ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 657 by ph10, Mon Aug 15 17:39:09 2011 UTC revision 1689 by ph10, Wed Mar 22 15:17:45 2017 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4  Version 8.13 02-Aug-2011  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 8.41
8    ------------
10    1.  Fixed typo in CMakeLists.txt (wrong number of arguments for
11    PCRE_STATIC_RUNTIME (affects MSVC only).
13    2.  Issue 1 for 8.40 below was not correctly fixed. If pcregrep in multiline
14    mode with --only-matching matched several lines, it restarted scanning at the
15    next line instead of moving on to the end of the matched string, which can be
16    several lines after the start.
18    3.  Fix a missing else in the JIT compiler reported by 'idaifish'.
20    4.  A (?# style comment is now ignored between a basic quantifier and a
21    following '+' or '?' (example: /X+(?#comment)?Y/.
23    5.  Avoid use of a potentially overflowing buffer in pcregrep (patch by Petr
24    Pisar).
26    6.  Fuzzers have reported issues in pcretest. These are NOT serious (it is,
27    after all, just a test program). However, to stop the reports, some easy ones
28    are fixed:
30        (a) Check for values < 256 when calling isprint() in pcretest.
31        (b) Give an error for too big a number after \O.
33    7.  In the 32-bit library in non-UTF mode, an attempt to find a Unicode
34    property for a character with a code point greater than 0x10ffff (the Unicode
35    maximum) caused a crash.
37    8. The alternative matching function, pcre_dfa_exec() misbehaved if it
38    encountered a character class with a possessive repeat, for example [a-f]{3}+.
41    Version 8.40 11-January-2017
42    ----------------------------
44    1.  Using -o with -M in pcregrep could cause unnecessary repeated output when
45        the match extended over a line boundary.
47    2.  Applied Chris Wilson's second patch (Bugzilla #1681) to CMakeLists.txt for
48        MSVC static compilation, putting the first patch under a new option.
50    3.  Fix register overwite in JIT when SSE2 acceleration is enabled.
52    4.  Ignore "show all captures" (/=) for DFA matching.
54    5.  Fix JIT unaligned accesses on x86. Patch by Marc Mutz.
56    6.  In any wide-character mode (8-bit UTF or any 16-bit or 32-bit mode),
57        without PCRE_UCP set, a negative character type such as \D in a positive
58        class should cause all characters greater than 255 to match, whatever else
59        is in the class. There was a bug that caused this not to happen if a
60        Unicode property item was added to such a class, for example [\D\P{Nd}] or
61        [\W\pL].
63    7.  When pcretest was outputing information from a callout, the caret indicator
64        for the current position in the subject line was incorrect if it was after
65        an escape sequence for a character whose code point was greater than
66        \x{ff}.
68    8.  A pattern such as (?<RA>abc)(?(R)xyz) was incorrectly compiled such that
69        the conditional was interpreted as a reference to capturing group 1 instead
70        of a test for recursion. Any group whose name began with R was
71        misinterpreted in this way. (The reference interpretation should only
72        happen if the group's name is precisely "R".)
74    9.  A number of bugs have been mended relating to match start-up optimizations
75        when the first thing in a pattern is a positive lookahead. These all
76        applied only when PCRE_NO_START_OPTIMIZE was *not* set:
78        (a) A pattern such as (?=.*X)X$ was incorrectly optimized as if it needed
79            both an initial 'X' and a following 'X'.
80        (b) Some patterns starting with an assertion that started with .* were
81            incorrectly optimized as having to match at the start of the subject or
82            after a newline. There are cases where this is not true, for example,
83            (?=.*[A-Z])(?=.{8,16})(?!.*[\s]) matches after the start in lines that
84            start with spaces. Starting .* in an assertion is no longer taken as an
85            indication of matching at the start (or after a newline).
88    Version 8.39 14-June-2016
89    -------------------------
91    1.  If PCRE_AUTO_CALLOUT was set on a pattern that had a (?# comment between
92        an item and its qualifier (for example, A(?#comment)?B) pcre_compile()
93        misbehaved. This bug was found by the LLVM fuzzer.
95    2.  Similar to the above, if an isolated \E was present between an item and its
96        qualifier when PCRE_AUTO_CALLOUT was set, pcre_compile() misbehaved. This
97        bug was found by the LLVM fuzzer.
99    3.  Further to 8.38/46, negated classes such as [^[:^ascii:]\d] were also not
100        working correctly in UCP mode.
102    4.  The POSIX wrapper function regexec() crashed if the option REG_STARTEND
103        was set when the pmatch argument was NULL. It now returns REG_INVARG.
105    5.  Allow for up to 32-bit numbers in the ordin() function in pcregrep.
107    6.  An empty \Q\E sequence between an item and its qualifier caused
108        pcre_compile() to misbehave when auto callouts were enabled. This bug was
109        found by the LLVM fuzzer.
111    7.  If a pattern that was compiled with PCRE_EXTENDED started with white
112        space or a #-type comment that was followed by (?-x), which turns off
113        PCRE_EXTENDED, and there was no subsequent (?x) to turn it on again,
114        pcre_compile() assumed that (?-x) applied to the whole pattern and
115        consequently mis-compiled it. This bug was found by the LLVM fuzzer.
117    8.  A call of pcre_copy_named_substring() for a named substring whose number
118        was greater than the space in the ovector could cause a crash.
120    9.  Yet another buffer overflow bug involved duplicate named groups with a
121        group that reset capture numbers (compare 8.38/7 below). Once again, I have
122        just allowed for more memory, even if not needed. (A proper fix is
123        implemented in PCRE2, but it involves a lot of refactoring.)
125    10. pcre_get_substring_list() crashed if the use of \K in a match caused the
126        start of the match to be earlier than the end.
128    11. Migrating appropriate PCRE2 JIT improvements to PCRE.
130    12. A pattern such as /(?<=((?C)0))/, which has a callout inside a lookbehind
131        assertion, caused pcretest to generate incorrect output, and also to read
132        uninitialized memory (detected by ASAN or valgrind).
134    13. A pattern that included (*ACCEPT) in the middle of a sufficiently deeply
135        nested set of parentheses of sufficient size caused an overflow of the
136        compiling workspace (which was diagnosed, but of course is not desirable).
138    14. And yet another buffer overflow bug involving duplicate named groups, this
139        time nested, with a nested back reference. Yet again, I have just allowed
140        for more memory, because anything more needs all the refactoring that has
141        been done for PCRE2. An example pattern that provoked this bug is:
142        /((?J)(?'R'(?'R'(?'R'(?'R'(?'R'(?|(\k'R'))))))))/ and the bug was
143        registered as CVE-2016-1283.
145    15. pcretest went into a loop if global matching was requested with an ovector
146        size less than 2. It now gives an error message. This bug was found by
147        afl-fuzz.
149    16. An invalid pattern fragment such as (?(?C)0 was not diagnosing an error
150        ("assertion expected") when (?(?C) was not followed by an opening
151        parenthesis.
153    17. Fixed typo ("&&" for "&") in pcre_study(). Fortunately, this could not
154        actually affect anything, by sheer luck.
156    18. Applied Chris Wilson's patch (Bugzilla #1681) to CMakeLists.txt for MSVC
157        static compilation.
159    19. Modified the RunTest script to incorporate a valgrind suppressions file so
160        that certain errors, provoked by the SSE2 instruction set when JIT is used,
161        are ignored.
163    20. A racing condition is fixed in JIT reported by Mozilla.
165    21. Minor code refactor to avoid "array subscript is below array bounds"
166        compiler warning.
168    22. Minor code refactor to avoid "left shift of negative number" warning.
170    23. Fix typo causing compile error when 16- or 32-bit JIT is compiled without
171        UCP support.
173    24. Refactor to avoid compiler warnings in pcrecpp.cc.
175    25. Refactor to fix a typo in pcre_jit_test.c
177    26. Patch to support compiling pcrecpp.cc with Intel compiler.
180    Version 8.38 23-November-2015
181    -----------------------------
183    1.  If a group that contained a recursive back reference also contained a
184        forward reference subroutine call followed by a non-forward-reference
185        subroutine call, for example /.((?2)(?R)\1)()/, pcre_compile() failed to
186        compile correct code, leading to undefined behaviour or an internally
187        detected error. This bug was discovered by the LLVM fuzzer.
189    2.  Quantification of certain items (e.g. atomic back references) could cause
190        incorrect code to be compiled when recursive forward references were
191        involved. For example, in this pattern: /(?1)()((((((\1++))\x85)+)|))/.
192        This bug was discovered by the LLVM fuzzer.
194    3.  A repeated conditional group whose condition was a reference by name caused
195        a buffer overflow if there was more than one group with the given name.
196        This bug was discovered by the LLVM fuzzer.
198    4.  A recursive back reference by name within a group that had the same name as
199        another group caused a buffer overflow. For example:
200        /(?J)(?'d'(?'d'\g{d}))/. This bug was discovered by the LLVM fuzzer.
202    5.  A forward reference by name to a group whose number is the same as the
203        current group, for example in this pattern: /(?|(\k'Pm')|(?'Pm'))/, caused
204        a buffer overflow at compile time. This bug was discovered by the LLVM
205        fuzzer.
207    6.  A lookbehind assertion within a set of mutually recursive subpatterns could
208        provoke a buffer overflow. This bug was discovered by the LLVM fuzzer.
210    7.  Another buffer overflow bug involved duplicate named groups with a
211        reference between their definition, with a group that reset capture
212        numbers, for example: /(?J:(?|(?'R')(\k'R')|((?'R'))))/. This has been
213        fixed by always allowing for more memory, even if not needed. (A proper fix
214        is implemented in PCRE2, but it involves more refactoring.)
216    8.  There was no check for integer overflow in subroutine calls such as (?123).
218    9.  The table entry for \l in EBCDIC environments was incorrect, leading to its
219        being treated as a literal 'l' instead of causing an error.
221    10. There was a buffer overflow if pcre_exec() was called with an ovector of
222        size 1. This bug was found by american fuzzy lop.
224    11. If a non-capturing group containing a conditional group that could match
225        an empty string was repeated, it was not identified as matching an empty
226        string itself. For example: /^(?:(?(1)x|)+)+$()/.
228    12. In an EBCDIC environment, pcretest was mishandling the escape sequences
229        \a and \e in test subject lines.
231    13. In an EBCDIC environment, \a in a pattern was converted to the ASCII
232        instead of the EBCDIC value.
234    14. The handling of \c in an EBCDIC environment has been revised so that it is
235        now compatible with the specification in Perl's perlebcdic page.
237    15. The EBCDIC character 0x41 is a non-breaking space, equivalent to 0xa0 in
238        ASCII/Unicode. This has now been added to the list of characters that are
239        recognized as white space in EBCDIC.
241    16. When PCRE was compiled without UCP support, the use of \p and \P gave an
242        error (correctly) when used outside a class, but did not give an error
243        within a class.
245    17. \h within a class was incorrectly compiled in EBCDIC environments.
247    18. A pattern with an unmatched closing parenthesis that contained a backward
248        assertion which itself contained a forward reference caused buffer
249        overflow. And example pattern is: /(?=di(?<=(?1))|(?=(.))))/.
251    19. JIT should return with error when the compiled pattern requires more stack
252        space than the maximum.
254    20. A possessively repeated conditional group that could match an empty string,
255        for example, /(?(R))*+/, was incorrectly compiled.
257    21. Fix infinite recursion in the JIT compiler when certain patterns such as
258        /(?:|a|){100}x/ are analysed.
260    22. Some patterns with character classes involving [: and \\ were incorrectly
261        compiled and could cause reading from uninitialized memory or an incorrect
262        error diagnosis.
264    23. Pathological patterns containing many nested occurrences of [: caused
265        pcre_compile() to run for a very long time.
267    24. A conditional group with only one branch has an implicit empty alternative
268        branch and must therefore be treated as potentially matching an empty
269        string.
271    25. If (?R was followed by - or + incorrect behaviour happened instead of a
272        diagnostic.
274    26. Arrange to give up on finding the minimum matching length for overly
275        complex patterns.
277    27. Similar to (4) above: in a pattern with duplicated named groups and an
278        occurrence of (?| it is possible for an apparently non-recursive back
279        reference to become recursive if a later named group with the relevant
280        number is encountered. This could lead to a buffer overflow. Wen Guanxing
281        from Venustech ADLAB discovered this bug.
283    28. If pcregrep was given the -q option with -c or -l, or when handling a
284        binary file, it incorrectly wrote output to stdout.
286    29. The JIT compiler did not restore the control verb head in case of *THEN
287        control verbs. This issue was found by Karl Skomski with a custom LLVM
288        fuzzer.
290    30. Error messages for syntax errors following \g and \k were giving inaccurate
291        offsets in the pattern.
293    31. Added a check for integer overflow in conditions (?(<digits>) and
294        (?(R<digits>). This omission was discovered by Karl Skomski with the LLVM
295        fuzzer.
297    32. Handling recursive references such as (?2) when the reference is to a group
298        later in the pattern uses code that is very hacked about and error-prone.
299        It has been re-written for PCRE2. Here in PCRE1, a check has been added to
300        give an internal error if it is obvious that compiling has gone wrong.
302    33. The JIT compiler should not check repeats after a {0,1} repeat byte code.
303        This issue was found by Karl Skomski with a custom LLVM fuzzer.
305    34. The JIT compiler should restore the control chain for empty possessive
306        repeats. This issue was found by Karl Skomski with a custom LLVM fuzzer.
308    35. Match limit check added to JIT recursion. This issue was found by Karl
309        Skomski with a custom LLVM fuzzer.
311    36. Yet another case similar to 27 above has been circumvented by an
312        unconditional allocation of extra memory. This issue is fixed "properly" in
313        PCRE2 by refactoring the way references are handled. Wen Guanxing
314        from Venustech ADLAB discovered this bug.
316    37. Fix two assertion fails in JIT. These issues were found by Karl Skomski
317        with a custom LLVM fuzzer.
319    38. Fixed a corner case of range optimization in JIT.
321    39. An incorrect error "overran compiling workspace" was given if there were
322        exactly enough group forward references such that the last one extended
323        into the workspace safety margin. The next one would have expanded the
324        workspace. The test for overflow was not including the safety margin.
326    40. A match limit issue is fixed in JIT which was found by Karl Skomski
327        with a custom LLVM fuzzer.
329    41. Remove the use of /dev/null in testdata/testinput2, because it doesn't
330        work under Windows. (Why has it taken so long for anyone to notice?)
332    42. In a character class such as [\W\p{Any}] where both a negative-type escape
333        ("not a word character") and a property escape were present, the property
334        escape was being ignored.
336    43. Fix crash caused by very long (*MARK) or (*THEN) names.
338    44. A sequence such as [[:punct:]b] that is, a POSIX character class followed
339        by a single ASCII character in a class item, was incorrectly compiled in
340        UCP mode. The POSIX class got lost, but only if the single character
341        followed it.
343    45. [:punct:] in UCP mode was matching some characters in the range 128-255
344        that should not have been matched.
346    46. If [:^ascii:] or [:^xdigit:] or [:^cntrl:] are present in a non-negated
347        class, all characters with code points greater than 255 are in the class.
348        When a Unicode property was also in the class (if PCRE_UCP is set, escapes
349        such as \w are turned into Unicode properties), wide characters were not
350        correctly handled, and could fail to match.
353    Version 8.37 28-April-2015
354    --------------------------
356    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
357        for those parentheses to be closed with whatever has been captured so far.
358        However, it was failing to mark any other groups between the hightest
359        capture so far and the currrent group as "unset". Thus, the ovector for
360        those groups contained whatever was previously there. An example is the
361        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
363    2.  If an assertion condition was quantified with a minimum of zero (an odd
364        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
366    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
367        unrecognized modifier, a crash could occur.
369    4.  An attempt to do global matching in pcretest with a zero-length ovector
370        caused a crash.
372    5.  Fixed a memory leak during matching that could occur for a subpattern
373        subroutine call (recursive or otherwise) if the number of captured groups
374        that had to be saved was greater than ten.
376    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
377        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
378        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
380    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
381        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
383    8. When a pattern is compiled, it remembers the highest back reference so that
384       when matching, if the ovector is too small, extra memory can be obtained to
385       use instead. A conditional subpattern whose condition is a check on a
386       capture having happened, such as, for example in the pattern
387       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
388       setting the highest backreference number. This mattered only if pcre_exec()
389       was called with an ovector that was too small to hold the capture, and there
390       was no other kind of back reference (a situation which is probably quite
391       rare). The effect of the bug was that the condition was always treated as
392       FALSE when the capture could not be consulted, leading to a incorrect
393       behaviour by pcre_exec(). This bug has been fixed.
395    9. A reference to a duplicated named group (either a back reference or a test
396       for being set in a conditional) that occurred in a part of the pattern where
397       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
398       to be incorrectly calculated, leading to overwriting.
400    10. A mutually recursive set of back references such as (\2)(\1) caused a
401        segfault at study time (while trying to find the minimum matching length).
402        The infinite loop is now broken (with the minimum length unset, that is,
403        zero).
405    11. If an assertion that was used as a condition was quantified with a minimum
406        of zero, matching went wrong. In particular, if the whole group had
407        unlimited repetition and could match an empty string, a segfault was
408        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
409        assertions to be quantified, but not if they are being used as conditions,
410        so the above pattern is faulted by Perl. PCRE has now been changed so that
411        it also rejects such patterns.
413    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
414        failed to allow the zero-repeat case if pcre2_exec() was called with an
415        ovector too small to capture the group.
417    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
418        Red Hat Product Security:
420        (a) A crash if /K and /F were both set with the option to save the compiled
421        pattern.
423        (b) Another crash if the option to print captured substrings in a callout
424        was combined with setting a null ovector, for example \O\C+ as a subject
425        string.
427    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
428        forward reference repeated a large (but limited) number of times within a
429        repeated outer group that has a zero minimum quantifier, caused incorrect
430        code to be compiled, leading to the error "internal error:
431        previously-checked referenced subpattern not found" when an incorrect
432        memory address was read. This bug was reported as "heap overflow",
433        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
434        CVE-2015-2325.
436    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
437        call within a group that also contained a recursive back reference caused
438        incorrect code to be compiled. This bug was reported as "heap overflow",
439        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
440        number CVE-2015-2326.
442    24. Computing the size of the JIT read-only data in advance has been a source
443        of various issues, and new ones are still appear unfortunately. To fix
444        existing and future issues, size computation is eliminated from the code,
445        and replaced by on-demand memory allocation.
447    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
448        adjacent to the end of the range, and the range contained characters with
449        more than one other case, caused incorrect behaviour when compiled in UTF
450        mode. In that example, the range a-j was left out of the class.
452    26. Fix JIT compilation of conditional blocks, which assertion
453        is converted to (*FAIL). E.g: /(?(?!))/.
455    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
456        discovered by the LLVM fuzzer.
458    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
459        when this assertion was used as a condition, for example (?(?!)a|b). In
460        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
461        error about an unsupported item.
463    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
464        possessification code could take exponential time to complete. A recursion
465        depth limit of 1000 has been imposed to limit the resources used by this
466        optimization.
468    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
469        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
470        because \S ensures they are all in the class. The code for doing this was
471        interacting badly with the code for computing the amount of space needed to
472        compile the pattern, leading to a buffer overflow. This bug was discovered
473        by the LLVM fuzzer.
475    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
476        other kinds of group caused stack overflow at compile time. This bug was
477        discovered by the LLVM fuzzer.
479    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
480        between a subroutine call and its quantifier was incorrectly compiled,
481        leading to buffer overflow or other errors. This bug was discovered by the
482        LLVM fuzzer.
484    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
485        assertion after (?(. The code was failing to check the character after
486        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
487        was discovered by the LLVM fuzzer.
489    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
490        a fixed maximum following a group that contains a subroutine reference was
491        incorrectly compiled and could trigger buffer overflow. This bug was
492        discovered by the LLVM fuzzer.
494    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
495        caused a stack overflow instead of the diagnosis of a non-fixed length
496        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
498    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
499        (e.g. /(?<=\Ka)/) could make pcregrep loop.
501    37. There was a similar problem to 36 in pcretest for global matches.
503    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
504        and a subsequent item in the pattern caused a non-match, backtracking over
505        the repeated \X did not stop, but carried on past the start of the subject,
506        causing reference to random memory and/or a segfault. There were also some
507        other cases where backtracking after \C could crash. This set of bugs was
508        discovered by the LLVM fuzzer.
510    39. The function for finding the minimum length of a matching string could take
511        a very long time if mutual recursion was present many times in a pattern,
512        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
513        method has been implemented. This infelicity was discovered by the LLVM
514        fuzzer.
516    40. Static linking against the PCRE library using the pkg-config module was
517        failing on missing pthread symbols.
520    Version 8.36 26-September-2014
521    ------------------------------
523    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
524        -Wmissing-field-initializers and -Wunused-parameter.
526    2.  The tests for quantifiers being too big (greater than 65535) were being
527        applied after reading the number, and stupidly assuming that integer
528        overflow would give a negative number. The tests are now applied as the
529        numbers are read.
531    3.  Tidy code in pcre_exec.c where two branches that used to be different are
532        now the same.
534    4.  The JIT compiler did not generate match limit checks for certain
535        bracketed expressions with quantifiers. This may lead to exponential
536        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
537        issue should be resolved now.
539    5.  Fixed an issue, which occures when nested alternatives are optimized
540        with table jumps.
542    6.  Inserted two casts and changed some ints to size_t in the light of some
543        reported 64-bit compiler warnings (Bugzilla 1477).
545    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
546        an empty string, which sometimes were behaving incorrectly in the
547        interpreter (though correctly in the JIT matcher). This pcretest input is
548        an example:
550          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
553        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
554        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
555        for an empty string was breaking the inner loop and carrying on at a lower
556        level, when possessive repeated groups should always return to a higher
557        level as they have no backtrack points in them. The empty string test now
558        occurs at the outer level.
560    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
561        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
563    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
564        doesn't).
566    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
567        to hold a pattern to be incorrectly computed (too small) when there were
568        named back references to duplicated names. This could cause "internal
569        error: code overflow" or "double free or corruption" or other memory
570        handling errors.
572    11. When named subpatterns had the same prefixes, back references could be
573        confused. For example, in this pattern:
575          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
577        the reference to 'Name' was incorrectly treated as a reference to a
578        duplicate name.
580    12. A pattern such as /^s?c/mi8 where the optional character has more than
581        one "other case" was incorrectly compiled such that it would only try to
582        match starting at "c".
584    13. When a pattern starting with \s was studied, VT was not included in the
585        list of possible starting characters; this should have been part of the
586        8.34/18 patch.
588    14. If a character class started [\Qx]... where x is any character, the class
589        was incorrectly terminated at the ].
591    15. If a pattern that started with a caseless match for a character with more
592        than one "other case" was studied, PCRE did not set up the starting code
593        unit bit map for the list of possible characters. Now it does. This is an
594        optimization improvement, not a bug fix.
596    16. The Unicode data tables have been updated to Unicode 7.0.0.
598    17. Fixed a number of memory leaks in pcregrep.
600    18. Avoid a compiler warning (from some compilers) for a function call with
601        a cast that removes "const" from an lvalue by using an intermediate
602        variable (to which the compiler does not object).
604    19. Incorrect code was compiled if a group that contained an internal recursive
605        back reference was optional (had quantifier with a minimum of zero). This
606        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
607        caused segmentation faults because of stack overflows at compile time.
609    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
610        group that is quantified with an indefinite repeat, caused a compile-time
611        loop which used up all the system stack and provoked a segmentation fault.
612        This was not the same bug as 19 above.
614    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
615        Patch by Mike Frysinger.
618    Version 8.35 04-April-2014
619    --------------------------
621    1.  A new flag is set, when property checks are present in an XCLASS.
622        When this flag is not set, PCRE can perform certain optimizations
623        such as studying these XCLASS-es.
625    2.  The auto-possessification of character sets were improved: a normal
626        and an extended character set can be compared now. Furthermore
627        the JIT compiler optimizes more character set checks.
629    3.  Got rid of some compiler warnings for potentially uninitialized variables
630        that show up only when compiled with -O2.
632    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
633        of a match later then the end of the match. The pcretest program was not
634        handling the case sensibly - it was outputting from the start to the next
635        binary zero. It now reports this situation in a message, and outputs the
636        text from the end to the start.
638    5.  Fast forward search is improved in JIT. Instead of the first three
639        characters, any three characters with fixed position can be searched.
640        Search order: first, last, middle.
642    6.  Improve character range checks in JIT. Characters are read by an inprecise
643        function now, which returns with an unknown value if the character code is
644        above a certain threshold (e.g: 256). The only limitation is that the value
645        must be bigger than the threshold as well. This function is useful when
646        the characters above the threshold are handled in the same way.
648    7.  The macros whose names start with RAWUCHAR are placeholders for a future
649        mode in which only the bottom 21 bits of 32-bit data items are used. To
650        make this more memorable for those maintaining the code, the names have
651        been changed to start with UCHAR21, and an extensive comment has been added
652        to their definition.
654    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
655        to the export list in Makefile.am (they were accidentally omitted from the
656        8.34 tarball).
658    9.  The informational output from pcretest used the phrase "starting byte set"
659        which is inappropriate for the 16-bit and 32-bit libraries. As the output
660        for "first char" and "need char" really means "non-UTF-char", I've changed
661        "byte" to "char", and slightly reworded the output. The documentation about
662        these values has also been (I hope) clarified.
664    10. Another JIT related optimization: use table jumps for selecting the correct
665        backtracking path, when more than four alternatives are present inside a
666        bracket.
668    11. Empty match is not possible, when the minimum length is greater than zero,
669        and there is no \K in the pattern. JIT should avoid empty match checks in
670        such cases.
672    12. In a caseless character class with UCP support, when a character with more
673        than one alternative case was not the first character of a range, not all
674        the alternative cases were added to the class. For example, s and \x{17f}
675        are both alternative cases for S: the class [RST] was handled correctly,
676        but [R-T] was not.
678    13. The configure.ac file always checked for pthread support when JIT was
679        enabled. This is not used in Windows, so I have put this test inside a
680        check for the presence of windows.h (which was already tested for).
682    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
683        The algorithm provides a way to skip certain starting offsets, and usually
684        faster than linear prefix searches.
686    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
687        as for 'fr_FR' and 'french'. For some reason, however, it then used the
688        Windows-specific input and output files, which have 'french' screwed in.
689        So this could never have worked. One of the problems with locales is that
690        they aren't always the same. I have now updated RunTest so that it checks
691        the output of the locale test (test 3) against three different output
692        files, and it allows the test to pass if any one of them matches. With luck
693        this should make the test pass on some versions of Solaris where it was
694        failing. Because of the uncertainty, the script did not used to stop if
695        test 3 failed; it now does. If further versions of a French locale ever
696        come to light, they can now easily be added.
698    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
699        there was a message during ./configure, but it did not stop. This now
700        provokes an error. The invalid example in README has been corrected.
701        If a value less than the minimum is given, the minimum value has always
702        been used, but now a warning is given.
704    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
705        was a bug in the test system, which is now fixed. Also, the list of various
706        configurations that are tested for each release did not have one with both
707        16/32 bits and --enable-bar-anycrlf. It now does.
709    18. pcretest was missing "-C bsr" for displaying the \R default setting.
711    19. Little endian PowerPC systems are supported now by the JIT compiler.
713    20. The fast forward newline mechanism could enter to an infinite loop on
714        certain invalid UTF-8 input. Although we don't support these cases
715        this issue can be fixed by a performance optimization.
717    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
718        not take account if existing stack usage. There is now a new global
719        variable called pcre_stack_guard that can be set to point to an external
720        function to check stack availability. It is called at the start of
721        processing every parenthesized group.
723    22. A typo in the code meant that in ungreedy mode the max/min qualifier
724        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
725        not match "ab".
727    23. When UTF was disabled, the JIT program reported some incorrect compile
728        errors. These messages are silenced now.
730    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
731        compiler.
733    25. Change all the temporary files used in RunGrepTest to be different to those
734        used by RunTest so that the tests can be run simultaneously, for example by
735        "make -j check".
738    Version 8.34 15-December-2013
739    -----------------------------
741    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
742        executable memory. Patch inspired by Carsten Klein.
744    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
745        this macro is never tested and has no effect, because the work to support
746        coverage involves only compiling and linking options and special targets in
747        the Makefile. The comment in config.h implied that defining the macro would
748        enable coverage support, which is totally false. There was also support for
749        setting this macro in the CMake files (my fault, I just copied it from
750        configure). SUPPORT_GCOV has now been removed.
752    3.  Make a small performance improvement in strlen16() and strlen32() in
753        pcretest.
755    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
756        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
757        these cases). There was also one in pcretest.c.
759    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
761    6.  In UTF mode, the code for checking whether a group could match an empty
762        string (which is used for indefinitely repeated groups to allow for
763        breaking an infinite loop) was broken when the group contained a repeated
764        negated single-character class with a character that occupied more than one
765        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
766        UTF-8 mode). The effect was undefined: the group might or might not be
767        deemed as matching an empty string, or the program might have crashed.
769    7.  The code for checking whether a group could match an empty string was not
770        recognizing that \h, \H, \v, \V, and \R must match a character.
772    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
773        an empty string. If it can, pcretest shows this in its information output.
775    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
776        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
777        matched by pcre_exec() without using JIT:
779        (a) If the rest of the pattern did not match after a maximal run of
780            grapheme clusters, the code for backing up to try with fewer of them
781            did not always back up over a full grapheme when characters that do not
782            have the modifier quality were involved, e.g. Hangul syllables.
784        (b) If the match point in a subject started with modifier character, and
785            there was no match, the code could incorrectly back up beyond the match
786            point, and potentially beyond the first character in the subject,
787            leading to a segfault or an incorrect match result.
789    10. A conditional group with an assertion condition could lead to PCRE
790        recording an incorrect first data item for a match if no other first data
791        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
792        first data item, and therefore matched "ca" after "c" instead of at the
793        start.
795    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
796        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
798    12. The source of pcregrep now includes z/OS-specific code so that it can be
799        compiled for z/OS as part of the special z/OS distribution.
801    13. Added the -T and -TM options to pcretest.
803    14. The code in pcre_compile.c for creating the table of named capturing groups
804        has been refactored. Instead of creating the table dynamically during the
805        actual compiling pass, the information is remembered during the pre-compile
806        pass (on the stack unless there are more than 20 named groups, in which
807        case malloc() is used) and the whole table is created before the actual
808        compile happens. This has simplified the code (it is now nearly 150 lines
809        shorter) and prepared the way for better handling of references to groups
810        with duplicate names.
812    15. A back reference to a named subpattern when there is more than one of the
813        same name now checks them in the order in which they appear in the pattern.
814        The first one that is set is used for the reference. Previously only the
815        first one was inspected. This change makes PCRE more compatible with Perl.
817    16. Unicode character properties were updated from Unicode 6.3.0.
819    17. The compile-time code for auto-possessification has been refactored, based
820        on a patch by Zoltan Herczeg. It now happens after instead of during
821        compilation. The code is cleaner, and more cases are handled. The option
822        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
823        options in pcretest are provided to set it. It can also be set by
824        (*NO_AUTO_POSSESS) at the start of a pattern.
826    18. The character VT has been added to the default ("C" locale) set of
827        characters that match \s and are generally treated as white space,
828        following this same change in Perl 5.18. There is now no difference between
829        "Perl space" and "POSIX space". Whether VT is treated as white space in
830        other locales depends on the locale.
832    19. The code for checking named groups as conditions, either for being set or
833        for being recursed, has been refactored (this is related to 14 and 15
834        above). Processing unduplicated named groups should now be as fast at
835        numerical groups, and processing duplicated groups should be faster than
836        before.
838    20. Two patches to the CMake build system, by Alexander Barkov:
840          (1) Replace the "source" command by "." in CMakeLists.txt because
841              "source" is a bash-ism.
843          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
844              without these the CMake build does not work on Solaris.
846    21. Perl has changed its handling of \8 and \9. If there is no previously
847        encountered capturing group of those numbers, they are treated as the
848        literal characters 8 and 9 instead of a binary zero followed by the
849        literals. PCRE now does the same.
851    22. Following Perl, added \o{} to specify codepoints in octal, making it
852        possible to specify values greater than 0777 and also making them
853        unambiguous.
855    23. Perl now gives an error for missing closing braces after \x{... instead of
856        treating the string as literal. PCRE now does the same.
858    24. RunTest used to grumble if an inappropriate test was selected explicitly,
859        but just skip it when running all tests. This make it awkward to run ranges
860        of tests when one of them was inappropriate. Now it just skips any
861        inappropriate tests, as it always did when running all tests.
863    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
864        character types such as \d or \w, too many callouts were inserted, and the
865        data that they returned was rubbish.
867    26. In UCP mode, \s was not matching two of the characters that Perl matches,
868        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
869        were matched by \h. The code has now been refactored so that the lists of
870        the horizontal and vertical whitespace characters used for \h and \v (which
871        are defined only in one place) are now also used for \s.
873    27. Add JIT support for the 64 bit TileGX architecture.
874        Patch by Jiong Wang (Tilera Corporation).
876    28. Possessive quantifiers for classes (both explicit and automatically
877        generated) now use special opcodes instead of wrapping in ONCE brackets.
879    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
880        (because it's meaningless), this was not happening when PCRE_CASELESS was
881        set. Not wrong, but inefficient.
883    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
884        properties for \w, \d, etc) is present in a test regex. Otherwise if the
885        test contains no characters greater than 255, Perl doesn't realise it
886        should be using Unicode semantics.
888    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
889        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
890        does in Unicode mode.
892    32. Added the "forbid" facility to pcretest so that putting tests into the
893        wrong test files can sometimes be quickly detected.
895    33. There is now a limit (default 250) on the depth of nesting of parentheses.
896        This limit is imposed to control the amount of system stack used at compile
897        time. It can be changed at build time by --with-parens-nest-limit=xxx or
898        the equivalent in CMake.
900    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
901        errors. Perl warns for these when in warning mode, but PCRE has no facility
902        for giving warnings.
904    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
905        However, this was not working for (?!) because it is optimized to (*FAIL),
906        for which PCRE does not allow quantifiers. The optimization is now disabled
907        when a quantifier follows (?!). I can't see any use for this, but it makes
908        things uniform.
910    36. Perl no longer allows group names to start with digits, so I have made this
911        change also in PCRE. It simplifies the code a bit.
913    37. In extended mode, Perl ignores spaces before a + that indicates a
914        possessive quantifier. PCRE allowed a space before the quantifier, but not
915        before the possessive +. It now does.
917    38. The use of \K (reset reported match start) within a repeated possessive
918        group such as (a\Kb)*+ was not working.
920    40. Document that the same character tables must be used at compile time and
921        run time, and that the facility to pass tables to pcre_exec() and
922        pcre_dfa_exec() is for use only with saved/restored patterns.
924    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
925        features for Builds with MSVC:
927        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
928           builds). The .rc files can be used to set FileDescription and many other
929           attributes.
931        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
932           This allows higher-level build scripts which want .pdb files to avoid
933           hard-coding the exact files needed."
935    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
936        mean "start of word" and "end of word", respectively, as a transition aid.
938    43. A minimizing repeat of a class containing codepoints greater than 255 in
939        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
940        compiled to use the heap for recursion.
942    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
943        is configured.
946    Version 8.33 28-May-2013
947    ------------------------
949    1.  Added 'U' to some constants that are compared to unsigned integers, to
950        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
951        variables that are added to signed variables, to ensure the result is
952        signed and can be negated.
954    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
955        CMake config files.
957    3.  Revise the creation of config.h.generic so that all boolean macros are
958        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
959        overriding via -D on the command line possible.
961    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
962        to unsigned int is reported to make a quite noticeable speed difference in
963        a specific Windows environment. Testing on Linux did also appear to show
964        some benefit (and it is clearly not harmful). Also fixed the definition of
965        Xop which should be unsigned.
967    5.  Related to (4), changing the definition of the intermediate variable cc
968        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
969        improvements.
971    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
972        unnecessary spaces.
974    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
975        and later.
977    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
979    9.  Optimizing fast_forward_start_bits in JIT.
981    10. Adding support for callouts in JIT, and fixing some issues revealed
982        during this work. Namely:
984        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
986        (b) Minimum length was not checked before the matching is started.
988    11. The value of capture_last that is passed to callouts was incorrect in some
989        cases when there was a capture on one path that was subsequently abandoned
990        after a backtrack. Also, the capture_last value is now reset after a
991        recursion, since all captures are also reset in this case.
993    12. The interpreter no longer returns the "too many substrings" error in the
994        case when an overflowing capture is in a branch that is subsequently
995        abandoned after a backtrack.
997    13. In the pathological case when an offset vector of size 2 is used, pcretest
998        now prints out the matched string after a yield of 0 or 1.
1000    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
1001        Only supported by the JIT compiler at the moment.
1003    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
1005    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
1006        offset of the starting point of the matching process, provided the offsets
1007        vector is large enough.
1009    17. The \A escape now records a lookbehind value of 1, though its execution
1010        does not actually inspect the previous character. This is to ensure that,
1011        in partial multi-segment matching, at least one character from the old
1012        segment is retained when a new segment is processed. Otherwise, if there
1013        are no lookbehinds in the pattern, \A might match incorrectly at the start
1014        of a new segment.
1016    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
1018    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
1019        gives some modest performance improvement in 8-bit mode.
1021    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
1022        be expressed in certain programming languages using Universal Character
1023        Names.
1025    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
1026        which points out that "non characters" are not "characters that may not
1027        appear in Unicode strings" but rather "characters that are reserved for
1028        internal use and have only local meaning".
1030    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
1031        there was a conditional group that depended on an assertion, if the
1032        assertion was false, the callout that immediately followed the alternation
1033        in the condition was skipped when pcre_exec() was used for matching.
1035    23. Allow an explicit callout to be inserted before an assertion that is the
1036        condition for a conditional group, for compatibility with automatic
1037        callouts, which always insert a callout at this point.
1039    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
1040        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
1042    25. (*PRUNE) is now supported by the JIT compiler.
1044    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
1046    27. Fix the case where there are two or more SKIPs with arguments that may be
1047        ignored.
1049    28. (*SKIP) is now supported by the JIT compiler.
1051    29. (*THEN) is now supported by the JIT compiler.
1053    30. Update RunTest with additional test selector options.
1055    31. The way PCRE handles backtracking verbs has been changed in two ways.
1057        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
1058        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
1059        backtracking. In some cases this makes it more Perl-compatible, but Perl's
1060        rather obscure rules do not always do the same thing.
1062        (2) Previously, backtracking verbs were confined within assertions. This is
1063        no longer the case for positive assertions, except for (*ACCEPT). Again,
1064        this sometimes improves Perl compatibility, and sometimes does not.
1066    32. A number of tests that were in test 2 because Perl did things differently
1067        have been moved to test 1, because either Perl or PCRE has changed, and
1068        these tests are now compatible.
1070    32. Backtracking control verbs are now handled in the same way in JIT and
1071        interpreter.
1073    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
1074        contained a forward subroutine reference caused a compile error.
1076    34. Auto-detect and optimize limited repetitions in JIT.
1078    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
1079        blocking (*UTF) etc.
1081    36. In the interpreter, maximizing pattern repetitions for characters and
1082        character types now use tail recursion, which reduces stack usage.
1084    37. The value of the max lookbehind was not correctly preserved if a compiled
1085        and saved regex was reloaded on a host of different endianness.
1087    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
1088        of the compiled pattern block, expand the flags field from 16 to 32 bits
1089        because it was almost full.
1091    39. Try madvise first before posix_madvise.
1093    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
1094        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
1095        original change locked out all matches of empty strings. This has been
1096        changed so that one match of an empty string per line is recognized.
1097        Subsequent searches on the same line (for colouring or for --only-matching,
1098        for example) do not recognize empty strings.
1100    41. Applied a user patch to fix a number of spelling mistakes in comments.
1102    42. Data lines longer than 65536 caused pcretest to crash.
1104    43. Clarified the data type for length and startoffset arguments for pcre_exec
1105        and pcre_dfa_exec in the function-specific man pages, where they were
1106        explicitly stated to be in bytes, never having been updated. I also added
1107        some clarification to the pcreapi man page.
1109    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
1110        a segmentation fault.
1113    Version 8.32 30-November-2012
1114    -----------------------------
1116    1.  Improved JIT compiler optimizations for first character search and single
1117        character iterators.
1119    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
1120        Patch by Daniel Richard G.
1122    3.  Single character iterator optimizations in the JIT compiler.
1124    4.  Improved JIT compiler optimizations for character ranges.
1126    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
1127        Reported by Giuseppe D'Angelo.
1129    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
1130        of a line, was being set incorrectly in cases where .* appeared inside
1131        atomic brackets at the start of a pattern, or where there was a subsequent
1132        *PRUNE or *SKIP.
1134    7.  Improved instruction cache flush for POWER/PowerPC.
1135        Patch by Daniel Richard G.
1137    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
1138        grep:
1140        (a) There is now no limit to the number of patterns to be matched.
1142        (b) An error is given if a pattern is too long.
1144        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
1145            are now supported.
1147        (d) --exclude-from and --include-from (multiple use) have been added.
1149        (e) Exclusions and inclusions now apply to all files and directories, not
1150            just to those obtained from scanning a directory recursively.
1152        (f) Multiple uses of -f and --file-list are now supported.
1154        (g) In a Windows environment, the default for -d has been changed from
1155            "read" (the GNU grep default) to "skip", because otherwise the presence
1156            of a directory in the file list provokes an error.
1158        (h) The documentation has been revised and clarified in places.
1160    9.  Improve the matching speed of capturing brackets.
1162    10. Changed the meaning of \X so that it now matches a Unicode extended
1163        grapheme cluster.
1165    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
1166        out POSIX threads when JIT support is configured.
1168    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
1170    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
1171        structure could go wrong in environments where size_t is not the same size
1172        as int.
1174    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
1175        set.
1177    15. The EBCDIC support had decayed; later updates to the code had included
1178        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
1179        general tidy up of EBCDIC-related issues, and the documentation was also
1180        not quite right. There is now a test that can be run on ASCII systems to
1181        check some of the EBCDIC-related things (but is it not a full test).
1183    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
1184        in a small tidy to the code.
1186    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
1188    18. If the --only-matching (-o) option in pcregrep is specified multiple
1189        times, each one causes appropriate output. For example, -o1 -o2 outputs the
1190        substrings matched by the 1st and 2nd capturing parentheses. A separating
1191        string can be specified by --om-separator (default empty).
1193    19. Improving the first n character searches.
1195    20. Turn case lists for horizontal and vertical white space into macros so that
1196        they are defined only once.
1198    21. This set of changes together give more compatible Unicode case-folding
1199        behaviour for characters that have more than one other case when UCP
1200        support is available.
1202        (a) The Unicode property table now has offsets into a new table of sets of
1203            three or more characters that are case-equivalent. The MultiStage2.py
1204            script that generates these tables (the pcre_ucd.c file) now scans
1205            CaseFolding.txt instead of UnicodeData.txt for character case
1206            information.
1208        (b) The code for adding characters or ranges of characters to a character
1209            class has been abstracted into a generalized function that also handles
1210            case-independence. In UTF-mode with UCP support, this uses the new data
1211            to handle characters with more than one other case.
1213        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
1214            whose other case is greater than 256 are now correctly matched
1215            caselessly. Previously, the high codepoint matched the low one, but not
1216            vice versa.
1218        (d) The processing of \h, \H, \v, and \ in character classes now makes use
1219            of the new class addition function, using character lists defined as
1220            macros alongside the case definitions of 20 above.
1222        (e) Caseless back references now work with characters that have more than
1223            one other case.
1225        (f) General caseless matching of characters with more than one other case
1226            is supported.
1228    22. Unicode character properties were updated from Unicode 6.2.0
1230    23. Improved CMake support under Windows. Patch by Daniel Richard G.
1232    24. Add support for 32-bit character strings, and UTF-32
1234    25. Major JIT compiler update (code refactoring and bugfixing).
1235        Experimental Sparc 32 support is added.
1237    26. Applied a modified version of Daniel Richard G's patch to create
1238        pcre.h.generic and config.h.generic by "make" instead of in the
1239        PrepareRelease script.
1241    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
1242        pcre_compile.c when checking for a zero character.
1244    28. Introducing a native interface for JIT. Through this interface, the compiled
1245        machine code can be directly executed. The purpose of this interface is to
1246        provide fast pattern matching, so several sanity checks are not performed.
1247        However, feature tests are still performed. The new interface provides
1248        1.4x speedup compared to the old one.
1250    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
1251        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
1252        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
1254    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
1255        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
1256        the "old" RFC 2279). Instead, it ended up passing a negative length to
1257        pcre_exec().
1259    31. Add support for GCC's visibility feature to hide internal functions.
1261    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
1262        "unknown -C option" after outputting 0 or 1.
1264    33. There is now support for generating a code coverage report for the test
1265        suite in environments where gcc is the compiler and lcov is installed. This
1266        is mainly for the benefit of the developers.
1268    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
1269        unaddressable using valgrind annotations, allowing valgrind to detect
1270        invalid memory accesses. This is mainly for the benefit of the developers.
1272    25. (*UTF) can now be used to start a pattern in any of the three libraries.
1274    26. Give configure error if --enable-cpp but no C++ compiler found.
1277    Version 8.31 06-July-2012
1278    -------------------------
1280    1.  Fixing a wrong JIT test case and some compiler warnings.
1282    2.  Removed a bashism from the RunTest script.
1284    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
1285        to unsigned type, result still unsigned" that was given by an MS compiler
1286        on encountering the code "-sizeof(xxx)".
1288    4.  Partial matching support is added to the JIT compiler.
1290    5.  Fixed several bugs concerned with partial matching of items that consist
1291        of more than one character:
1293        (a) /^(..)\1/ did not partially match "aba" because checking references was
1294            done on an "all or nothing" basis. This also applied to repeated
1295            references.
1297        (b) \R did not give a hard partial match if \r was found at the end of the
1298            subject.
1300        (c) \X did not give a hard partial match after matching one or more
1301            characters at the end of the subject.
1303        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1304            a partial match for the string "\r".
1306        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1307            a partial match for a CR character at the end of the subject string.
1309    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1310        running pcretest, the text "(JIT)" added to the output whenever JIT is
1311        actually used to run the match.
1313    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1314        or /S+[+] with a digit between 1 and 7.
1316    8.  OP_NOT now supports any UTF character not just single-byte ones.
1318    9.  (*MARK) control verb is now supported by the JIT compiler.
1320    10. The command "./RunTest list" lists the available tests without actually
1321        running any of them. (Because I keep forgetting what they all are.)
1325    12. Applied a (slightly modified) user-supplied patch that improves performance
1326        when the heap is used for recursion (compiled with --disable-stack-for-
1327        recursion). Instead of malloc and free for each heap frame each time a
1328        logical recursion happens, frames are retained on a chain and re-used where
1329        possible. This sometimes gives as much as 30% improvement.
1331    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1332        call.
1334    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1336    15. It is now possible to link pcretest with libedit as an alternative to
1337        libreadline.
1339    16. (*COMMIT) control verb is now supported by the JIT compiler.
1341    17. The Unicode data tables have been updated to Unicode 6.1.0.
1343    18. Added --file-list option to pcregrep.
1345    19. Added binary file support to pcregrep, including the -a, --binary-files,
1346        -I, and --text options.
1348    20. The madvise function is renamed for posix_madvise for QNX compatibility
1349        reasons. Fixed by Giuseppe D'Angelo.
1351    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1353    22. Changed the option for creating symbolic links for 16-bit man pages from
1354        -s to -sf so that re-installing does not cause issues.
1356    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1358    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1359        matches in certain environments (the workspace was not being correctly
1360        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1361        some of the workspace data at the beginning of a restart.
1363    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1364        was not doing so when it should - probably a typo introduced by SVN 528
1365        (change 8.10/14).
1367    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1368        \w+ when the character tables indicated that \x{c4} was a word character.
1369        There were several related cases, all because the tests for doing a table
1370        lookup were testing for characters less than 127 instead of 255.
1372    27. If a pattern contains capturing parentheses that are not used in a match,
1373        their slots in the ovector are set to -1. For those that are higher than
1374        any matched groups, this happens at the end of processing. In the case when
1375        there were back references that the ovector was too small to contain
1376        (causing temporary malloc'd memory to be used during matching), and the
1377        highest capturing number was not used, memory off the end of the ovector
1378        was incorrectly being set to -1. (It was using the size of the temporary
1379        memory instead of the true size.)
1381    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1382        ovector size, it uses memory at the end of the block that it has got.
1384    29. Check for an overlong MARK name and give an error at compile time. The
1385        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1387    30. JIT compiler update.
1389    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1390        Rill for the patch.
1392    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1394    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1396    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1397        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1398        was enabled.
1400    35. Improve JIT code generation for greedy plus quantifier.
1402    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1403        1 to "aa" instead of to an empty string. The bug affected repeated groups
1404        that could potentially match an empty string.
1406    37. Optimizing single character iterators in JIT.
1408    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1409        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1410        codepoints that are too big for the mode are faulted, and in a UTF mode,
1411        disallowed codepoints are also faulted.
1413    39. If PCRE was compiled with UTF support, in three places in the DFA
1414        matcher there was code that should only have been obeyed in UTF mode, but
1415        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1416        processing when bytes with values greater than 127 were present. In 16-bit
1417        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1418        both cases the values are those that cannot be the first data item in a UTF
1419        character. The three items that might have provoked this were recursions,
1420        possessively repeated groups, and atomic groups.
1422    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1423        and pcregrep, because some OS require shared objects to be explicitly
1424        passed to ld, causing the link step to fail if they are not.
1426    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1427        mode, patterns that started with \h* or \R* might be incorrectly matched.
1430    Version 8.30 04-February-2012
1431    -----------------------------
1433    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1434        name is defined in ctype.h.
1436    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1437        only in quite long subpatterns.
1439    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1440        since it was replaced by pcre_fullinfo() in February 2000.
1442    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1443        match a (*MARK), and the match failed at the start of the subject, a
1444        reference to memory before the start of the subject could occur. This bug
1445        was introduced by fix 17 of release 8.21.
1447    5.  A reference to an unset group with zero minimum repetition was giving
1448        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1449        /(another)?(\1?)test/ matched against "hello world test". This bug was
1450        introduced in release 8.13.
1452    6.  Add support for 16-bit character strings (a large amount of work involving
1453        many changes and refactorings).
1455    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1456        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1457        from a file.
1459    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1460        rounding is not applied in this particular case).
1462    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1463        if they appear, or are escaped, in patterns.
1465    10. Get rid of a number of -Wunused-but-set-variable warnings.
1467    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1468        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1469        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1470        also returns the mark "x". This bug applied to capturing parentheses,
1471        non-capturing parentheses, and atomic parentheses. It also applied to some
1472        assertions.
1474    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1475        information out of configure.ac instead of relying on pcre.h.generic, which
1476        is not stored in the repository.
1478    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1479        -lreadline.
1481    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1483    15. Applied Graycode's patch to put the top-level frame on the stack rather
1484        than the heap when not using the stack for recursion. This gives a
1485        performance improvement in many cases when recursion is not deep.
1487    16. Experimental code added to "pcretest -C" to output the stack frame size.
1490    Version 8.21 12-Dec-2011
1491    ------------------------
1493    1.  Updating the JIT compiler.
1495    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1496        are added as well.
1498    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1499        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1500        calling _pcre_jit_exec. Some extra comments are added.
1502    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1503        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1504        was introduced by change 18 for 8.20.
1506    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1507        ECMA-262 standard.
1509    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1510        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1511        This bug was probably introduced by change 9 of 8.13.
1513    7.  While fixing 6 above, I noticed that a number of other items were being
1514        incorrectly rejected as "not fixed length". This arose partly because newer
1515        opcodes had not been added to the fixed-length checking code. I have (a)
1516        corrected the bug and added tests for these items, and (b) arranged for an
1517        error to occur if an unknown opcode is encountered while checking for fixed
1518        length instead of just assuming "not fixed length". The items that were
1519        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1520        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1521        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1523    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1524        being incorrectly compiled and would have given unpredicatble results.
1526    9.  A possessively repeated subpattern with minimum repeat count greater than
1527        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1528        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1529        the first (A) could occur when it should not.
1531    10. Add a cast and remove a redundant test from the code.
1533    11. JIT should use pcre_malloc/pcre_free for allocation.
1535    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1536        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1537        is anything other than /usr, -L is still shown).
1539    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1541    14. Perl does not support \N without a following name in a [] class; PCRE now
1542        also gives an error.
1544    15. If a forward reference was repeated with an upper limit of around 2000,
1545        it caused the error "internal error: overran compiling workspace". The
1546        maximum number of forward references (including repeats) was limited by the
1547        internal workspace, and dependent on the LINK_SIZE. The code has been
1548        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1549        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1550        of around 200,000 forward references. While doing this, I also speeded up
1551        the filling in of repeated forward references.
1553    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1554        incorrectly expecting the subject to contain another "a" after the start.
1556    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1557        in the match, the SKIP should be ignored. This was not happening; instead
1558        the SKIP was being treated as NOMATCH. For patterns such as
1559        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1560        tested.
1562    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1563        now much more compatible with Perl, in particular in cases where the result
1564        is a non-match for a non-anchored pattern. For example, if
1565        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1566        "m", where previously it did not return a name. A side effect of this
1567        change is that for partial matches, the last encountered mark name is
1568        returned, as for non matches. A number of tests that were previously not
1569        Perl-compatible have been moved into the Perl-compatible test files. The
1570        refactoring has had the pleasing side effect of removing one argument from
1571        the match() function, thus reducing its stack requirements.
1573    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1574        subsequent uses of /S (without +) incorrectly behaved like /S+.
1576    21. Retrieve executable code size support for the JIT compiler and fixing
1577        some warnings.
1579    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1580        not work when the shorter character appeared right at the end of the
1581        subject string.
1583    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1584        systems.
1586    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1587        output it when the /M option is used in pcretest.
1589    25. The CheckMan script was not being included in the distribution. Also, added
1590        an explicit "perl" to run Perl scripts from the PrepareRelease script
1591        because this is reportedly needed in Windows.
1593    26. If study data was being save in a file and studying had not found a set of
1594        "starts with" bytes for the pattern, the data written to the file (though
1595        never used) was taken from uninitialized memory and so caused valgrind to
1596        complain.
1598    27. Updated RunTest.bat as provided by Sheri Pierce.
1600    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1602    29. Computation of memory usage for the table of capturing group names was
1603        giving an unnecessarily large value.
1606    Version 8.20 21-Oct-2011
1607    ------------------------
1609    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1610        a POSIX class. After further experiments with Perl, which convinced me that
1611        Perl has bugs and confusions, a closing square bracket is no longer allowed
1612        in a POSIX name. This bug also affected patterns with classes that started
1613        with full stops.
1615    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1616        captured substring, but while checking the failing first alternative,
1617        substring 1 is temporarily captured. If the output vector supplied to
1618        pcre_exec() was not big enough for this capture, the yield of the function
1619        was still zero ("insufficient space for captured substrings"). This cannot
1620        be totally fixed without adding another stack variable, which seems a lot
1621        of expense for a edge case. However, I have improved the situation in cases
1622        such as /(a)(b)x|abc/ matched against "abc", where the return code
1623        indicates that fewer than the maximum number of slots in the ovector have
1624        been set.
1626    3.  Related to (2) above: when there are more back references in a pattern than
1627        slots in the output vector, pcre_exec() uses temporary memory during
1628        matching, and copies in the captures as far as possible afterwards. It was
1629        using the entire output vector, but this conflicts with the specification
1630        that only 2/3 is used for passing back captured substrings. Now it uses
1631        only the first 2/3, for compatibility. This is, of course, another edge
1632        case.
1634    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1635        main code base, and can be used by building with --enable-jit. When this is
1636        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1637        runtime --no-jit option is given.
1639    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1640        ovector, the return from the function was zero, implying that there were
1641        other matches that did not fit. The correct "exactly full" value is now
1642        returned.
1644    6.  If a subpattern that was called recursively or as a subroutine contained
1645        (*PRUNE) or any other control that caused it to give a non-standard return,
1646        invalid errors such as "Error -26 (nested recursion at the same subject
1647        position)" or even infinite loops could occur.
1649    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1650        computing the minimum length on reaching *ACCEPT, and so ended up with the
1651        wrong value of 1 rather than 0. Further investigation indicates that
1652        computing a minimum subject length in the presence of *ACCEPT is difficult
1653        (think back references, subroutine calls), and so I have changed the code
1654        so that no minimum is registered for a pattern that contains *ACCEPT.
1656    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1657        it was not handled as intended. [But see 16 below.]
1659    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1660        Sheri Pierce.
1662    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1663        the first byte in a match must be "a".
1665    11. Change 17 for 8.13 increased the recursion depth for patterns like
1666        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1667        pattern contains any instances of (*THEN). If it does not, the old
1668        optimizations are restored. It would be nice to do this on a per-group
1669        basis, but at the moment that is not feasible.
1671    12. In some environments, the output of pcretest -C is CRLF terminated. This
1672        broke RunTest's code that checks for the link size. A single white space
1673        character after the value is now allowed for.
1675    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1676        For "fr", it uses the Windows-specific input and output files.
1678    14. If (*THEN) appeared in a group that was called recursively or as a
1679        subroutine, it did not work as intended. [But see next item.]
1681    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1682        pattern fragments (but not containing any | characters). If A and B are
1683        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1684        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1685        D. In other words, Perl considers parentheses that do not contain any |
1686        characters to be part of a surrounding alternative, whereas PCRE was
1687        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1688        differently. PCRE now behaves in the same way as Perl, except in the case
1689        of subroutine/recursion calls such as (?1) which have in any case always
1690        been different (but PCRE had them first :-).
1692    16. Related to 15 above: Perl does not treat the | in a conditional group as
1693        creating alternatives. Such a group is treated in the same way as an
1694        ordinary group without any | characters when processing (*THEN). PCRE has
1695        been changed to match Perl's behaviour.
1697    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1698        RunGrepTest script failed.
1700    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1701        inevitable for groups that contain captures, but it can lead to a lot of
1702        stack use in large patterns. The old behaviour has been restored for atomic
1703        groups that do not contain any capturing parentheses.
1705    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1706        suppress the check for a minimum subject length at run time. (If it was
1707        given to pcre_exec() or pcre_dfa_exec() it did work.)
1709    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1710        fail to work when decoding hex characters in data strings in EBCDIC
1711        environments.
1713    21. It appears that in at least one Mac OS environment, the isxdigit() function
1714        is implemented as a macro that evaluates to its argument more than once,
1715        contravening the C 90 Standard (I haven't checked a later standard). There
1716        was an instance in pcretest which caused it to go wrong when processing
1717        \x{...} escapes in subject strings. The has been rewritten to avoid using
1718        things like p++ in the argument of isxdigit().
1721    Version 8.13 16-Aug-2011
1722  ------------------------  ------------------------
1724  1.  The Unicode data tables have been updated to Unicode 6.0.0.  1.  The Unicode data tables have been updated to Unicode 6.0.0.
# Line 98  Version 8.13 02-Aug-2011 Line 1815  Version 8.13 02-Aug-2011
1815      tail recursion to cut down on stack usage. Unfortunately, now that there is      tail recursion to cut down on stack usage. Unfortunately, now that there is
1816      the possibility of (*THEN) occurring in these branches, tail recursion is      the possibility of (*THEN) occurring in these branches, tail recursion is
1817      no longer possible because the return has to be checked for (*THEN). These      no longer possible because the return has to be checked for (*THEN). These
1818      two optimizations have therefore been removed.      two optimizations have therefore been removed. [But see 8.20/11 above.]
1820  18. If a pattern containing \R was studied, it was assumed that \R always  18. If a pattern containing \R was studied, it was assumed that \R always
1821      matched two bytes, thus causing the minimum subject length to be      matched two bytes, thus causing the minimum subject length to be
# Line 189  Version 8.13 02-Aug-2011 Line 1906  Version 8.13 02-Aug-2011
1906      For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,      For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1907      unescaped square brackets may also appear as part of class names. For      unescaped square brackets may also appear as part of class names. For
1908      example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves      example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1909      more like Perl.      more like Perl. (But see 8.20/1 above.)
1911  38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this  38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1912      was because it thought it was \N{name}, which is not supported).      was because it thought it was \N{name}, which is not supported).
# Line 224  Version 8.13 02-Aug-2011 Line 1941  Version 8.13 02-Aug-2011
1941      (a) The default value of the buffer size parameter has been increased from      (a) The default value of the buffer size parameter has been increased from
1942          8K to 20K. (The actual buffer used is three times this size.)          8K to 20K. (The actual buffer used is three times this size.)
1944      (b) The default can be changed by ./configure --with-pcregrep-bufsiz when      (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1945          PCRE is built.          PCRE is built.
1947      (c) A --buffer-size=n option has been added to pcregrep, to allow the size      (c) A --buffer-size=n option has been added to pcregrep, to allow the size
# Line 245  Version 8.13 02-Aug-2011 Line 1962  Version 8.13 02-Aug-2011
1962  47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a  47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1963      complete match instead of a partial match. This bug was dependent on both      complete match instead of a partial match. This bug was dependent on both
1964      the PCRE_UTF8 and PCRE_DOTALL options being set.      the PCRE_UTF8 and PCRE_DOTALL options being set.
1966  48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the  48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1967      starting byte set, because \b was not being ignored.      starting byte set, because \b was not being ignored.
1970  Version 8.12 15-Jan-2011  Version 8.12 15-Jan-2011
# Line 900  Version 7.9 11-Apr-09 Line 2617  Version 7.9 11-Apr-09
2617  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
2618      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
2619      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2620      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2621        8.33.]
2623  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
2624      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 1143  Version 7.7 07-May-08 Line 2861  Version 7.7 07-May-08
2861      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2862      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2863      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2864      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2865      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2866      was subsequently set up correctly.)      was subsequently set up correctly.)

Removed from v.657  
changed lines
  Added in v.1689

  ViewVC Help
Powered by ViewVC 1.1.5