ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 740 by zherczeg, Mon Oct 31 06:10:14 2011 UTC revision 1686 by ph10, Thu Feb 23 16:24:08 2017 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4  Version 8.21  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.
34    Version 8.40 11-January-2017
35    ----------------------------
37    1.  Using -o with -M in pcregrep could cause unnecessary repeated output when
38        the match extended over a line boundary.
40    2.  Applied Chris Wilson's second patch (Bugzilla #1681) to CMakeLists.txt for
41        MSVC static compilation, putting the first patch under a new option.
43    3.  Fix register overwite in JIT when SSE2 acceleration is enabled.
45    4.  Ignore "show all captures" (/=) for DFA matching.
47    5.  Fix JIT unaligned accesses on x86. Patch by Marc Mutz.
49    6.  In any wide-character mode (8-bit UTF or any 16-bit or 32-bit mode),
50        without PCRE_UCP set, a negative character type such as \D in a positive
51        class should cause all characters greater than 255 to match, whatever else
52        is in the class. There was a bug that caused this not to happen if a
53        Unicode property item was added to such a class, for example [\D\P{Nd}] or
54        [\W\pL].
56    7.  When pcretest was outputing information from a callout, the caret indicator
57        for the current position in the subject line was incorrect if it was after
58        an escape sequence for a character whose code point was greater than
59        \x{ff}.
61    8.  A pattern such as (?<RA>abc)(?(R)xyz) was incorrectly compiled such that
62        the conditional was interpreted as a reference to capturing group 1 instead
63        of a test for recursion. Any group whose name began with R was
64        misinterpreted in this way. (The reference interpretation should only
65        happen if the group's name is precisely "R".)
67    9.  A number of bugs have been mended relating to match start-up optimizations
68        when the first thing in a pattern is a positive lookahead. These all
69        applied only when PCRE_NO_START_OPTIMIZE was *not* set:
71        (a) A pattern such as (?=.*X)X$ was incorrectly optimized as if it needed
72            both an initial 'X' and a following 'X'.
73        (b) Some patterns starting with an assertion that started with .* were
74            incorrectly optimized as having to match at the start of the subject or
75            after a newline. There are cases where this is not true, for example,
76            (?=.*[A-Z])(?=.{8,16})(?!.*[\s]) matches after the start in lines that
77            start with spaces. Starting .* in an assertion is no longer taken as an
78            indication of matching at the start (or after a newline).
81    Version 8.39 14-June-2016
82    -------------------------
84    1.  If PCRE_AUTO_CALLOUT was set on a pattern that had a (?# comment between
85        an item and its qualifier (for example, A(?#comment)?B) pcre_compile()
86        misbehaved. This bug was found by the LLVM fuzzer.
88    2.  Similar to the above, if an isolated \E was present between an item and its
89        qualifier when PCRE_AUTO_CALLOUT was set, pcre_compile() misbehaved. This
90        bug was found by the LLVM fuzzer.
92    3.  Further to 8.38/46, negated classes such as [^[:^ascii:]\d] were also not
93        working correctly in UCP mode.
95    4.  The POSIX wrapper function regexec() crashed if the option REG_STARTEND
96        was set when the pmatch argument was NULL. It now returns REG_INVARG.
98    5.  Allow for up to 32-bit numbers in the ordin() function in pcregrep.
100    6.  An empty \Q\E sequence between an item and its qualifier caused
101        pcre_compile() to misbehave when auto callouts were enabled. This bug was
102        found by the LLVM fuzzer.
104    7.  If a pattern that was compiled with PCRE_EXTENDED started with white
105        space or a #-type comment that was followed by (?-x), which turns off
106        PCRE_EXTENDED, and there was no subsequent (?x) to turn it on again,
107        pcre_compile() assumed that (?-x) applied to the whole pattern and
108        consequently mis-compiled it. This bug was found by the LLVM fuzzer.
110    8.  A call of pcre_copy_named_substring() for a named substring whose number
111        was greater than the space in the ovector could cause a crash.
113    9.  Yet another buffer overflow bug involved duplicate named groups with a
114        group that reset capture numbers (compare 8.38/7 below). Once again, I have
115        just allowed for more memory, even if not needed. (A proper fix is
116        implemented in PCRE2, but it involves a lot of refactoring.)
118    10. pcre_get_substring_list() crashed if the use of \K in a match caused the
119        start of the match to be earlier than the end.
121    11. Migrating appropriate PCRE2 JIT improvements to PCRE.
123    12. A pattern such as /(?<=((?C)0))/, which has a callout inside a lookbehind
124        assertion, caused pcretest to generate incorrect output, and also to read
125        uninitialized memory (detected by ASAN or valgrind).
127    13. A pattern that included (*ACCEPT) in the middle of a sufficiently deeply
128        nested set of parentheses of sufficient size caused an overflow of the
129        compiling workspace (which was diagnosed, but of course is not desirable).
131    14. And yet another buffer overflow bug involving duplicate named groups, this
132        time nested, with a nested back reference. Yet again, I have just allowed
133        for more memory, because anything more needs all the refactoring that has
134        been done for PCRE2. An example pattern that provoked this bug is:
135        /((?J)(?'R'(?'R'(?'R'(?'R'(?'R'(?|(\k'R'))))))))/ and the bug was
136        registered as CVE-2016-1283.
138    15. pcretest went into a loop if global matching was requested with an ovector
139        size less than 2. It now gives an error message. This bug was found by
140        afl-fuzz.
142    16. An invalid pattern fragment such as (?(?C)0 was not diagnosing an error
143        ("assertion expected") when (?(?C) was not followed by an opening
144        parenthesis.
146    17. Fixed typo ("&&" for "&") in pcre_study(). Fortunately, this could not
147        actually affect anything, by sheer luck.
149    18. Applied Chris Wilson's patch (Bugzilla #1681) to CMakeLists.txt for MSVC
150        static compilation.
152    19. Modified the RunTest script to incorporate a valgrind suppressions file so
153        that certain errors, provoked by the SSE2 instruction set when JIT is used,
154        are ignored.
156    20. A racing condition is fixed in JIT reported by Mozilla.
158    21. Minor code refactor to avoid "array subscript is below array bounds"
159        compiler warning.
161    22. Minor code refactor to avoid "left shift of negative number" warning.
163    23. Fix typo causing compile error when 16- or 32-bit JIT is compiled without
164        UCP support.
166    24. Refactor to avoid compiler warnings in pcrecpp.cc.
168    25. Refactor to fix a typo in pcre_jit_test.c
170    26. Patch to support compiling pcrecpp.cc with Intel compiler.
173    Version 8.38 23-November-2015
174    -----------------------------
176    1.  If a group that contained a recursive back reference also contained a
177        forward reference subroutine call followed by a non-forward-reference
178        subroutine call, for example /.((?2)(?R)\1)()/, pcre_compile() failed to
179        compile correct code, leading to undefined behaviour or an internally
180        detected error. This bug was discovered by the LLVM fuzzer.
182    2.  Quantification of certain items (e.g. atomic back references) could cause
183        incorrect code to be compiled when recursive forward references were
184        involved. For example, in this pattern: /(?1)()((((((\1++))\x85)+)|))/.
185        This bug was discovered by the LLVM fuzzer.
187    3.  A repeated conditional group whose condition was a reference by name caused
188        a buffer overflow if there was more than one group with the given name.
189        This bug was discovered by the LLVM fuzzer.
191    4.  A recursive back reference by name within a group that had the same name as
192        another group caused a buffer overflow. For example:
193        /(?J)(?'d'(?'d'\g{d}))/. This bug was discovered by the LLVM fuzzer.
195    5.  A forward reference by name to a group whose number is the same as the
196        current group, for example in this pattern: /(?|(\k'Pm')|(?'Pm'))/, caused
197        a buffer overflow at compile time. This bug was discovered by the LLVM
198        fuzzer.
200    6.  A lookbehind assertion within a set of mutually recursive subpatterns could
201        provoke a buffer overflow. This bug was discovered by the LLVM fuzzer.
203    7.  Another buffer overflow bug involved duplicate named groups with a
204        reference between their definition, with a group that reset capture
205        numbers, for example: /(?J:(?|(?'R')(\k'R')|((?'R'))))/. This has been
206        fixed by always allowing for more memory, even if not needed. (A proper fix
207        is implemented in PCRE2, but it involves more refactoring.)
209    8.  There was no check for integer overflow in subroutine calls such as (?123).
211    9.  The table entry for \l in EBCDIC environments was incorrect, leading to its
212        being treated as a literal 'l' instead of causing an error.
214    10. There was a buffer overflow if pcre_exec() was called with an ovector of
215        size 1. This bug was found by american fuzzy lop.
217    11. If a non-capturing group containing a conditional group that could match
218        an empty string was repeated, it was not identified as matching an empty
219        string itself. For example: /^(?:(?(1)x|)+)+$()/.
221    12. In an EBCDIC environment, pcretest was mishandling the escape sequences
222        \a and \e in test subject lines.
224    13. In an EBCDIC environment, \a in a pattern was converted to the ASCII
225        instead of the EBCDIC value.
227    14. The handling of \c in an EBCDIC environment has been revised so that it is
228        now compatible with the specification in Perl's perlebcdic page.
230    15. The EBCDIC character 0x41 is a non-breaking space, equivalent to 0xa0 in
231        ASCII/Unicode. This has now been added to the list of characters that are
232        recognized as white space in EBCDIC.
234    16. When PCRE was compiled without UCP support, the use of \p and \P gave an
235        error (correctly) when used outside a class, but did not give an error
236        within a class.
238    17. \h within a class was incorrectly compiled in EBCDIC environments.
240    18. A pattern with an unmatched closing parenthesis that contained a backward
241        assertion which itself contained a forward reference caused buffer
242        overflow. And example pattern is: /(?=di(?<=(?1))|(?=(.))))/.
244    19. JIT should return with error when the compiled pattern requires more stack
245        space than the maximum.
247    20. A possessively repeated conditional group that could match an empty string,
248        for example, /(?(R))*+/, was incorrectly compiled.
250    21. Fix infinite recursion in the JIT compiler when certain patterns such as
251        /(?:|a|){100}x/ are analysed.
253    22. Some patterns with character classes involving [: and \\ were incorrectly
254        compiled and could cause reading from uninitialized memory or an incorrect
255        error diagnosis.
257    23. Pathological patterns containing many nested occurrences of [: caused
258        pcre_compile() to run for a very long time.
260    24. A conditional group with only one branch has an implicit empty alternative
261        branch and must therefore be treated as potentially matching an empty
262        string.
264    25. If (?R was followed by - or + incorrect behaviour happened instead of a
265        diagnostic.
267    26. Arrange to give up on finding the minimum matching length for overly
268        complex patterns.
270    27. Similar to (4) above: in a pattern with duplicated named groups and an
271        occurrence of (?| it is possible for an apparently non-recursive back
272        reference to become recursive if a later named group with the relevant
273        number is encountered. This could lead to a buffer overflow. Wen Guanxing
274        from Venustech ADLAB discovered this bug.
276    28. If pcregrep was given the -q option with -c or -l, or when handling a
277        binary file, it incorrectly wrote output to stdout.
279    29. The JIT compiler did not restore the control verb head in case of *THEN
280        control verbs. This issue was found by Karl Skomski with a custom LLVM
281        fuzzer.
283    30. Error messages for syntax errors following \g and \k were giving inaccurate
284        offsets in the pattern.
286    31. Added a check for integer overflow in conditions (?(<digits>) and
287        (?(R<digits>). This omission was discovered by Karl Skomski with the LLVM
288        fuzzer.
290    32. Handling recursive references such as (?2) when the reference is to a group
291        later in the pattern uses code that is very hacked about and error-prone.
292        It has been re-written for PCRE2. Here in PCRE1, a check has been added to
293        give an internal error if it is obvious that compiling has gone wrong.
295    33. The JIT compiler should not check repeats after a {0,1} repeat byte code.
296        This issue was found by Karl Skomski with a custom LLVM fuzzer.
298    34. The JIT compiler should restore the control chain for empty possessive
299        repeats. This issue was found by Karl Skomski with a custom LLVM fuzzer.
301    35. Match limit check added to JIT recursion. This issue was found by Karl
302        Skomski with a custom LLVM fuzzer.
304    36. Yet another case similar to 27 above has been circumvented by an
305        unconditional allocation of extra memory. This issue is fixed "properly" in
306        PCRE2 by refactoring the way references are handled. Wen Guanxing
307        from Venustech ADLAB discovered this bug.
309    37. Fix two assertion fails in JIT. These issues were found by Karl Skomski
310        with a custom LLVM fuzzer.
312    38. Fixed a corner case of range optimization in JIT.
314    39. An incorrect error "overran compiling workspace" was given if there were
315        exactly enough group forward references such that the last one extended
316        into the workspace safety margin. The next one would have expanded the
317        workspace. The test for overflow was not including the safety margin.
319    40. A match limit issue is fixed in JIT which was found by Karl Skomski
320        with a custom LLVM fuzzer.
322    41. Remove the use of /dev/null in testdata/testinput2, because it doesn't
323        work under Windows. (Why has it taken so long for anyone to notice?)
325    42. In a character class such as [\W\p{Any}] where both a negative-type escape
326        ("not a word character") and a property escape were present, the property
327        escape was being ignored.
329    43. Fix crash caused by very long (*MARK) or (*THEN) names.
331    44. A sequence such as [[:punct:]b] that is, a POSIX character class followed
332        by a single ASCII character in a class item, was incorrectly compiled in
333        UCP mode. The POSIX class got lost, but only if the single character
334        followed it.
336    45. [:punct:] in UCP mode was matching some characters in the range 128-255
337        that should not have been matched.
339    46. If [:^ascii:] or [:^xdigit:] or [:^cntrl:] are present in a non-negated
340        class, all characters with code points greater than 255 are in the class.
341        When a Unicode property was also in the class (if PCRE_UCP is set, escapes
342        such as \w are turned into Unicode properties), wide characters were not
343        correctly handled, and could fail to match.
346    Version 8.37 28-April-2015
347    --------------------------
349    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
350        for those parentheses to be closed with whatever has been captured so far.
351        However, it was failing to mark any other groups between the hightest
352        capture so far and the currrent group as "unset". Thus, the ovector for
353        those groups contained whatever was previously there. An example is the
354        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
356    2.  If an assertion condition was quantified with a minimum of zero (an odd
357        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
359    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
360        unrecognized modifier, a crash could occur.
362    4.  An attempt to do global matching in pcretest with a zero-length ovector
363        caused a crash.
365    5.  Fixed a memory leak during matching that could occur for a subpattern
366        subroutine call (recursive or otherwise) if the number of captured groups
367        that had to be saved was greater than ten.
369    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
370        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
371        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
373    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
374        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
376    8. When a pattern is compiled, it remembers the highest back reference so that
377       when matching, if the ovector is too small, extra memory can be obtained to
378       use instead. A conditional subpattern whose condition is a check on a
379       capture having happened, such as, for example in the pattern
380       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
381       setting the highest backreference number. This mattered only if pcre_exec()
382       was called with an ovector that was too small to hold the capture, and there
383       was no other kind of back reference (a situation which is probably quite
384       rare). The effect of the bug was that the condition was always treated as
385       FALSE when the capture could not be consulted, leading to a incorrect
386       behaviour by pcre_exec(). This bug has been fixed.
388    9. A reference to a duplicated named group (either a back reference or a test
389       for being set in a conditional) that occurred in a part of the pattern where
390       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
391       to be incorrectly calculated, leading to overwriting.
393    10. A mutually recursive set of back references such as (\2)(\1) caused a
394        segfault at study time (while trying to find the minimum matching length).
395        The infinite loop is now broken (with the minimum length unset, that is,
396        zero).
398    11. If an assertion that was used as a condition was quantified with a minimum
399        of zero, matching went wrong. In particular, if the whole group had
400        unlimited repetition and could match an empty string, a segfault was
401        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
402        assertions to be quantified, but not if they are being used as conditions,
403        so the above pattern is faulted by Perl. PCRE has now been changed so that
404        it also rejects such patterns.
406    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
407        failed to allow the zero-repeat case if pcre2_exec() was called with an
408        ovector too small to capture the group.
410    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
411        Red Hat Product Security:
413        (a) A crash if /K and /F were both set with the option to save the compiled
414        pattern.
416        (b) Another crash if the option to print captured substrings in a callout
417        was combined with setting a null ovector, for example \O\C+ as a subject
418        string.
420    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
421        forward reference repeated a large (but limited) number of times within a
422        repeated outer group that has a zero minimum quantifier, caused incorrect
423        code to be compiled, leading to the error "internal error:
424        previously-checked referenced subpattern not found" when an incorrect
425        memory address was read. This bug was reported as "heap overflow",
426        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
427        CVE-2015-2325.
429    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
430        call within a group that also contained a recursive back reference caused
431        incorrect code to be compiled. This bug was reported as "heap overflow",
432        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
433        number CVE-2015-2326.
435    24. Computing the size of the JIT read-only data in advance has been a source
436        of various issues, and new ones are still appear unfortunately. To fix
437        existing and future issues, size computation is eliminated from the code,
438        and replaced by on-demand memory allocation.
440    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
441        adjacent to the end of the range, and the range contained characters with
442        more than one other case, caused incorrect behaviour when compiled in UTF
443        mode. In that example, the range a-j was left out of the class.
445    26. Fix JIT compilation of conditional blocks, which assertion
446        is converted to (*FAIL). E.g: /(?(?!))/.
448    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
449        discovered by the LLVM fuzzer.
451    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
452        when this assertion was used as a condition, for example (?(?!)a|b). In
453        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
454        error about an unsupported item.
456    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
457        possessification code could take exponential time to complete. A recursion
458        depth limit of 1000 has been imposed to limit the resources used by this
459        optimization.
461    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
462        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
463        because \S ensures they are all in the class. The code for doing this was
464        interacting badly with the code for computing the amount of space needed to
465        compile the pattern, leading to a buffer overflow. This bug was discovered
466        by the LLVM fuzzer.
468    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
469        other kinds of group caused stack overflow at compile time. This bug was
470        discovered by the LLVM fuzzer.
472    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
473        between a subroutine call and its quantifier was incorrectly compiled,
474        leading to buffer overflow or other errors. This bug was discovered by the
475        LLVM fuzzer.
477    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
478        assertion after (?(. The code was failing to check the character after
479        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
480        was discovered by the LLVM fuzzer.
482    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
483        a fixed maximum following a group that contains a subroutine reference was
484        incorrectly compiled and could trigger buffer overflow. This bug was
485        discovered by the LLVM fuzzer.
487    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
488        caused a stack overflow instead of the diagnosis of a non-fixed length
489        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
491    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
492        (e.g. /(?<=\Ka)/) could make pcregrep loop.
494    37. There was a similar problem to 36 in pcretest for global matches.
496    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
497        and a subsequent item in the pattern caused a non-match, backtracking over
498        the repeated \X did not stop, but carried on past the start of the subject,
499        causing reference to random memory and/or a segfault. There were also some
500        other cases where backtracking after \C could crash. This set of bugs was
501        discovered by the LLVM fuzzer.
503    39. The function for finding the minimum length of a matching string could take
504        a very long time if mutual recursion was present many times in a pattern,
505        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
506        method has been implemented. This infelicity was discovered by the LLVM
507        fuzzer.
509    40. Static linking against the PCRE library using the pkg-config module was
510        failing on missing pthread symbols.
513    Version 8.36 26-September-2014
514    ------------------------------
516    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
517        -Wmissing-field-initializers and -Wunused-parameter.
519    2.  The tests for quantifiers being too big (greater than 65535) were being
520        applied after reading the number, and stupidly assuming that integer
521        overflow would give a negative number. The tests are now applied as the
522        numbers are read.
524    3.  Tidy code in pcre_exec.c where two branches that used to be different are
525        now the same.
527    4.  The JIT compiler did not generate match limit checks for certain
528        bracketed expressions with quantifiers. This may lead to exponential
529        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
530        issue should be resolved now.
532    5.  Fixed an issue, which occures when nested alternatives are optimized
533        with table jumps.
535    6.  Inserted two casts and changed some ints to size_t in the light of some
536        reported 64-bit compiler warnings (Bugzilla 1477).
538    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
539        an empty string, which sometimes were behaving incorrectly in the
540        interpreter (though correctly in the JIT matcher). This pcretest input is
541        an example:
543          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
546        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
547        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
548        for an empty string was breaking the inner loop and carrying on at a lower
549        level, when possessive repeated groups should always return to a higher
550        level as they have no backtrack points in them. The empty string test now
551        occurs at the outer level.
553    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
554        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
556    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
557        doesn't).
559    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
560        to hold a pattern to be incorrectly computed (too small) when there were
561        named back references to duplicated names. This could cause "internal
562        error: code overflow" or "double free or corruption" or other memory
563        handling errors.
565    11. When named subpatterns had the same prefixes, back references could be
566        confused. For example, in this pattern:
568          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
570        the reference to 'Name' was incorrectly treated as a reference to a
571        duplicate name.
573    12. A pattern such as /^s?c/mi8 where the optional character has more than
574        one "other case" was incorrectly compiled such that it would only try to
575        match starting at "c".
577    13. When a pattern starting with \s was studied, VT was not included in the
578        list of possible starting characters; this should have been part of the
579        8.34/18 patch.
581    14. If a character class started [\Qx]... where x is any character, the class
582        was incorrectly terminated at the ].
584    15. If a pattern that started with a caseless match for a character with more
585        than one "other case" was studied, PCRE did not set up the starting code
586        unit bit map for the list of possible characters. Now it does. This is an
587        optimization improvement, not a bug fix.
589    16. The Unicode data tables have been updated to Unicode 7.0.0.
591    17. Fixed a number of memory leaks in pcregrep.
593    18. Avoid a compiler warning (from some compilers) for a function call with
594        a cast that removes "const" from an lvalue by using an intermediate
595        variable (to which the compiler does not object).
597    19. Incorrect code was compiled if a group that contained an internal recursive
598        back reference was optional (had quantifier with a minimum of zero). This
599        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
600        caused segmentation faults because of stack overflows at compile time.
602    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
603        group that is quantified with an indefinite repeat, caused a compile-time
604        loop which used up all the system stack and provoked a segmentation fault.
605        This was not the same bug as 19 above.
607    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
608        Patch by Mike Frysinger.
611    Version 8.35 04-April-2014
612    --------------------------
614    1.  A new flag is set, when property checks are present in an XCLASS.
615        When this flag is not set, PCRE can perform certain optimizations
616        such as studying these XCLASS-es.
618    2.  The auto-possessification of character sets were improved: a normal
619        and an extended character set can be compared now. Furthermore
620        the JIT compiler optimizes more character set checks.
622    3.  Got rid of some compiler warnings for potentially uninitialized variables
623        that show up only when compiled with -O2.
625    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
626        of a match later then the end of the match. The pcretest program was not
627        handling the case sensibly - it was outputting from the start to the next
628        binary zero. It now reports this situation in a message, and outputs the
629        text from the end to the start.
631    5.  Fast forward search is improved in JIT. Instead of the first three
632        characters, any three characters with fixed position can be searched.
633        Search order: first, last, middle.
635    6.  Improve character range checks in JIT. Characters are read by an inprecise
636        function now, which returns with an unknown value if the character code is
637        above a certain threshold (e.g: 256). The only limitation is that the value
638        must be bigger than the threshold as well. This function is useful when
639        the characters above the threshold are handled in the same way.
641    7.  The macros whose names start with RAWUCHAR are placeholders for a future
642        mode in which only the bottom 21 bits of 32-bit data items are used. To
643        make this more memorable for those maintaining the code, the names have
644        been changed to start with UCHAR21, and an extensive comment has been added
645        to their definition.
647    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
648        to the export list in Makefile.am (they were accidentally omitted from the
649        8.34 tarball).
651    9.  The informational output from pcretest used the phrase "starting byte set"
652        which is inappropriate for the 16-bit and 32-bit libraries. As the output
653        for "first char" and "need char" really means "non-UTF-char", I've changed
654        "byte" to "char", and slightly reworded the output. The documentation about
655        these values has also been (I hope) clarified.
657    10. Another JIT related optimization: use table jumps for selecting the correct
658        backtracking path, when more than four alternatives are present inside a
659        bracket.
661    11. Empty match is not possible, when the minimum length is greater than zero,
662        and there is no \K in the pattern. JIT should avoid empty match checks in
663        such cases.
665    12. In a caseless character class with UCP support, when a character with more
666        than one alternative case was not the first character of a range, not all
667        the alternative cases were added to the class. For example, s and \x{17f}
668        are both alternative cases for S: the class [RST] was handled correctly,
669        but [R-T] was not.
671    13. The configure.ac file always checked for pthread support when JIT was
672        enabled. This is not used in Windows, so I have put this test inside a
673        check for the presence of windows.h (which was already tested for).
675    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
676        The algorithm provides a way to skip certain starting offsets, and usually
677        faster than linear prefix searches.
679    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
680        as for 'fr_FR' and 'french'. For some reason, however, it then used the
681        Windows-specific input and output files, which have 'french' screwed in.
682        So this could never have worked. One of the problems with locales is that
683        they aren't always the same. I have now updated RunTest so that it checks
684        the output of the locale test (test 3) against three different output
685        files, and it allows the test to pass if any one of them matches. With luck
686        this should make the test pass on some versions of Solaris where it was
687        failing. Because of the uncertainty, the script did not used to stop if
688        test 3 failed; it now does. If further versions of a French locale ever
689        come to light, they can now easily be added.
691    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
692        there was a message during ./configure, but it did not stop. This now
693        provokes an error. The invalid example in README has been corrected.
694        If a value less than the minimum is given, the minimum value has always
695        been used, but now a warning is given.
697    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
698        was a bug in the test system, which is now fixed. Also, the list of various
699        configurations that are tested for each release did not have one with both
700        16/32 bits and --enable-bar-anycrlf. It now does.
702    18. pcretest was missing "-C bsr" for displaying the \R default setting.
704    19. Little endian PowerPC systems are supported now by the JIT compiler.
706    20. The fast forward newline mechanism could enter to an infinite loop on
707        certain invalid UTF-8 input. Although we don't support these cases
708        this issue can be fixed by a performance optimization.
710    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
711        not take account if existing stack usage. There is now a new global
712        variable called pcre_stack_guard that can be set to point to an external
713        function to check stack availability. It is called at the start of
714        processing every parenthesized group.
716    22. A typo in the code meant that in ungreedy mode the max/min qualifier
717        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
718        not match "ab".
720    23. When UTF was disabled, the JIT program reported some incorrect compile
721        errors. These messages are silenced now.
723    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
724        compiler.
726    25. Change all the temporary files used in RunGrepTest to be different to those
727        used by RunTest so that the tests can be run simultaneously, for example by
728        "make -j check".
731    Version 8.34 15-December-2013
732    -----------------------------
734    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
735        executable memory. Patch inspired by Carsten Klein.
737    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
738        this macro is never tested and has no effect, because the work to support
739        coverage involves only compiling and linking options and special targets in
740        the Makefile. The comment in config.h implied that defining the macro would
741        enable coverage support, which is totally false. There was also support for
742        setting this macro in the CMake files (my fault, I just copied it from
743        configure). SUPPORT_GCOV has now been removed.
745    3.  Make a small performance improvement in strlen16() and strlen32() in
746        pcretest.
748    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
749        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
750        these cases). There was also one in pcretest.c.
752    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
754    6.  In UTF mode, the code for checking whether a group could match an empty
755        string (which is used for indefinitely repeated groups to allow for
756        breaking an infinite loop) was broken when the group contained a repeated
757        negated single-character class with a character that occupied more than one
758        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
759        UTF-8 mode). The effect was undefined: the group might or might not be
760        deemed as matching an empty string, or the program might have crashed.
762    7.  The code for checking whether a group could match an empty string was not
763        recognizing that \h, \H, \v, \V, and \R must match a character.
765    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
766        an empty string. If it can, pcretest shows this in its information output.
768    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
769        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
770        matched by pcre_exec() without using JIT:
772        (a) If the rest of the pattern did not match after a maximal run of
773            grapheme clusters, the code for backing up to try with fewer of them
774            did not always back up over a full grapheme when characters that do not
775            have the modifier quality were involved, e.g. Hangul syllables.
777        (b) If the match point in a subject started with modifier character, and
778            there was no match, the code could incorrectly back up beyond the match
779            point, and potentially beyond the first character in the subject,
780            leading to a segfault or an incorrect match result.
782    10. A conditional group with an assertion condition could lead to PCRE
783        recording an incorrect first data item for a match if no other first data
784        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
785        first data item, and therefore matched "ca" after "c" instead of at the
786        start.
788    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
789        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
791    12. The source of pcregrep now includes z/OS-specific code so that it can be
792        compiled for z/OS as part of the special z/OS distribution.
794    13. Added the -T and -TM options to pcretest.
796    14. The code in pcre_compile.c for creating the table of named capturing groups
797        has been refactored. Instead of creating the table dynamically during the
798        actual compiling pass, the information is remembered during the pre-compile
799        pass (on the stack unless there are more than 20 named groups, in which
800        case malloc() is used) and the whole table is created before the actual
801        compile happens. This has simplified the code (it is now nearly 150 lines
802        shorter) and prepared the way for better handling of references to groups
803        with duplicate names.
805    15. A back reference to a named subpattern when there is more than one of the
806        same name now checks them in the order in which they appear in the pattern.
807        The first one that is set is used for the reference. Previously only the
808        first one was inspected. This change makes PCRE more compatible with Perl.
810    16. Unicode character properties were updated from Unicode 6.3.0.
812    17. The compile-time code for auto-possessification has been refactored, based
813        on a patch by Zoltan Herczeg. It now happens after instead of during
814        compilation. The code is cleaner, and more cases are handled. The option
815        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
816        options in pcretest are provided to set it. It can also be set by
817        (*NO_AUTO_POSSESS) at the start of a pattern.
819    18. The character VT has been added to the default ("C" locale) set of
820        characters that match \s and are generally treated as white space,
821        following this same change in Perl 5.18. There is now no difference between
822        "Perl space" and "POSIX space". Whether VT is treated as white space in
823        other locales depends on the locale.
825    19. The code for checking named groups as conditions, either for being set or
826        for being recursed, has been refactored (this is related to 14 and 15
827        above). Processing unduplicated named groups should now be as fast at
828        numerical groups, and processing duplicated groups should be faster than
829        before.
831    20. Two patches to the CMake build system, by Alexander Barkov:
833          (1) Replace the "source" command by "." in CMakeLists.txt because
834              "source" is a bash-ism.
836          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
837              without these the CMake build does not work on Solaris.
839    21. Perl has changed its handling of \8 and \9. If there is no previously
840        encountered capturing group of those numbers, they are treated as the
841        literal characters 8 and 9 instead of a binary zero followed by the
842        literals. PCRE now does the same.
844    22. Following Perl, added \o{} to specify codepoints in octal, making it
845        possible to specify values greater than 0777 and also making them
846        unambiguous.
848    23. Perl now gives an error for missing closing braces after \x{... instead of
849        treating the string as literal. PCRE now does the same.
851    24. RunTest used to grumble if an inappropriate test was selected explicitly,
852        but just skip it when running all tests. This make it awkward to run ranges
853        of tests when one of them was inappropriate. Now it just skips any
854        inappropriate tests, as it always did when running all tests.
856    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
857        character types such as \d or \w, too many callouts were inserted, and the
858        data that they returned was rubbish.
860    26. In UCP mode, \s was not matching two of the characters that Perl matches,
861        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
862        were matched by \h. The code has now been refactored so that the lists of
863        the horizontal and vertical whitespace characters used for \h and \v (which
864        are defined only in one place) are now also used for \s.
866    27. Add JIT support for the 64 bit TileGX architecture.
867        Patch by Jiong Wang (Tilera Corporation).
869    28. Possessive quantifiers for classes (both explicit and automatically
870        generated) now use special opcodes instead of wrapping in ONCE brackets.
872    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
873        (because it's meaningless), this was not happening when PCRE_CASELESS was
874        set. Not wrong, but inefficient.
876    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
877        properties for \w, \d, etc) is present in a test regex. Otherwise if the
878        test contains no characters greater than 255, Perl doesn't realise it
879        should be using Unicode semantics.
881    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
882        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
883        does in Unicode mode.
885    32. Added the "forbid" facility to pcretest so that putting tests into the
886        wrong test files can sometimes be quickly detected.
888    33. There is now a limit (default 250) on the depth of nesting of parentheses.
889        This limit is imposed to control the amount of system stack used at compile
890        time. It can be changed at build time by --with-parens-nest-limit=xxx or
891        the equivalent in CMake.
893    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
894        errors. Perl warns for these when in warning mode, but PCRE has no facility
895        for giving warnings.
897    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
898        However, this was not working for (?!) because it is optimized to (*FAIL),
899        for which PCRE does not allow quantifiers. The optimization is now disabled
900        when a quantifier follows (?!). I can't see any use for this, but it makes
901        things uniform.
903    36. Perl no longer allows group names to start with digits, so I have made this
904        change also in PCRE. It simplifies the code a bit.
906    37. In extended mode, Perl ignores spaces before a + that indicates a
907        possessive quantifier. PCRE allowed a space before the quantifier, but not
908        before the possessive +. It now does.
910    38. The use of \K (reset reported match start) within a repeated possessive
911        group such as (a\Kb)*+ was not working.
913    40. Document that the same character tables must be used at compile time and
914        run time, and that the facility to pass tables to pcre_exec() and
915        pcre_dfa_exec() is for use only with saved/restored patterns.
917    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
918        features for Builds with MSVC:
920        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
921           builds). The .rc files can be used to set FileDescription and many other
922           attributes.
924        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
925           This allows higher-level build scripts which want .pdb files to avoid
926           hard-coding the exact files needed."
928    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
929        mean "start of word" and "end of word", respectively, as a transition aid.
931    43. A minimizing repeat of a class containing codepoints greater than 255 in
932        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
933        compiled to use the heap for recursion.
935    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
936        is configured.
939    Version 8.33 28-May-2013
940    ------------------------
942    1.  Added 'U' to some constants that are compared to unsigned integers, to
943        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
944        variables that are added to signed variables, to ensure the result is
945        signed and can be negated.
947    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
948        CMake config files.
950    3.  Revise the creation of config.h.generic so that all boolean macros are
951        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
952        overriding via -D on the command line possible.
954    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
955        to unsigned int is reported to make a quite noticeable speed difference in
956        a specific Windows environment. Testing on Linux did also appear to show
957        some benefit (and it is clearly not harmful). Also fixed the definition of
958        Xop which should be unsigned.
960    5.  Related to (4), changing the definition of the intermediate variable cc
961        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
962        improvements.
964    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
965        unnecessary spaces.
967    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
968        and later.
970    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
972    9.  Optimizing fast_forward_start_bits in JIT.
974    10. Adding support for callouts in JIT, and fixing some issues revealed
975        during this work. Namely:
977        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
979        (b) Minimum length was not checked before the matching is started.
981    11. The value of capture_last that is passed to callouts was incorrect in some
982        cases when there was a capture on one path that was subsequently abandoned
983        after a backtrack. Also, the capture_last value is now reset after a
984        recursion, since all captures are also reset in this case.
986    12. The interpreter no longer returns the "too many substrings" error in the
987        case when an overflowing capture is in a branch that is subsequently
988        abandoned after a backtrack.
990    13. In the pathological case when an offset vector of size 2 is used, pcretest
991        now prints out the matched string after a yield of 0 or 1.
993    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
994        Only supported by the JIT compiler at the moment.
996    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
998    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
999        offset of the starting point of the matching process, provided the offsets
1000        vector is large enough.
1002    17. The \A escape now records a lookbehind value of 1, though its execution
1003        does not actually inspect the previous character. This is to ensure that,
1004        in partial multi-segment matching, at least one character from the old
1005        segment is retained when a new segment is processed. Otherwise, if there
1006        are no lookbehinds in the pattern, \A might match incorrectly at the start
1007        of a new segment.
1009    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
1011    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
1012        gives some modest performance improvement in 8-bit mode.
1014    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
1015        be expressed in certain programming languages using Universal Character
1016        Names.
1018    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
1019        which points out that "non characters" are not "characters that may not
1020        appear in Unicode strings" but rather "characters that are reserved for
1021        internal use and have only local meaning".
1023    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
1024        there was a conditional group that depended on an assertion, if the
1025        assertion was false, the callout that immediately followed the alternation
1026        in the condition was skipped when pcre_exec() was used for matching.
1028    23. Allow an explicit callout to be inserted before an assertion that is the
1029        condition for a conditional group, for compatibility with automatic
1030        callouts, which always insert a callout at this point.
1032    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
1033        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
1035    25. (*PRUNE) is now supported by the JIT compiler.
1037    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
1039    27. Fix the case where there are two or more SKIPs with arguments that may be
1040        ignored.
1042    28. (*SKIP) is now supported by the JIT compiler.
1044    29. (*THEN) is now supported by the JIT compiler.
1046    30. Update RunTest with additional test selector options.
1048    31. The way PCRE handles backtracking verbs has been changed in two ways.
1050        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
1051        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
1052        backtracking. In some cases this makes it more Perl-compatible, but Perl's
1053        rather obscure rules do not always do the same thing.
1055        (2) Previously, backtracking verbs were confined within assertions. This is
1056        no longer the case for positive assertions, except for (*ACCEPT). Again,
1057        this sometimes improves Perl compatibility, and sometimes does not.
1059    32. A number of tests that were in test 2 because Perl did things differently
1060        have been moved to test 1, because either Perl or PCRE has changed, and
1061        these tests are now compatible.
1063    32. Backtracking control verbs are now handled in the same way in JIT and
1064        interpreter.
1066    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
1067        contained a forward subroutine reference caused a compile error.
1069    34. Auto-detect and optimize limited repetitions in JIT.
1071    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
1072        blocking (*UTF) etc.
1074    36. In the interpreter, maximizing pattern repetitions for characters and
1075        character types now use tail recursion, which reduces stack usage.
1077    37. The value of the max lookbehind was not correctly preserved if a compiled
1078        and saved regex was reloaded on a host of different endianness.
1080    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
1081        of the compiled pattern block, expand the flags field from 16 to 32 bits
1082        because it was almost full.
1084    39. Try madvise first before posix_madvise.
1086    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
1087        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
1088        original change locked out all matches of empty strings. This has been
1089        changed so that one match of an empty string per line is recognized.
1090        Subsequent searches on the same line (for colouring or for --only-matching,
1091        for example) do not recognize empty strings.
1093    41. Applied a user patch to fix a number of spelling mistakes in comments.
1095    42. Data lines longer than 65536 caused pcretest to crash.
1097    43. Clarified the data type for length and startoffset arguments for pcre_exec
1098        and pcre_dfa_exec in the function-specific man pages, where they were
1099        explicitly stated to be in bytes, never having been updated. I also added
1100        some clarification to the pcreapi man page.
1102    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
1103        a segmentation fault.
1106    Version 8.32 30-November-2012
1107    -----------------------------
1109    1.  Improved JIT compiler optimizations for first character search and single
1110        character iterators.
1112    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
1113        Patch by Daniel Richard G.
1115    3.  Single character iterator optimizations in the JIT compiler.
1117    4.  Improved JIT compiler optimizations for character ranges.
1119    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
1120        Reported by Giuseppe D'Angelo.
1122    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
1123        of a line, was being set incorrectly in cases where .* appeared inside
1124        atomic brackets at the start of a pattern, or where there was a subsequent
1125        *PRUNE or *SKIP.
1127    7.  Improved instruction cache flush for POWER/PowerPC.
1128        Patch by Daniel Richard G.
1130    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
1131        grep:
1133        (a) There is now no limit to the number of patterns to be matched.
1135        (b) An error is given if a pattern is too long.
1137        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
1138            are now supported.
1140        (d) --exclude-from and --include-from (multiple use) have been added.
1142        (e) Exclusions and inclusions now apply to all files and directories, not
1143            just to those obtained from scanning a directory recursively.
1145        (f) Multiple uses of -f and --file-list are now supported.
1147        (g) In a Windows environment, the default for -d has been changed from
1148            "read" (the GNU grep default) to "skip", because otherwise the presence
1149            of a directory in the file list provokes an error.
1151        (h) The documentation has been revised and clarified in places.
1153    9.  Improve the matching speed of capturing brackets.
1155    10. Changed the meaning of \X so that it now matches a Unicode extended
1156        grapheme cluster.
1158    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
1159        out POSIX threads when JIT support is configured.
1161    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
1163    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
1164        structure could go wrong in environments where size_t is not the same size
1165        as int.
1167    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
1168        set.
1170    15. The EBCDIC support had decayed; later updates to the code had included
1171        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
1172        general tidy up of EBCDIC-related issues, and the documentation was also
1173        not quite right. There is now a test that can be run on ASCII systems to
1174        check some of the EBCDIC-related things (but is it not a full test).
1176    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
1177        in a small tidy to the code.
1179    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
1181    18. If the --only-matching (-o) option in pcregrep is specified multiple
1182        times, each one causes appropriate output. For example, -o1 -o2 outputs the
1183        substrings matched by the 1st and 2nd capturing parentheses. A separating
1184        string can be specified by --om-separator (default empty).
1186    19. Improving the first n character searches.
1188    20. Turn case lists for horizontal and vertical white space into macros so that
1189        they are defined only once.
1191    21. This set of changes together give more compatible Unicode case-folding
1192        behaviour for characters that have more than one other case when UCP
1193        support is available.
1195        (a) The Unicode property table now has offsets into a new table of sets of
1196            three or more characters that are case-equivalent. The MultiStage2.py
1197            script that generates these tables (the pcre_ucd.c file) now scans
1198            CaseFolding.txt instead of UnicodeData.txt for character case
1199            information.
1201        (b) The code for adding characters or ranges of characters to a character
1202            class has been abstracted into a generalized function that also handles
1203            case-independence. In UTF-mode with UCP support, this uses the new data
1204            to handle characters with more than one other case.
1206        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
1207            whose other case is greater than 256 are now correctly matched
1208            caselessly. Previously, the high codepoint matched the low one, but not
1209            vice versa.
1211        (d) The processing of \h, \H, \v, and \ in character classes now makes use
1212            of the new class addition function, using character lists defined as
1213            macros alongside the case definitions of 20 above.
1215        (e) Caseless back references now work with characters that have more than
1216            one other case.
1218        (f) General caseless matching of characters with more than one other case
1219            is supported.
1221    22. Unicode character properties were updated from Unicode 6.2.0
1223    23. Improved CMake support under Windows. Patch by Daniel Richard G.
1225    24. Add support for 32-bit character strings, and UTF-32
1227    25. Major JIT compiler update (code refactoring and bugfixing).
1228        Experimental Sparc 32 support is added.
1230    26. Applied a modified version of Daniel Richard G's patch to create
1231        pcre.h.generic and config.h.generic by "make" instead of in the
1232        PrepareRelease script.
1234    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
1235        pcre_compile.c when checking for a zero character.
1237    28. Introducing a native interface for JIT. Through this interface, the compiled
1238        machine code can be directly executed. The purpose of this interface is to
1239        provide fast pattern matching, so several sanity checks are not performed.
1240        However, feature tests are still performed. The new interface provides
1241        1.4x speedup compared to the old one.
1243    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
1244        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
1245        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
1247    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
1248        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
1249        the "old" RFC 2279). Instead, it ended up passing a negative length to
1250        pcre_exec().
1252    31. Add support for GCC's visibility feature to hide internal functions.
1254    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
1255        "unknown -C option" after outputting 0 or 1.
1257    33. There is now support for generating a code coverage report for the test
1258        suite in environments where gcc is the compiler and lcov is installed. This
1259        is mainly for the benefit of the developers.
1261    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
1262        unaddressable using valgrind annotations, allowing valgrind to detect
1263        invalid memory accesses. This is mainly for the benefit of the developers.
1265    25. (*UTF) can now be used to start a pattern in any of the three libraries.
1267    26. Give configure error if --enable-cpp but no C++ compiler found.
1270    Version 8.31 06-July-2012
1271    -------------------------
1273    1.  Fixing a wrong JIT test case and some compiler warnings.
1275    2.  Removed a bashism from the RunTest script.
1277    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
1278        to unsigned type, result still unsigned" that was given by an MS compiler
1279        on encountering the code "-sizeof(xxx)".
1281    4.  Partial matching support is added to the JIT compiler.
1283    5.  Fixed several bugs concerned with partial matching of items that consist
1284        of more than one character:
1286        (a) /^(..)\1/ did not partially match "aba" because checking references was
1287            done on an "all or nothing" basis. This also applied to repeated
1288            references.
1290        (b) \R did not give a hard partial match if \r was found at the end of the
1291            subject.
1293        (c) \X did not give a hard partial match after matching one or more
1294            characters at the end of the subject.
1296        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1297            a partial match for the string "\r".
1299        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1300            a partial match for a CR character at the end of the subject string.
1302    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1303        running pcretest, the text "(JIT)" added to the output whenever JIT is
1304        actually used to run the match.
1306    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1307        or /S+[+] with a digit between 1 and 7.
1309    8.  OP_NOT now supports any UTF character not just single-byte ones.
1311    9.  (*MARK) control verb is now supported by the JIT compiler.
1313    10. The command "./RunTest list" lists the available tests without actually
1314        running any of them. (Because I keep forgetting what they all are.)
1318    12. Applied a (slightly modified) user-supplied patch that improves performance
1319        when the heap is used for recursion (compiled with --disable-stack-for-
1320        recursion). Instead of malloc and free for each heap frame each time a
1321        logical recursion happens, frames are retained on a chain and re-used where
1322        possible. This sometimes gives as much as 30% improvement.
1324    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1325        call.
1327    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1329    15. It is now possible to link pcretest with libedit as an alternative to
1330        libreadline.
1332    16. (*COMMIT) control verb is now supported by the JIT compiler.
1334    17. The Unicode data tables have been updated to Unicode 6.1.0.
1336    18. Added --file-list option to pcregrep.
1338    19. Added binary file support to pcregrep, including the -a, --binary-files,
1339        -I, and --text options.
1341    20. The madvise function is renamed for posix_madvise for QNX compatibility
1342        reasons. Fixed by Giuseppe D'Angelo.
1344    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1346    22. Changed the option for creating symbolic links for 16-bit man pages from
1347        -s to -sf so that re-installing does not cause issues.
1349    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1351    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1352        matches in certain environments (the workspace was not being correctly
1353        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1354        some of the workspace data at the beginning of a restart.
1356    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1357        was not doing so when it should - probably a typo introduced by SVN 528
1358        (change 8.10/14).
1360    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1361        \w+ when the character tables indicated that \x{c4} was a word character.
1362        There were several related cases, all because the tests for doing a table
1363        lookup were testing for characters less than 127 instead of 255.
1365    27. If a pattern contains capturing parentheses that are not used in a match,
1366        their slots in the ovector are set to -1. For those that are higher than
1367        any matched groups, this happens at the end of processing. In the case when
1368        there were back references that the ovector was too small to contain
1369        (causing temporary malloc'd memory to be used during matching), and the
1370        highest capturing number was not used, memory off the end of the ovector
1371        was incorrectly being set to -1. (It was using the size of the temporary
1372        memory instead of the true size.)
1374    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1375        ovector size, it uses memory at the end of the block that it has got.
1377    29. Check for an overlong MARK name and give an error at compile time. The
1378        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1380    30. JIT compiler update.
1382    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1383        Rill for the patch.
1385    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1387    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1389    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1390        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1391        was enabled.
1393    35. Improve JIT code generation for greedy plus quantifier.
1395    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1396        1 to "aa" instead of to an empty string. The bug affected repeated groups
1397        that could potentially match an empty string.
1399    37. Optimizing single character iterators in JIT.
1401    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1402        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1403        codepoints that are too big for the mode are faulted, and in a UTF mode,
1404        disallowed codepoints are also faulted.
1406    39. If PCRE was compiled with UTF support, in three places in the DFA
1407        matcher there was code that should only have been obeyed in UTF mode, but
1408        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1409        processing when bytes with values greater than 127 were present. In 16-bit
1410        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1411        both cases the values are those that cannot be the first data item in a UTF
1412        character. The three items that might have provoked this were recursions,
1413        possessively repeated groups, and atomic groups.
1415    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1416        and pcregrep, because some OS require shared objects to be explicitly
1417        passed to ld, causing the link step to fail if they are not.
1419    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1420        mode, patterns that started with \h* or \R* might be incorrectly matched.
1423    Version 8.30 04-February-2012
1424    -----------------------------
1426    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1427        name is defined in ctype.h.
1429    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1430        only in quite long subpatterns.
1432    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1433        since it was replaced by pcre_fullinfo() in February 2000.
1435    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1436        match a (*MARK), and the match failed at the start of the subject, a
1437        reference to memory before the start of the subject could occur. This bug
1438        was introduced by fix 17 of release 8.21.
1440    5.  A reference to an unset group with zero minimum repetition was giving
1441        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1442        /(another)?(\1?)test/ matched against "hello world test". This bug was
1443        introduced in release 8.13.
1445    6.  Add support for 16-bit character strings (a large amount of work involving
1446        many changes and refactorings).
1448    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1449        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1450        from a file.
1452    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1453        rounding is not applied in this particular case).
1455    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1456        if they appear, or are escaped, in patterns.
1458    10. Get rid of a number of -Wunused-but-set-variable warnings.
1460    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1461        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1462        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1463        also returns the mark "x". This bug applied to capturing parentheses,
1464        non-capturing parentheses, and atomic parentheses. It also applied to some
1465        assertions.
1467    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1468        information out of configure.ac instead of relying on pcre.h.generic, which
1469        is not stored in the repository.
1471    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1472        -lreadline.
1474    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1476    15. Applied Graycode's patch to put the top-level frame on the stack rather
1477        than the heap when not using the stack for recursion. This gives a
1478        performance improvement in many cases when recursion is not deep.
1480    16. Experimental code added to "pcretest -C" to output the stack frame size.
1483    Version 8.21 12-Dec-2011
1484    ------------------------
1486  1.  Updating the JIT compiler.  1.  Updating the JIT compiler.
1488    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1489        are added as well.
1491    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1492        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1493        calling _pcre_jit_exec. Some extra comments are added.
1495    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1496        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1497        was introduced by change 18 for 8.20.
1499    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1500        ECMA-262 standard.
1502    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1503        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1504        This bug was probably introduced by change 9 of 8.13.
1506    7.  While fixing 6 above, I noticed that a number of other items were being
1507        incorrectly rejected as "not fixed length". This arose partly because newer
1508        opcodes had not been added to the fixed-length checking code. I have (a)
1509        corrected the bug and added tests for these items, and (b) arranged for an
1510        error to occur if an unknown opcode is encountered while checking for fixed
1511        length instead of just assuming "not fixed length". The items that were
1512        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1513        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1514        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1516    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1517        being incorrectly compiled and would have given unpredicatble results.
1519    9.  A possessively repeated subpattern with minimum repeat count greater than
1520        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1521        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1522        the first (A) could occur when it should not.
1524    10. Add a cast and remove a redundant test from the code.
1526    11. JIT should use pcre_malloc/pcre_free for allocation.
1528    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1529        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1530        is anything other than /usr, -L is still shown).
1532    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1534    14. Perl does not support \N without a following name in a [] class; PCRE now
1535        also gives an error.
1537    15. If a forward reference was repeated with an upper limit of around 2000,
1538        it caused the error "internal error: overran compiling workspace". The
1539        maximum number of forward references (including repeats) was limited by the
1540        internal workspace, and dependent on the LINK_SIZE. The code has been
1541        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1542        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1543        of around 200,000 forward references. While doing this, I also speeded up
1544        the filling in of repeated forward references.
1546    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1547        incorrectly expecting the subject to contain another "a" after the start.
1549    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1550        in the match, the SKIP should be ignored. This was not happening; instead
1551        the SKIP was being treated as NOMATCH. For patterns such as
1552        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1553        tested.
1555    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1556        now much more compatible with Perl, in particular in cases where the result
1557        is a non-match for a non-anchored pattern. For example, if
1558        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1559        "m", where previously it did not return a name. A side effect of this
1560        change is that for partial matches, the last encountered mark name is
1561        returned, as for non matches. A number of tests that were previously not
1562        Perl-compatible have been moved into the Perl-compatible test files. The
1563        refactoring has had the pleasing side effect of removing one argument from
1564        the match() function, thus reducing its stack requirements.
1566    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1567        subsequent uses of /S (without +) incorrectly behaved like /S+.
1569    21. Retrieve executable code size support for the JIT compiler and fixing
1570        some warnings.
1572    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1573        not work when the shorter character appeared right at the end of the
1574        subject string.
1576    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1577        systems.
1579    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1580        output it when the /M option is used in pcretest.
1582    25. The CheckMan script was not being included in the distribution. Also, added
1583        an explicit "perl" to run Perl scripts from the PrepareRelease script
1584        because this is reportedly needed in Windows.
1586    26. If study data was being save in a file and studying had not found a set of
1587        "starts with" bytes for the pattern, the data written to the file (though
1588        never used) was taken from uninitialized memory and so caused valgrind to
1589        complain.
1591    27. Updated RunTest.bat as provided by Sheri Pierce.
1593    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1595    29. Computation of memory usage for the table of capturing group names was
1596        giving an unnecessarily large value.
1599  Version 8.20 21-Oct-2011  Version 8.20 21-Oct-2011
1600  ------------------------  ------------------------
# Line 1021  Version 7.9 11-Apr-09 Line 2610  Version 7.9 11-Apr-09
2610  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
2611      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
2612      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2613      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2614        8.33.]
2616  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
2617      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 1264  Version 7.7 07-May-08 Line 2854  Version 7.7 07-May-08
2854      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2855      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2856      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2857      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2858      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2859      was subsequently set up correctly.)      was subsequently set up correctly.)

Removed from v.740  
changed lines
  Added in v.1686

  ViewVC Help
Powered by ViewVC 1.1.5