1097    16. (*COMMIT) control verb is now supported by the JIT compiler.
1099    17. The Unicode data tables have been updated to Unicode 6.1.0.
1101    18. Added --file-list option to pcregrep.
1103    19. Added binary file support to pcregrep, including the -a, --binary-files,
1104        -I, and --text options.
1106    20. The madvise function is renamed for posix_madvise for QNX compatibility
1107        reasons. Fixed by Giuseppe D'Angelo.
1109    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1111    22. Changed the option for creating symbolic links for 16-bit man pages from
1112        -s to -sf so that re-installing does not cause issues.
1114    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1116    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1117        matches in certain environments (the workspace was not being correctly
1118        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1119        some of the workspace data at the beginning of a restart.
1121    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1122        was not doing so when it should - probably a typo introduced by SVN 528
1123        (change 8.10/14).
1125    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1126        \w+ when the character tables indicated that \x{c4} was a word character.
1127        There were several related cases, all because the tests for doing a table
1128        lookup were testing for characters less than 127 instead of 255.
1130    27. If a pattern contains capturing parentheses that are not used in a match,
1131        their slots in the ovector are set to -1. For those that are higher than
1132        any matched groups, this happens at the end of processing. In the case when
1133        there were back references that the ovector was too small to contain
1134        (causing temporary malloc'd memory to be used during matching), and the
1135        highest capturing number was not used, memory off the end of the ovector
1136        was incorrectly being set to -1. (It was using the size of the temporary
1137        memory instead of the true size.)
1139    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1140        ovector size, it uses memory at the end of the block that it has got.
1142    29. Check for an overlong MARK name and give an error at compile time. The
1143        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1145    30. JIT compiler update.
1147    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1148        Rill for the patch.
1150    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1152    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1154    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1155        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1156        was enabled.
1158    35. Improve JIT code generation for greedy plus quantifier.
1160    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1161        1 to "aa" instead of to an empty string. The bug affected repeated groups
1162        that could potentially match an empty string.
1164    37. Optimizing single character iterators in JIT.
1166    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1167        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1168        codepoints that are too big for the mode are faulted, and in a UTF mode,
1169        disallowed codepoints are also faulted.
1171    39. If PCRE was compiled with UTF support, in three places in the DFA
1172        matcher there was code that should only have been obeyed in UTF mode, but
1173        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1174        processing when bytes with values greater than 127 were present. In 16-bit
1175        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1176        both cases the values are those that cannot be the first data item in a UTF
1177        character. The three items that might have provoked this were recursions,
1178        possessively repeated groups, and atomic groups.
1180    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1181        and pcregrep, because some OS require shared objects to be explicitly
1182        passed to ld, causing the link step to fail if they are not.
1184    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1185        mode, patterns that started with \h* or \R* might be incorrectly matched.
1188    Version 8.30 04-February-2012
1189    -----------------------------
1191    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1192        name is defined in ctype.h.
1194    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1195        only in quite long subpatterns.
1197    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1198        since it was replaced by pcre_fullinfo() in February 2000.
1200    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1201        match a (*MARK), and the match failed at the start of the subject, a
1202        reference to memory before the start of the subject could occur. This bug
1203        was introduced by fix 17 of release 8.21.
1205    5.  A reference to an unset group with zero minimum repetition was giving
1206        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1207        /(another)?(\1?)test/ matched against "hello world test". This bug was
1208        introduced in release 8.13.
1210    6.  Add support for 16-bit character strings (a large amount of work involving
1211        many changes and refactorings).
1213    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1214        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1215        from a file.
1217    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1218        rounding is not applied in this particular case).
1220    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1221        if they appear, or are escaped, in patterns.
1223    10. Get rid of a number of -Wunused-but-set-variable warnings.
1225    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1226        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1227        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1228        also returns the mark "x". This bug applied to capturing parentheses,
1229        non-capturing parentheses, and atomic parentheses. It also applied to some
1230        assertions.
1232    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1233        information out of configure.ac instead of relying on pcre.h.generic, which
1234        is not stored in the repository.
1236    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1237        -lreadline.
1239    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1241    15. Applied Graycode's patch to put the top-level frame on the stack rather
1242        than the heap when not using the stack for recursion. This gives a
1243        performance improvement in many cases when recursion is not deep.
1245    16. Experimental code added to "pcretest -C" to output the stack frame size.
1248    Version 8.21 12-Dec-2011
1249    ------------------------
1251    1.  Updating the JIT compiler.
1253    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1254        are added as well.
1256    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1257        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1258        calling _pcre_jit_exec. Some extra comments are added.
1260    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1261        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1262        was introduced by change 18 for 8.20.
1264    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1265        ECMA-262 standard.
1267    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1268        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1269        This bug was probably introduced by change 9 of 8.13.
1271    7.  While fixing 6 above, I noticed that a number of other items were being
1272        incorrectly rejected as "not fixed length". This arose partly because newer
1273        opcodes had not been added to the fixed-length checking code. I have (a)
1274        corrected the bug and added tests for these items, and (b) arranged for an
1275        error to occur if an unknown opcode is encountered while checking for fixed
1276        length instead of just assuming "not fixed length". The items that were
1277        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1278        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1279        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1281    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1282        being incorrectly compiled and would have given unpredicatble results.
1284    9.  A possessively repeated subpattern with minimum repeat count greater than
1285        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1286        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1287        the first (A) could occur when it should not.
1289    10. Add a cast and remove a redundant test from the code.
1291    11. JIT should use pcre_malloc/pcre_free for allocation.
1293    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1294        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1295        is anything other than /usr, -L is still shown).
1297    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1299    14. Perl does not support \N without a following name in a [] class; PCRE now
1300        also gives an error.
1302    15. If a forward reference was repeated with an upper limit of around 2000,
1303        it caused the error "internal error: overran compiling workspace". The
1304        maximum number of forward references (including repeats) was limited by the
1305        internal workspace, and dependent on the LINK_SIZE. The code has been
1306        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1307        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1308        of around 200,000 forward references. While doing this, I also speeded up
1309        the filling in of repeated forward references.
1311    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1312        incorrectly expecting the subject to contain another "a" after the start.
1314    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1315        in the match, the SKIP should be ignored. This was not happening; instead
1316        the SKIP was being treated as NOMATCH. For patterns such as
1317        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1318        tested.
1320    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1321        now much more compatible with Perl, in particular in cases where the result
1322        is a non-match for a non-anchored pattern. For example, if
1323        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1324        "m", where previously it did not return a name. A side effect of this
1325        change is that for partial matches, the last encountered mark name is
1326        returned, as for non matches. A number of tests that were previously not
1327        Perl-compatible have been moved into the Perl-compatible test files. The
1328        refactoring has had the pleasing side effect of removing one argument from
1329        the match() function, thus reducing its stack requirements.
1331    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1332        subsequent uses of /S (without +) incorrectly behaved like /S+.
1334    21. Retrieve executable code size support for the JIT compiler and fixing
1335        some warnings.
1337    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1338        not work when the shorter character appeared right at the end of the
1339        subject string.
1341    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1342        systems.
1344    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1345        output it when the /M option is used in pcretest.
1347    25. The CheckMan script was not being included in the distribution. Also, added
1348        an explicit "perl" to run Perl scripts from the PrepareRelease script
1349        because this is reportedly needed in Windows.
1351    26. If study data was being save in a file and studying had not found a set of
1352        "starts with" bytes for the pattern, the data written to the file (though
1353        never used) was taken from uninitialized memory and so caused valgrind to
1354        complain.
1356    27. Updated RunTest.bat as provided by Sheri Pierce.
1358    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1360    29. Computation of memory usage for the table of capturing group names was
1361        giving an unnecessarily large value.
1364    Version 8.20 21-Oct-2011
1365    ------------------------
1367    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1368        a POSIX class. After further experiments with Perl, which convinced me that
1369        Perl has bugs and confusions, a closing square bracket is no longer allowed
1370        in a POSIX name. This bug also affected patterns with classes that started
1371        with full stops.
1373    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1374        captured substring, but while checking the failing first alternative,
1375        substring 1 is temporarily captured. If the output vector supplied to
1376        pcre_exec() was not big enough for this capture, the yield of the function
1377        was still zero ("insufficient space for captured substrings"). This cannot
1378        be totally fixed without adding another stack variable, which seems a lot
1379        of expense for a edge case. However, I have improved the situation in cases
1380        such as /(a)(b)x|abc/ matched against "abc", where the return code
1381        indicates that fewer than the maximum number of slots in the ovector have
1382        been set.
1384    3.  Related to (2) above: when there are more back references in a pattern than
1385        slots in the output vector, pcre_exec() uses temporary memory during
1386        matching, and copies in the captures as far as possible afterwards. It was
1387        using the entire output vector, but this conflicts with the specification
1388        that only 2/3 is used for passing back captured substrings. Now it uses
1389        only the first 2/3, for compatibility. This is, of course, another edge
1390        case.
1392    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1393        main code base, and can be used by building with --enable-jit. When this is
1394        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1395        runtime --no-jit option is given.
1397    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1398        ovector, the return from the function was zero, implying that there were
1399        other matches that did not fit. The correct "exactly full" value is now
1400        returned.
1402    6.  If a subpattern that was called recursively or as a subroutine contained
1403        (*PRUNE) or any other control that caused it to give a non-standard return,
1404        invalid errors such as "Error -26 (nested recursion at the same subject
1405        position)" or even infinite loops could occur.
1407    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1408        computing the minimum length on reaching *ACCEPT, and so ended up with the
1409        wrong value of 1 rather than 0. Further investigation indicates that
1410        computing a minimum subject length in the presence of *ACCEPT is difficult
1411        (think back references, subroutine calls), and so I have changed the code
1412        so that no minimum is registered for a pattern that contains *ACCEPT.
1414    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1415        it was not handled as intended. [But see 16 below.]
1417    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1418        Sheri Pierce.
1420    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1421        the first byte in a match must be "a".
1423    11. Change 17 for 8.13 increased the recursion depth for patterns like
1424        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1425        pattern contains any instances of (*THEN). If it does not, the old
1426        optimizations are restored. It would be nice to do this on a per-group
1427        basis, but at the moment that is not feasible.
1429    12. In some environments, the output of pcretest -C is CRLF terminated. This
1430        broke RunTest's code that checks for the link size. A single white space
1431        character after the value is now allowed for.
1433    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1434        For "fr", it uses the Windows-specific input and output files.
1436    14. If (*THEN) appeared in a group that was called recursively or as a
1437        subroutine, it did not work as intended. [But see next item.]
1439    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1440        pattern fragments (but not containing any | characters). If A and B are
1441        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1442        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1443        D. In other words, Perl considers parentheses that do not contain any |
1444        characters to be part of a surrounding alternative, whereas PCRE was
1445        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1446        differently. PCRE now behaves in the same way as Perl, except in the case
1447        of subroutine/recursion calls such as (?1) which have in any case always
1448        been different (but PCRE had them first :-).
1450    16. Related to 15 above: Perl does not treat the | in a conditional group as
1451        creating alternatives. Such a group is treated in the same way as an
1452        ordinary group without any | characters when processing (*THEN). PCRE has
1453        been changed to match Perl's behaviour.
1455    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1456        RunGrepTest script failed.
1458    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1459        inevitable for groups that contain captures, but it can lead to a lot of
1460        stack use in large patterns. The old behaviour has been restored for atomic
1461        groups that do not contain any capturing parentheses.
1463    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1464        suppress the check for a minimum subject length at run time. (If it was
1465        given to pcre_exec() or pcre_dfa_exec() it did work.)
1467    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1468        fail to work when decoding hex characters in data strings in EBCDIC
1469        environments.
1471    21. It appears that in at least one Mac OS environment, the isxdigit() function
1472        is implemented as a macro that evaluates to its argument more than once,
1473        contravening the C 90 Standard (I haven't checked a later standard). There
1474        was an instance in pcretest which caused it to go wrong when processing
1475        \x{...} escapes in subject strings. The has been rewritten to avoid using
1476        things like p++ in the argument of isxdigit().
1479    Version 8.13 16-Aug-2011
1480    ------------------------
1482    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1484    2.  Two minor typos in pcre_internal.h have been fixed.
1486    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1487        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1488        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1490    4.  There were a number of related bugs in the code for matching backrefences
1491        caselessly in UTF-8 mode when codes for the characters concerned were
1492        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1493        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1494        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1495        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1496        2-byte code at the end of the subject (it thought there wasn't enough data
1497        left).
1499    5.  Comprehensive information about what went wrong is now returned by
1500        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1501        as the output vector has at least 2 elements. The offset of the start of
1502        the failing character and a reason code are placed in the vector.
1504    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1505        now returned is for the first byte of the failing character, instead of the
1506        last byte inspected. This is an incompatible change, but I hope it is small
1507        enough not to be a problem. It makes the returned offset consistent with
1508        pcre_exec() and pcre_dfa_exec().
1510    7.  pcretest now gives a text phrase as well as the error number when
1511        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1512        failure, the offset and reason code are output.
1514    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1515        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1516        back over a single character (\n). This seems wrong (because it treated the
1517        two characters as a single entity when going forwards), conflicts with the
1518        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1519        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1520        has been changed.
1522    9.  Some internal refactoring has changed the processing so that the handling
1523        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1524        time (the PCRE_DOTALL option was changed this way some time ago: version
1525        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1526        which was always a bit of a fudge. It also means that there is one less
1527        argument for the match() function, which reduces its stack requirements
1528        slightly. This change also fixes an incompatibility with Perl: the pattern
1529        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1531    10. More internal refactoring has drastically reduced the number of recursive
1532        calls to match() for possessively repeated groups such as (abc)++ when
1533        using pcre_exec().
1535    11. While implementing 10, a number of bugs in the handling of groups were
1536        discovered and fixed:
1538        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1539        (a|)*(?1) gave a compile-time internal error.
1540        ((a|)+)+  did not notice that the outer group could match an empty string.
1541        (^a|^)+   was not marked as anchored.
1542        (.*a|.*)+ was not marked as matching at start or after a newline.
1544    12. Yet more internal refactoring has removed another argument from the match()
1545        function. Special calls to this function are now indicated by setting a
1546        value in a variable in the "match data" data block.
1548    13. Be more explicit in pcre_study() instead of relying on "default" for
1549        opcodes that mean there is no starting character; this means that when new
1550        ones are added and accidentally left out of pcre_study(), testing should
1551        pick them up.
1553    14. The -s option of pcretest has been documented for ages as being an old
1554        synonym of -m (show memory usage). I have changed it to mean "force study
1555        for every regex", that is, assume /S for every regex. This is similar to -i
1556        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1557        using it. It makes it easier to run collections of tests with and without
1558        study enabled, and thereby test pcre_study() more easily. All the standard
1559        tests are now run with and without -s (but some patterns can be marked as
1560        "never study" - see 20 below).
1562    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1563        restoration of the capturing data to the outer values was not happening
1564        correctly.
1566    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1567        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1568        pattern had matched an empty string, and so incorrectly returned a no
1569        match.
1571    17. There was optimizing code for the last branch of non-capturing parentheses,
1572        and also for the obeyed branch of a conditional subexpression, which used
1573        tail recursion to cut down on stack usage. Unfortunately, now that there is
1574        the possibility of (*THEN) occurring in these branches, tail recursion is
1575        no longer possible because the return has to be checked for (*THEN). These
1576        two optimizations have therefore been removed. [But see 8.20/11 above.]
1578    18. If a pattern containing \R was studied, it was assumed that \R always
1579        matched two bytes, thus causing the minimum subject length to be
1580        incorrectly computed because \R can also match just one byte.
1582    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1583        was incorrectly computed.
1585    20. If /S is present twice on a test pattern in pcretest input, it now
1586        *disables* studying, thereby overriding the use of -s on the command line
1587        (see 14 above). This is necessary for one or two tests to keep the output
1588        identical in both cases.
1590    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1591        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1593    22. When an atomic group that contained a capturing parenthesis was
1594        successfully matched, but the branch in which it appeared failed, the
1595        capturing was not being forgotten if a higher numbered group was later
1596        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1597        group 1 to "a", when in fact it should be unset. This applied to multi-
1598        branched capturing and non-capturing groups, repeated or not, and also to
1599        positive assertions (capturing in negative assertions does not happen
1600        in PCRE) and also to nested atomic groups.
1602    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1603        subject after a captured substring, to make it easier to tell which of a
1604        number of identical substrings has been captured.
1606    24. The way atomic groups are processed by pcre_exec() has been changed so that
1607        if they are repeated, backtracking one repetition now resets captured
1608        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1609        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1610        "aaa". Previously, it would have been "a". As part of this code
1611        refactoring, the way recursive calls are handled has also been changed.
1613    25. If an assertion condition captured any substrings, they were not passed
1614        back unless some other capturing happened later. For example, if
1615        (?(?=(a))a) was matched against "a", no capturing was returned.
1617    26. When studying a pattern that contained subroutine calls or assertions,
1618        the code for finding the minimum length of a possible match was handling
1619        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1620        group 1 called group 2 while simultaneously a separate group 2 called group
1621        1). A stack overflow occurred in this case. I have fixed this by limiting
1622        the recursion depth to 10.
1624    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1625        Fortmann. This supports explicit test numbers on the command line, and has
1626        argument validation and error reporting.
1628    28. An instance of \X with an unlimited repeat could fail if at any point the
1629        first character it looked at was a mark character.
1631    29. Some minor code refactoring concerning Unicode properties and scripts
1632        should reduce the stack requirement of match() slightly.
1634    30. Added the '=' option to pcretest to check the setting of unused capturing
1635        slots at the end of the pattern, which are documented as being -1, but are
1636        not included in the return count.
1638    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1639        compiled something random. Now it gives a compile-time error (as does
1640        Perl).
1642    32. A *MARK encountered during the processing of a positive assertion is now
1643        recorded and passed back (compatible with Perl).
1645    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1646        had alternative anchored branches, the search for a second match in a line
1647        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1648        matched the line "0102" twice. The same bug affected patterns that started
1649        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1650        twice.
1652    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1653        does, and because of capturing effects, quantifying parenthesized
1654        assertions may at times be useful. Quantifiers are now allowed for
1655        parenthesized assertions.
1657    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1659    36. \g was being checked for fancy things in a character class, when it should
1660        just be a literal "g".
1662    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1663        appearance of a nested POSIX class supersedes an apparent external class.
1664        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1665        unescaped square brackets may also appear as part of class names. For
1666        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1667        more like Perl. (But see 8.20/1 above.)
1669    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1670        was because it thought it was \N{name}, which is not supported).
1672    39. Add minix to OS list not supporting the -S option in pcretest.
1674    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1675        cannot analyze patterns in sufficient detail to catch mutual recursions
1676        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1677        subgroup is called recursively as a subpattern for a second time at the
1678        same position in the subject string. In previous releases this might have
1679        been caught by the recursion limit, or it might have run out of stack.
1681    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1682        happen only once. PCRE was, however incorrectly giving a compile time error
1683        "recursive call could loop indefinitely" because it cannot analyze the
1684        pattern in sufficient detail. The compile time test no longer happens when
1685        PCRE is compiling a conditional subpattern, but actual runaway loops are
1686        now caught at runtime (see 40 above).
1688    42. It seems that Perl allows any characters other than a closing parenthesis
1689        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1690        has been changed to be the same.
1692    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1693        as not to get warnings when autogen.sh is called. Also changed
1694        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1696    44. To help people who use pcregrep to scan files containing exceedingly long
1697        lines, the following changes have been made:
1699        (a) The default value of the buffer size parameter has been increased from
1700            8K to 20K. (The actual buffer used is three times this size.)
1702        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1703            PCRE is built.
1705        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1706            to be set at run time.
1708        (d) Numerical values in pcregrep options can be followed by K or M, for
1709            example --buffer-size=50K.
1711        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1712            given and the return code is set to 2.
1714    45. Add a pointer to the latest mark to the callout data block.
1716    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1717        partial match of an empty string instead of no match. This was specific to
1718        the use of ".".
1720    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1721        complete match instead of a partial match. This bug was dependent on both
1722        the PCRE_UTF8 and PCRE_DOTALL options being set.
1724    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1725        starting byte set, because \b was not being ignored.
1728    Version 8.12 15-Jan-2011
1729    ------------------------
1731    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1732        checks for such things as part of the documentation building process.
1734    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1735        --match-limit and --recursion-limit options (added for 8.11). In
1736        particular, this made one of the standard tests fail. (The integer value
1737        went into the wrong half of a long int.)
1739    3.  If the --colour option was given to pcregrep with -v (invert match), it
1740        did strange things, either producing crazy output, or crashing. It should,
1741        of course, ignore a request for colour when reporting lines that do not
1742        match.
1744    4.  Another pcregrep bug caused similar problems if --colour was specified with
1745        -M (multiline) and the pattern match finished with a line ending.
1747    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1748        matched in multiline mode, the following line was shown as part of the
1749        match. This seems wrong, so I have changed it.
1751    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1752        the check for further matches in the same line (so they could be coloured)
1753        to overrun the end of the current line. If another match was found, it was
1754        incorrectly shown (and then shown again when found in the next line).
1756    7.  If pcregrep was compiled under Windows, there was a reference to the
1757        function pcregrep_exit() before it was defined. I am assuming this was
1758        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1759        reported by a user. I've moved the definition above the reference.
1762    Version 8.11 10-Dec-2010
1763    ------------------------
1765    1.  (*THEN) was not working properly if there were untried alternatives prior
1766        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1767        backtracked to try for "b" instead of moving to the next alternative branch
1768        at the same level (in this case, to look for "c"). The Perl documentation
1769        is clear that when (*THEN) is backtracked onto, it goes to the "next
1770        alternative in the innermost enclosing group".
1772    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1773        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1774        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1775        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1776        (*THEN).
1778    3.  If \s appeared in a character class, it removed the VT character from
1779        the class, even if it had been included by some previous item, for example
1780        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1781        of \s, but is part of the POSIX "space" class.)
1783    4.  A partial match never returns an empty string (because you can always
1784        match an empty string at the end of the subject); however the checking for
1785        an empty string was starting at the "start of match" point. This has been
1786        changed to the "earliest inspected character" point, because the returned
1787        data for a partial match starts at this character. This means that, for
1788        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1789        (previously it gave "no match").
1791    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1792        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1793        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1794        has an implication that the given string is incomplete (because a partial
1795        match is preferred over a full match). For this reason, these items now
1796        give a partial match in this situation. [Aside: previously, the one case
1797        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1798        match rather than a full match, which was wrong by the old rules, but is
1799        now correct.]
1801    6.  There was a bug in the handling of #-introduced comments, recognized when
1802        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1803        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1804        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1805        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1806        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1807        places in pcre_compile().
1809    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1810        comments when looking ahead for named forward references to subpatterns,
1811        the only newline sequence it recognized was NL. It now handles newlines
1812        according to the set newline convention.
1814    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1815        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1816        cater for a lack of strerror(). These oversights have been fixed.
1818    9.  Added --match-limit and --recursion-limit to pcregrep.
1820    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1822    11. When the -o option was used, pcregrep was setting a return code of 1, even
1823        when matches were found, and --line-buffered was not being honoured.
1825    12. Added an optional parentheses number to the -o and --only-matching options
1826        of pcregrep.
1828    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1829        can match an empty string. The code to do it in pcretest and pcredemo
1830        needed fixing:
1832        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1833            only one byte after an empty string match just before CRLF (this case
1834            just got forgotten; "any" and "anycrlf" were OK).
1836        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1837            mode when an empty string match preceded an ASCII character followed by
1838            a non-ASCII character. (The code for advancing by one character rather
1839            than one byte was nonsense.)
1841        (c) The pcredemo.c sample program did not have any code at all to handle
1842            the cases when CRLF is a valid newline sequence.
1844    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1845        as a starting offset was within the subject string. There is now a new
1846        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1847        negative or greater than the length of the string. In order to test this,
1848        pcretest is extended to allow the setting of negative starting offsets.
1850    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1851        starting offset points to the beginning of a UTF-8 character was
1852        unnecessarily clumsy. I tidied it up.
1854    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1855        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1857    17. Nobody had reported that the --include_dir option, which was added in
1858        release 7.7 should have been called --include-dir (hyphen, not underscore)
1859        for compatibility with GNU grep. I have changed it to --include-dir, but
1860        left --include_dir as an undocumented synonym, and the same for
1861        --exclude-dir, though that is not available in GNU grep, at least as of
1862        release 2.5.4.
1864    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1865        characters from a string of bytes) have been redefined so as not to use
1866        loops, in order to improve performance in some environments. At the same
1867        time, I abstracted some of the common code into auxiliary macros to save
1868        repetition (this should not affect the compiled code).
1870    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1871        compile-time error is now given if \c is not followed by an ASCII
1872        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1873        different, and any byte value is allowed.)
1875    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1876        START_OPTIMIZE option, which is now allowed at compile time - but just
1877        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1878        to pcregrep and other applications that have no direct access to PCRE
1879        options. The new /Y option in pcretest sets this option when calling
1880        pcre_compile().
1882    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1883        back references. Groups containing recursive back references were forced to
1884        be atomic by that change, but in the case of named groups, the amount of
1885        memory required was incorrectly computed, leading to "Failed: internal
1886        error: code overflow". This has been fixed.
1888    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1889        pcretest.c, to avoid build problems in some Borland environments.
1892    Version 8.10 25-Jun-2010
1893    ------------------------
1895    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1896        THEN.
1898    2.  (*ACCEPT) was not working when inside an atomic group.
1900    3.  Inside a character class, \B is treated as a literal by default, but
1901        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1902        causes the error). The code is unchanged, but I tidied the documentation.
1904    4.  Inside a character class, PCRE always treated \R and \X as literals,
1905        whereas Perl faults them if its -w option is set. I have changed PCRE so
1906        that it faults them when PCRE_EXTRA is set.
1908    5.  Added support for \N, which always matches any character other than
1909        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1911    6.  When compiling pcregrep with newer versions of gcc which may have
1912        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1913        declared with attribute warn_unused_result" were given. Just casting the
1914        result to (void) does not stop the warnings; a more elaborate fudge is
1915        needed. I've used a macro to implement this.
1917    7.  Minor change to pcretest.c to avoid a compiler warning.
1919    8.  Added four artifical Unicode properties to help with an option to make
1920        \s etc use properties (see next item). The new properties are: Xan
1921        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1923    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1924        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1925        this option. Modified pcretest to add /W to test this facility. Added
1926        REG_UCP to make it available via the POSIX interface.
1928    10. Added --line-buffered to pcregrep.
1930    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1931        studied, and the match started with a letter with a code point greater than
1932        127 whose first byte was different to the first byte of the other case of
1933        the letter, the other case of this starting letter was not recognized
1934        (#976).
1936    12. If a pattern that was studied started with a repeated Unicode property
1937        test, for example, \p{Nd}+, there was the theoretical possibility of
1938        setting up an incorrect bitmap of starting bytes, but fortunately it could
1939        not have actually happened in practice until change 8 above was made (it
1940        added property types that matched character-matching opcodes).
1942    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1943        possible starting bytes for non-anchored patterns.
1945    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1946        \R, and also a number of cases that involve Unicode properties, both
1947        explicit and implicit when PCRE_UCP is set.
1949    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1950        input, it could crash or give wrong results if characters with values
1951        greater than 0xc0 were present in the subject string. (Detail: it assumed
1952        UTF-8 input when processing these items.)
1954    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1955        size_t is 64-bit (#991).
1957    17. Added a check for running out of memory when PCRE is compiled with
1958        --disable-stack-for-recursion (#990).
1960    18. If the last data line in a file for pcretest does not have a newline on
1961        the end, a newline was missing in the output.
1963    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1964        less than 128) in its various bitmaps. However, there is a facility for
1965        generating tables according to the current locale when PCRE is compiled. It
1966        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1967        characters, are recognized by isspace() and therefore were getting set in
1968        these tables, and indeed these tables seem to approximate to ISO 8859. This
1969        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1970        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1971        which of course cannot start UTF-8 characters. I have changed the code so
1972        that only real ASCII characters (less than 128) and the correct starting
1973        bytes for UTF-8 encodings are set for characters greater than 127 when in
1974        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1975        altogether.)
1977    20. Added the /T option to pcretest so as to be able to run tests with non-
1978        standard character tables, thus making it possible to include the tests
1979        used for 19 above in the standard set of tests.
1981    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1982        reference to a subpattern the other side of a comment that contains an
1983        opening parenthesis caused either an internal compiling error, or a
1984        reference to the wrong subpattern.
1987    Version 8.02 19-Mar-2010
1988    ------------------------
1990    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1992    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1993        configured.
1995    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1996        original author of that file, following a query about its status.
1998    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1999        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
2001    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
2002        quantifier applied to a forward-referencing subroutine call, could compile
2003        incorrect code or give the error "internal error: previously-checked
2004        referenced subpattern not found".
2006    6.  Both MS Visual Studio and Symbian OS have problems with initializing
2007        variables to point to external functions. For these systems, therefore,
2008        pcre_malloc etc. are now initialized to local functions that call the
2009        relevant global functions.
2011    7.  There were two entries missing in the vectors called coptable and poptable
2012        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
2013        I've fixed the data, and added a kludgy way of testing at compile time that
2014        the lengths are correct (equal to the number of opcodes).
2016    8.  Following on from 7, I added a similar kludge to check the length of the
2017        eint vector in pcreposix.c.
2019    9.  Error texts for pcre_compile() are held as one long string to avoid too
2020        much relocation at load time. To find a text, the string is searched,
2021        counting zeros. There was no check for running off the end of the string,
2022        which could happen if a new error number was added without updating the
2023        string.
2025    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
2027    11. \K was not working if it appeared in an atomic group or in a group that
2028        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
2029        \K is "not well defined" if used in an assertion. PCRE now accepts it if
2030        the assertion is positive, but not if it is negative.
2032    12. Change 11 fortuitously reduced the size of the stack frame used in the
2033        "match()" function of pcre_exec.c by one pointer. Forthcoming
2034        implementation of support for (*MARK) will need an extra pointer on the
2035        stack; I have reserved it now, so that the stack frame size does not
2036        decrease.
2038    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
2039        item in branch that calls a recursion is a subroutine call - as in the
2040        second branch in the above example - was incorrectly given the compile-
2041        time error "recursive call could loop indefinitely" because pcre_compile()
2042        was not correctly checking the subroutine for matching a non-empty string.
2044    14. The checks for overrunning compiling workspace could trigger after an
2045        overrun had occurred. This is a "should never occur" error, but it can be
2046        triggered by pathological patterns such as hundreds of nested parentheses.
2047        The checks now trigger 100 bytes before the end of the workspace.
2049    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
2052    Version 8.01 19-Jan-2010
2053    ------------------------
2055  1.  If a pattern contained a conditional subpattern with only one branch (in  1.  If a pattern contained a conditional subpattern with only one branch (in
2056      particular, this includes all (DEFINE) patterns), a call to pcre_study()      particular, this includes all (*DEFINE) patterns), a call to pcre_study()
2057      computed the wrong minimum data length (which is of course zero for such      computed the wrong minimum data length (which is of course zero for such
2058      subpatterns).      subpatterns). This could cause incorrect "no match" results.
2060  2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of  2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
2061      the pattern is reset in the first branch, pcre_compile() failed with      the pattern is reset in the first branch, pcre_compile() failed with
# Line 61  Version 8.01 11-Dec-09 Line 2109  Version 8.01 11-Dec-09
2109      custom one, because of the following reported problem in Windows:      custom one, because of the following reported problem in Windows:
2111        - libbz2 uses the Pascal calling convention (WINAPI) for the functions        - libbz2 uses the Pascal calling convention (WINAPI) for the functions
2112          under Win32.            under Win32.
2113        - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",        - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
2114          therefore missing the function definition.            therefore missing the function definition.
2115        - The compiler thus generates a "C" signature for the test function.        - The compiler thus generates a "C" signature for the test function.
2116        - The linker fails to find the "C" function.        - The linker fails to find the "C" function.
2117        - PCRE fails to configure if asked to do so against libbz2.        - PCRE fails to configure if asked to do so against libbz2.
2119  11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these  11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
2120      messages were output:      messages were output:
2122        Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and        Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
2123        rerunning libtoolize, to keep the correct libtool macros in-tree.        rerunning libtoolize, to keep the correct libtool macros in-tree.
2124        Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.        Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
2126      I have done both of these things.      I have done both of these things.
2128  12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()  12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
2129      most of the time, it *can* run out if it is given a pattern that contains a      most of the time, it *can* run out if it is given a pattern that contains a
2130      runaway infinite recursion. I updated the discussion in the pcrestack man      runaway infinite recursion. I updated the discussion in the pcrestack man
2131      page.      page.
2133  13. Now that we have gone to the x.xx style of version numbers, the minor  13. Now that we have gone to the x.xx style of version numbers, the minor
2134      version may start with zero. Using 08 or 09 is a bad idea because users      version may start with zero. Using 08 or 09 is a bad idea because users
2135      might check the value of PCRE_MINOR in their code, and 08 or 09 may be      might check the value of PCRE_MINOR in their code, and 08 or 09 may be
2136      interpreted as invalid octal numbers. I've updated the previous comment in      interpreted as invalid octal numbers. I've updated the previous comment in
2137      configure.ac, and also added a check that gives an error if 08 or 09 are      configure.ac, and also added a check that gives an error if 08 or 09 are
2138      used.      used.
2140  14. Change 8.00/11 was not quite complete: code had been accidentally omitted,  14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
2141      causing partial matching to fail when the end of the subject matched \W      causing partial matching to fail when the end of the subject matched \W
2142      in a UTF-8 pattern where \W was quantified with a minimum of 3.      in a UTF-8 pattern where \W was quantified with a minimum of 3.
2144  15. There were some discrepancies between the declarations in pcre_internal.h  15. There were some discrepancies between the declarations in pcre_internal.h
2145      of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and      of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
2146      their definitions. The declarations used "const uschar *" and the      their definitions. The declarations used "const uschar *" and the
2147      definitions used USPTR. Even though USPTR is normally defined as "const      definitions used USPTR. Even though USPTR is normally defined as "const
2148      unsigned char *" (and uschar is typedeffed as "unsigned char"), it was      unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
2149      reported that: "This difference in casting confuses some C++ compilers, for      reported that: "This difference in casting confuses some C++ compilers, for
2150      example, SunCC recognizes above declarations as different functions and      example, SunCC recognizes above declarations as different functions and
2151      generates broken code for hbpcre." I have changed the declarations to use      generates broken code for hbpcre." I have changed the declarations to use
2152      USPTR.      USPTR.
2154    16. GNU libtool is named differently on some systems. The autogen.sh script now
2155        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
2156        (FreeBSD).
2158    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
2159        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
2160        comment: "Figure out how to create a longlong from a string: strtoll and
2161        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
2162        instance, but it only takes 2 args instead of 3!"
2164    18. A subtle bug concerned with back references has been fixed by a change of
2165        specification, with a corresponding code fix. A pattern such as
2166        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
2167        refers, was giving matches when it shouldn't. For example, xa=xaaa would
2168        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
2169        same bug. Such groups have to be quantified to be useful, or contained
2170        inside another quantified group. (If there's no repetition, the reference
2171        can never match.) The problem arises because, having left the group and
2172        moved on to the rest of the pattern, a later failure that backtracks into
2173        the group uses the captured value from the final iteration of the group
2174        rather than the correct earlier one. I have fixed this in PCRE by forcing
2175        any group that contains a reference to itself to be an atomic group; that
2176        is, there cannot be any backtracking into it once it has completed. This is
2177        similar to recursive and subroutine calls.
2180  Version 8.00 19-Oct-09  Version 8.00 19-Oct-09
# Line 304  Version 7.9 11-Apr-09 Line 2375  Version 7.9 11-Apr-09
2375  7.  A pattern that could match an empty string could cause pcregrep to loop; it  7.  A pattern that could match an empty string could cause pcregrep to loop; it
2376      doesn't make sense to accept an empty string match in pcregrep, so I have      doesn't make sense to accept an empty string match in pcregrep, so I have
2377      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2378      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2379        8.33.]
2381  8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at  8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2382      start or after a newline", because the conditional assertion was not being      start or after a newline", because the conditional assertion was not being
# Line 547  Version 7.7 07-May-08 Line 2619  Version 7.7 07-May-08
2619      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2620      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2621      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2622      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2623      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2624      was subsequently set up correctly.)      was subsequently set up correctly.)

Removed from v.483  
changed lines
  Added in v.1585

  ViewVC Help
Powered by ViewVC 1.1.5