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

Diff of /code/trunk/ChangeLog

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

revision 753 by ph10, Sat Nov 19 17:03:35 2011 UTC revision 1582 by ph10, Wed Jul 22 09:33:04 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
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.
6    
7    Version 8.38 xx-xxx-xxxx
8    ------------------------
9    
10    1.  If a group that contained a recursive back reference also contained a
11        forward reference subroutine call followed by a non-forward-reference
12        subroutine call, for example /.((?2)(?R)\1)()/, pcre2_compile() failed to
13        compile correct code, leading to undefined behaviour or an internally
14        detected error. This bug was discovered by the LLVM fuzzer.
15    
16    2.  Quantification of certain items (e.g. atomic back references) could cause
17        incorrect code to be compiled when recursive forward references were
18        involved. For example, in this pattern: /(?1)()((((((\1++))\x85)+)|))/.
19        This bug was discovered by the LLVM fuzzer.
20    
21    3.  A repeated conditional group whose condition was a reference by name caused
22        a buffer overflow if there was more than one group with the given name.
23        This bug was discovered by the LLVM fuzzer.
24    
25    4.  A recursive back reference by name within a group that had the same name as
26        another group caused a buffer overflow. For example:
27        /(?J)(?'d'(?'d'\g{d}))/. This bug was discovered by the LLVM fuzzer.
28    
29    5.  A forward reference by name to a group whose number is the same as the
30        current group, for example in this pattern: /(?|(\k'Pm')|(?'Pm'))/, caused
31        a buffer overflow at compile time. This bug was discovered by the LLVM
32        fuzzer.
33    
34    6.  A lookbehind assertion within a set of mutually recursive subpatterns could
35        provoke a buffer overflow. This bug was discovered by the LLVM fuzzer.
36    
37    7.  Another buffer overflow bug involved duplicate named groups with a
38        reference between their definition, with a group that reset capture
39        numbers, for example: /(?J:(?|(?'R')(\k'R')|((?'R'))))/. This has been
40        fixed by always allowing for more memory, even if not needed. (A proper fix
41        is implemented in PCRE2, but it involves more refactoring.)
42    
43    8.  There was no check for integer overflow in subroutine calls such as (?123).
44    
45    9.  The table entry for \l in EBCDIC environments was incorrect, leading to its
46        being treated as a literal 'l' instead of causing an error.
47    
48    10. There was a buffer overflow if pcre_exec() was called with an ovector of
49        size 1. This bug was found by american fuzzy lop.
50    
51    11. If a non-capturing group containing a conditional group that could match
52        an empty string was repeated, it was not identified as matching an empty
53        string itself. For example: /^(?:(?(1)x|)+)+$()/.
54    
55    12. In an EBCDIC environment, pcretest was mishandling the escape sequences
56        \a and \e in test subject lines.
57    
58    13. In an EBCDIC environment, \a in a pattern was converted to the ASCII
59        instead of the EBCDIC value.
60    
61    14. The handling of \c in an EBCDIC environment has been revised so that it is
62        now compatible with the specification in Perl's perlebcdic page.
63    
64    15. The EBCDIC character 0x41 is a non-breaking space, equivalent to 0xa0 in
65        ASCII/Unicode. This has now been added to the list of characters that are
66        recognized as white space in EBCDIC.
67    
68    16. When PCRE was compiled without UCP support, the use of \p and \P gave an
69        error (correctly) when used outside a class, but did not give an error
70        within a class.
71    
72    17. \h within a class was incorrectly compiled in EBCDIC environments.
73    
74    18. A pattern with an unmatched closing parenthesis that contained a backward
75        assertion which itself contained a forward reference caused buffer
76        overflow. And example pattern is: /(?=di(?<=(?1))|(?=(.))))/.
77    
78    19. JIT should return with error when the compiled pattern requires more stack
79        space than the maximum.
80    
81    20. A possessively repeated conditional group that could match an empty string,
82        for example, /(?(R))*+/, was incorrectly compiled.
83    
84    21. Fix infinite recursion in the JIT compiler when certain patterns such as
85        /(?:|a|){100}x/ are analysed.
86    
87    22. Some patterns with character classes involving [: and \\ were incorrectly
88        compiled and could cause reading from uninitialized memory or an incorrect
89        error diagnosis.
90    
91    23. Pathological patterns containing many nested occurrences of [: caused
92        pcre_compile() to run for a very long time.
93    
94    24. A conditional group with only one branch has an implicit empty alternative
95        branch and must therefore be treated as potentially matching an empty
96        string.
97    
98    25. If (?R was followed by - or + incorrect behaviour happened instead of a
99        diagnostic.
100    
101    
102    Version 8.37 28-April-2015
103    --------------------------
104    
105    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
106        for those parentheses to be closed with whatever has been captured so far.
107        However, it was failing to mark any other groups between the hightest
108        capture so far and the currrent group as "unset". Thus, the ovector for
109        those groups contained whatever was previously there. An example is the
110        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
111    
112    2.  If an assertion condition was quantified with a minimum of zero (an odd
113        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
114    
115    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
116        unrecognized modifier, a crash could occur.
117    
118    4.  An attempt to do global matching in pcretest with a zero-length ovector
119        caused a crash.
120    
121    5.  Fixed a memory leak during matching that could occur for a subpattern
122        subroutine call (recursive or otherwise) if the number of captured groups
123        that had to be saved was greater than ten.
124    
125    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
126        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
127        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
128    
129    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
130        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
131    
132    8. When a pattern is compiled, it remembers the highest back reference so that
133       when matching, if the ovector is too small, extra memory can be obtained to
134       use instead. A conditional subpattern whose condition is a check on a
135       capture having happened, such as, for example in the pattern
136       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
137       setting the highest backreference number. This mattered only if pcre_exec()
138       was called with an ovector that was too small to hold the capture, and there
139       was no other kind of back reference (a situation which is probably quite
140       rare). The effect of the bug was that the condition was always treated as
141       FALSE when the capture could not be consulted, leading to a incorrect
142       behaviour by pcre_exec(). This bug has been fixed.
143    
144    9. A reference to a duplicated named group (either a back reference or a test
145       for being set in a conditional) that occurred in a part of the pattern where
146       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
147       to be incorrectly calculated, leading to overwriting.
148    
149    10. A mutually recursive set of back references such as (\2)(\1) caused a
150        segfault at study time (while trying to find the minimum matching length).
151        The infinite loop is now broken (with the minimum length unset, that is,
152        zero).
153    
154    11. If an assertion that was used as a condition was quantified with a minimum
155        of zero, matching went wrong. In particular, if the whole group had
156        unlimited repetition and could match an empty string, a segfault was
157        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
158        assertions to be quantified, but not if they are being used as conditions,
159        so the above pattern is faulted by Perl. PCRE has now been changed so that
160        it also rejects such patterns.
161    
162    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
163        failed to allow the zero-repeat case if pcre2_exec() was called with an
164        ovector too small to capture the group.
165    
166    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
167        Red Hat Product Security:
168    
169        (a) A crash if /K and /F were both set with the option to save the compiled
170        pattern.
171    
172        (b) Another crash if the option to print captured substrings in a callout
173        was combined with setting a null ovector, for example \O\C+ as a subject
174        string.
175    
176    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
177        forward reference repeated a large (but limited) number of times within a
178        repeated outer group that has a zero minimum quantifier, caused incorrect
179        code to be compiled, leading to the error "internal error:
180        previously-checked referenced subpattern not found" when an incorrect
181        memory address was read. This bug was reported as "heap overflow",
182        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
183        CVE-2015-2325.
184    
185    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
186        call within a group that also contained a recursive back reference caused
187        incorrect code to be compiled. This bug was reported as "heap overflow",
188        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
189        number CVE-2015-2326.
190    
191    24. Computing the size of the JIT read-only data in advance has been a source
192        of various issues, and new ones are still appear unfortunately. To fix
193        existing and future issues, size computation is eliminated from the code,
194        and replaced by on-demand memory allocation.
195    
196    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
197        adjacent to the end of the range, and the range contained characters with
198        more than one other case, caused incorrect behaviour when compiled in UTF
199        mode. In that example, the range a-j was left out of the class.
200    
201    26. Fix JIT compilation of conditional blocks, which assertion
202        is converted to (*FAIL). E.g: /(?(?!))/.
203    
204    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
205        discovered by the LLVM fuzzer.
206    
207    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
208        when this assertion was used as a condition, for example (?(?!)a|b). In
209        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
210        error about an unsupported item.
211    
212    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
213        possessification code could take exponential time to complete. A recursion
214        depth limit of 1000 has been imposed to limit the resources used by this
215        optimization.
216    
217    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
218        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
219        because \S ensures they are all in the class. The code for doing this was
220        interacting badly with the code for computing the amount of space needed to
221        compile the pattern, leading to a buffer overflow. This bug was discovered
222        by the LLVM fuzzer.
223    
224    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
225        other kinds of group caused stack overflow at compile time. This bug was
226        discovered by the LLVM fuzzer.
227    
228    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
229        between a subroutine call and its quantifier was incorrectly compiled,
230        leading to buffer overflow or other errors. This bug was discovered by the
231        LLVM fuzzer.
232    
233    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
234        assertion after (?(. The code was failing to check the character after
235        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
236        was discovered by the LLVM fuzzer.
237    
238    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
239        a fixed maximum following a group that contains a subroutine reference was
240        incorrectly compiled and could trigger buffer overflow. This bug was
241        discovered by the LLVM fuzzer.
242    
243    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
244        caused a stack overflow instead of the diagnosis of a non-fixed length
245        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
246    
247    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
248        (e.g. /(?<=\Ka)/) could make pcregrep loop.
249    
250    37. There was a similar problem to 36 in pcretest for global matches.
251    
252    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
253        and a subsequent item in the pattern caused a non-match, backtracking over
254        the repeated \X did not stop, but carried on past the start of the subject,
255        causing reference to random memory and/or a segfault. There were also some
256        other cases where backtracking after \C could crash. This set of bugs was
257        discovered by the LLVM fuzzer.
258    
259    39. The function for finding the minimum length of a matching string could take
260        a very long time if mutual recursion was present many times in a pattern,
261        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
262        method has been implemented. This infelicity was discovered by the LLVM
263        fuzzer.
264    
265    40. Static linking against the PCRE library using the pkg-config module was
266        failing on missing pthread symbols.
267    
268    
269    Version 8.36 26-September-2014
270    ------------------------------
271    
272    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
273        -Wmissing-field-initializers and -Wunused-parameter.
274    
275    2.  The tests for quantifiers being too big (greater than 65535) were being
276        applied after reading the number, and stupidly assuming that integer
277        overflow would give a negative number. The tests are now applied as the
278        numbers are read.
279    
280    3.  Tidy code in pcre_exec.c where two branches that used to be different are
281        now the same.
282    
283    4.  The JIT compiler did not generate match limit checks for certain
284        bracketed expressions with quantifiers. This may lead to exponential
285        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
286        issue should be resolved now.
287    
288    5.  Fixed an issue, which occures when nested alternatives are optimized
289        with table jumps.
290    
291    6.  Inserted two casts and changed some ints to size_t in the light of some
292        reported 64-bit compiler warnings (Bugzilla 1477).
293    
294    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
295        an empty string, which sometimes were behaving incorrectly in the
296        interpreter (though correctly in the JIT matcher). This pcretest input is
297        an example:
298    
299          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
300          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
301    
302        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
303        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
304        for an empty string was breaking the inner loop and carrying on at a lower
305        level, when possessive repeated groups should always return to a higher
306        level as they have no backtrack points in them. The empty string test now
307        occurs at the outer level.
308    
309    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
310        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
311    
312    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
313        doesn't).
314    
315    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
316        to hold a pattern to be incorrectly computed (too small) when there were
317        named back references to duplicated names. This could cause "internal
318        error: code overflow" or "double free or corruption" or other memory
319        handling errors.
320    
321    11. When named subpatterns had the same prefixes, back references could be
322        confused. For example, in this pattern:
323    
324          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
325    
326        the reference to 'Name' was incorrectly treated as a reference to a
327        duplicate name.
328    
329    12. A pattern such as /^s?c/mi8 where the optional character has more than
330        one "other case" was incorrectly compiled such that it would only try to
331        match starting at "c".
332    
333    13. When a pattern starting with \s was studied, VT was not included in the
334        list of possible starting characters; this should have been part of the
335        8.34/18 patch.
336    
337    14. If a character class started [\Qx]... where x is any character, the class
338        was incorrectly terminated at the ].
339    
340    15. If a pattern that started with a caseless match for a character with more
341        than one "other case" was studied, PCRE did not set up the starting code
342        unit bit map for the list of possible characters. Now it does. This is an
343        optimization improvement, not a bug fix.
344    
345    16. The Unicode data tables have been updated to Unicode 7.0.0.
346    
347    17. Fixed a number of memory leaks in pcregrep.
348    
349    18. Avoid a compiler warning (from some compilers) for a function call with
350        a cast that removes "const" from an lvalue by using an intermediate
351        variable (to which the compiler does not object).
352    
353    19. Incorrect code was compiled if a group that contained an internal recursive
354        back reference was optional (had quantifier with a minimum of zero). This
355        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
356        caused segmentation faults because of stack overflows at compile time.
357    
358    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
359        group that is quantified with an indefinite repeat, caused a compile-time
360        loop which used up all the system stack and provoked a segmentation fault.
361        This was not the same bug as 19 above.
362    
363    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
364        Patch by Mike Frysinger.
365    
366    
367    Version 8.35 04-April-2014
368    --------------------------
369    
370    1.  A new flag is set, when property checks are present in an XCLASS.
371        When this flag is not set, PCRE can perform certain optimizations
372        such as studying these XCLASS-es.
373    
374    2.  The auto-possessification of character sets were improved: a normal
375        and an extended character set can be compared now. Furthermore
376        the JIT compiler optimizes more character set checks.
377    
378    3.  Got rid of some compiler warnings for potentially uninitialized variables
379        that show up only when compiled with -O2.
380    
381    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
382        of a match later then the end of the match. The pcretest program was not
383        handling the case sensibly - it was outputting from the start to the next
384        binary zero. It now reports this situation in a message, and outputs the
385        text from the end to the start.
386    
387    5.  Fast forward search is improved in JIT. Instead of the first three
388        characters, any three characters with fixed position can be searched.
389        Search order: first, last, middle.
390    
391    6.  Improve character range checks in JIT. Characters are read by an inprecise
392        function now, which returns with an unknown value if the character code is
393        above a certain threshold (e.g: 256). The only limitation is that the value
394        must be bigger than the threshold as well. This function is useful when
395        the characters above the threshold are handled in the same way.
396    
397    7.  The macros whose names start with RAWUCHAR are placeholders for a future
398        mode in which only the bottom 21 bits of 32-bit data items are used. To
399        make this more memorable for those maintaining the code, the names have
400        been changed to start with UCHAR21, and an extensive comment has been added
401        to their definition.
402    
403    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
404        to the export list in Makefile.am (they were accidentally omitted from the
405        8.34 tarball).
406    
407    9.  The informational output from pcretest used the phrase "starting byte set"
408        which is inappropriate for the 16-bit and 32-bit libraries. As the output
409        for "first char" and "need char" really means "non-UTF-char", I've changed
410        "byte" to "char", and slightly reworded the output. The documentation about
411        these values has also been (I hope) clarified.
412    
413    10. Another JIT related optimization: use table jumps for selecting the correct
414        backtracking path, when more than four alternatives are present inside a
415        bracket.
416    
417    11. Empty match is not possible, when the minimum length is greater than zero,
418        and there is no \K in the pattern. JIT should avoid empty match checks in
419        such cases.
420    
421    12. In a caseless character class with UCP support, when a character with more
422        than one alternative case was not the first character of a range, not all
423        the alternative cases were added to the class. For example, s and \x{17f}
424        are both alternative cases for S: the class [RST] was handled correctly,
425        but [R-T] was not.
426    
427    13. The configure.ac file always checked for pthread support when JIT was
428        enabled. This is not used in Windows, so I have put this test inside a
429        check for the presence of windows.h (which was already tested for).
430    
431    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
432        The algorithm provides a way to skip certain starting offsets, and usually
433        faster than linear prefix searches.
434    
435    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
436        as for 'fr_FR' and 'french'. For some reason, however, it then used the
437        Windows-specific input and output files, which have 'french' screwed in.
438        So this could never have worked. One of the problems with locales is that
439        they aren't always the same. I have now updated RunTest so that it checks
440        the output of the locale test (test 3) against three different output
441        files, and it allows the test to pass if any one of them matches. With luck
442        this should make the test pass on some versions of Solaris where it was
443        failing. Because of the uncertainty, the script did not used to stop if
444        test 3 failed; it now does. If further versions of a French locale ever
445        come to light, they can now easily be added.
446    
447    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
448        there was a message during ./configure, but it did not stop. This now
449        provokes an error. The invalid example in README has been corrected.
450        If a value less than the minimum is given, the minimum value has always
451        been used, but now a warning is given.
452    
453    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
454        was a bug in the test system, which is now fixed. Also, the list of various
455        configurations that are tested for each release did not have one with both
456        16/32 bits and --enable-bar-anycrlf. It now does.
457    
458    18. pcretest was missing "-C bsr" for displaying the \R default setting.
459    
460    19. Little endian PowerPC systems are supported now by the JIT compiler.
461    
462    20. The fast forward newline mechanism could enter to an infinite loop on
463        certain invalid UTF-8 input. Although we don't support these cases
464        this issue can be fixed by a performance optimization.
465    
466    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
467        not take account if existing stack usage. There is now a new global
468        variable called pcre_stack_guard that can be set to point to an external
469        function to check stack availability. It is called at the start of
470        processing every parenthesized group.
471    
472    22. A typo in the code meant that in ungreedy mode the max/min qualifier
473        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
474        not match "ab".
475    
476    23. When UTF was disabled, the JIT program reported some incorrect compile
477        errors. These messages are silenced now.
478    
479    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
480        compiler.
481    
482    25. Change all the temporary files used in RunGrepTest to be different to those
483        used by RunTest so that the tests can be run simultaneously, for example by
484        "make -j check".
485    
486    
487    Version 8.34 15-December-2013
488    -----------------------------
489    
490    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
491        executable memory. Patch inspired by Carsten Klein.
492    
493    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
494        this macro is never tested and has no effect, because the work to support
495        coverage involves only compiling and linking options and special targets in
496        the Makefile. The comment in config.h implied that defining the macro would
497        enable coverage support, which is totally false. There was also support for
498        setting this macro in the CMake files (my fault, I just copied it from
499        configure). SUPPORT_GCOV has now been removed.
500    
501    3.  Make a small performance improvement in strlen16() and strlen32() in
502        pcretest.
503    
504    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
505        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
506        these cases). There was also one in pcretest.c.
507    
508    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
509    
510    6.  In UTF mode, the code for checking whether a group could match an empty
511        string (which is used for indefinitely repeated groups to allow for
512        breaking an infinite loop) was broken when the group contained a repeated
513        negated single-character class with a character that occupied more than one
514        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
515        UTF-8 mode). The effect was undefined: the group might or might not be
516        deemed as matching an empty string, or the program might have crashed.
517    
518    7.  The code for checking whether a group could match an empty string was not
519        recognizing that \h, \H, \v, \V, and \R must match a character.
520    
521    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
522        an empty string. If it can, pcretest shows this in its information output.
523    
524    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
525        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
526        matched by pcre_exec() without using JIT:
527    
528        (a) If the rest of the pattern did not match after a maximal run of
529            grapheme clusters, the code for backing up to try with fewer of them
530            did not always back up over a full grapheme when characters that do not
531            have the modifier quality were involved, e.g. Hangul syllables.
532    
533        (b) If the match point in a subject started with modifier character, and
534            there was no match, the code could incorrectly back up beyond the match
535            point, and potentially beyond the first character in the subject,
536            leading to a segfault or an incorrect match result.
537    
538    10. A conditional group with an assertion condition could lead to PCRE
539        recording an incorrect first data item for a match if no other first data
540        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
541        first data item, and therefore matched "ca" after "c" instead of at the
542        start.
543    
544    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
545        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
546    
547    12. The source of pcregrep now includes z/OS-specific code so that it can be
548        compiled for z/OS as part of the special z/OS distribution.
549    
550    13. Added the -T and -TM options to pcretest.
551    
552    14. The code in pcre_compile.c for creating the table of named capturing groups
553        has been refactored. Instead of creating the table dynamically during the
554        actual compiling pass, the information is remembered during the pre-compile
555        pass (on the stack unless there are more than 20 named groups, in which
556        case malloc() is used) and the whole table is created before the actual
557        compile happens. This has simplified the code (it is now nearly 150 lines
558        shorter) and prepared the way for better handling of references to groups
559        with duplicate names.
560    
561    15. A back reference to a named subpattern when there is more than one of the
562        same name now checks them in the order in which they appear in the pattern.
563        The first one that is set is used for the reference. Previously only the
564        first one was inspected. This change makes PCRE more compatible with Perl.
565    
566    16. Unicode character properties were updated from Unicode 6.3.0.
567    
568    17. The compile-time code for auto-possessification has been refactored, based
569        on a patch by Zoltan Herczeg. It now happens after instead of during
570        compilation. The code is cleaner, and more cases are handled. The option
571        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
572        options in pcretest are provided to set it. It can also be set by
573        (*NO_AUTO_POSSESS) at the start of a pattern.
574    
575    18. The character VT has been added to the default ("C" locale) set of
576        characters that match \s and are generally treated as white space,
577        following this same change in Perl 5.18. There is now no difference between
578        "Perl space" and "POSIX space". Whether VT is treated as white space in
579        other locales depends on the locale.
580    
581    19. The code for checking named groups as conditions, either for being set or
582        for being recursed, has been refactored (this is related to 14 and 15
583        above). Processing unduplicated named groups should now be as fast at
584        numerical groups, and processing duplicated groups should be faster than
585        before.
586    
587    20. Two patches to the CMake build system, by Alexander Barkov:
588    
589          (1) Replace the "source" command by "." in CMakeLists.txt because
590              "source" is a bash-ism.
591    
592          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
593              without these the CMake build does not work on Solaris.
594    
595    21. Perl has changed its handling of \8 and \9. If there is no previously
596        encountered capturing group of those numbers, they are treated as the
597        literal characters 8 and 9 instead of a binary zero followed by the
598        literals. PCRE now does the same.
599    
600    22. Following Perl, added \o{} to specify codepoints in octal, making it
601        possible to specify values greater than 0777 and also making them
602        unambiguous.
603    
604    23. Perl now gives an error for missing closing braces after \x{... instead of
605        treating the string as literal. PCRE now does the same.
606    
607    24. RunTest used to grumble if an inappropriate test was selected explicitly,
608        but just skip it when running all tests. This make it awkward to run ranges
609        of tests when one of them was inappropriate. Now it just skips any
610        inappropriate tests, as it always did when running all tests.
611    
612    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
613        character types such as \d or \w, too many callouts were inserted, and the
614        data that they returned was rubbish.
615    
616    26. In UCP mode, \s was not matching two of the characters that Perl matches,
617        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
618        were matched by \h. The code has now been refactored so that the lists of
619        the horizontal and vertical whitespace characters used for \h and \v (which
620        are defined only in one place) are now also used for \s.
621    
622    27. Add JIT support for the 64 bit TileGX architecture.
623        Patch by Jiong Wang (Tilera Corporation).
624    
625    28. Possessive quantifiers for classes (both explicit and automatically
626        generated) now use special opcodes instead of wrapping in ONCE brackets.
627    
628    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
629        (because it's meaningless), this was not happening when PCRE_CASELESS was
630        set. Not wrong, but inefficient.
631    
632    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
633        properties for \w, \d, etc) is present in a test regex. Otherwise if the
634        test contains no characters greater than 255, Perl doesn't realise it
635        should be using Unicode semantics.
636    
637    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
638        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
639        does in Unicode mode.
640    
641    32. Added the "forbid" facility to pcretest so that putting tests into the
642        wrong test files can sometimes be quickly detected.
643    
644    33. There is now a limit (default 250) on the depth of nesting of parentheses.
645        This limit is imposed to control the amount of system stack used at compile
646        time. It can be changed at build time by --with-parens-nest-limit=xxx or
647        the equivalent in CMake.
648    
649    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
650        errors. Perl warns for these when in warning mode, but PCRE has no facility
651        for giving warnings.
652    
653    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
654        However, this was not working for (?!) because it is optimized to (*FAIL),
655        for which PCRE does not allow quantifiers. The optimization is now disabled
656        when a quantifier follows (?!). I can't see any use for this, but it makes
657        things uniform.
658    
659    36. Perl no longer allows group names to start with digits, so I have made this
660        change also in PCRE. It simplifies the code a bit.
661    
662    37. In extended mode, Perl ignores spaces before a + that indicates a
663        possessive quantifier. PCRE allowed a space before the quantifier, but not
664        before the possessive +. It now does.
665    
666    38. The use of \K (reset reported match start) within a repeated possessive
667        group such as (a\Kb)*+ was not working.
668    
669    40. Document that the same character tables must be used at compile time and
670        run time, and that the facility to pass tables to pcre_exec() and
671        pcre_dfa_exec() is for use only with saved/restored patterns.
672    
673    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
674        features for Builds with MSVC:
675    
676        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
677           builds). The .rc files can be used to set FileDescription and many other
678           attributes.
679    
680        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
681           This allows higher-level build scripts which want .pdb files to avoid
682           hard-coding the exact files needed."
683    
684    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
685        mean "start of word" and "end of word", respectively, as a transition aid.
686    
687    43. A minimizing repeat of a class containing codepoints greater than 255 in
688        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
689        compiled to use the heap for recursion.
690    
691    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
692        is configured.
693    
694    
695    Version 8.33 28-May-2013
696    ------------------------
697    
698    1.  Added 'U' to some constants that are compared to unsigned integers, to
699        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
700        variables that are added to signed variables, to ensure the result is
701        signed and can be negated.
702    
703    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
704        CMake config files.
705    
706    3.  Revise the creation of config.h.generic so that all boolean macros are
707        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
708        overriding via -D on the command line possible.
709    
710    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
711        to unsigned int is reported to make a quite noticeable speed difference in
712        a specific Windows environment. Testing on Linux did also appear to show
713        some benefit (and it is clearly not harmful). Also fixed the definition of
714        Xop which should be unsigned.
715    
716    5.  Related to (4), changing the definition of the intermediate variable cc
717        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
718        improvements.
719    
720    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
721        unnecessary spaces.
722    
723    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
724        and later.
725    
726    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
727    
728    9.  Optimizing fast_forward_start_bits in JIT.
729    
730    10. Adding support for callouts in JIT, and fixing some issues revealed
731        during this work. Namely:
732    
733        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
734    
735        (b) Minimum length was not checked before the matching is started.
736    
737    11. The value of capture_last that is passed to callouts was incorrect in some
738        cases when there was a capture on one path that was subsequently abandoned
739        after a backtrack. Also, the capture_last value is now reset after a
740        recursion, since all captures are also reset in this case.
741    
742    12. The interpreter no longer returns the "too many substrings" error in the
743        case when an overflowing capture is in a branch that is subsequently
744        abandoned after a backtrack.
745    
746    13. In the pathological case when an offset vector of size 2 is used, pcretest
747        now prints out the matched string after a yield of 0 or 1.
748    
749    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
750        Only supported by the JIT compiler at the moment.
751    
752    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
753    
754    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
755        offset of the starting point of the matching process, provided the offsets
756        vector is large enough.
757    
758    17. The \A escape now records a lookbehind value of 1, though its execution
759        does not actually inspect the previous character. This is to ensure that,
760        in partial multi-segment matching, at least one character from the old
761        segment is retained when a new segment is processed. Otherwise, if there
762        are no lookbehinds in the pattern, \A might match incorrectly at the start
763        of a new segment.
764    
765    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
766    
767    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
768        gives some modest performance improvement in 8-bit mode.
769    
770    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
771        be expressed in certain programming languages using Universal Character
772        Names.
773    
774    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
775        which points out that "non characters" are not "characters that may not
776        appear in Unicode strings" but rather "characters that are reserved for
777        internal use and have only local meaning".
778    
779    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
780        there was a conditional group that depended on an assertion, if the
781        assertion was false, the callout that immediately followed the alternation
782        in the condition was skipped when pcre_exec() was used for matching.
783    
784    23. Allow an explicit callout to be inserted before an assertion that is the
785        condition for a conditional group, for compatibility with automatic
786        callouts, which always insert a callout at this point.
787    
788    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
789        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
790    
791    25. (*PRUNE) is now supported by the JIT compiler.
792    
793    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
794    
795    27. Fix the case where there are two or more SKIPs with arguments that may be
796        ignored.
797    
798    28. (*SKIP) is now supported by the JIT compiler.
799    
800    29. (*THEN) is now supported by the JIT compiler.
801    
802    30. Update RunTest with additional test selector options.
803    
804    31. The way PCRE handles backtracking verbs has been changed in two ways.
805    
806        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
807        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
808        backtracking. In some cases this makes it more Perl-compatible, but Perl's
809        rather obscure rules do not always do the same thing.
810    
811        (2) Previously, backtracking verbs were confined within assertions. This is
812        no longer the case for positive assertions, except for (*ACCEPT). Again,
813        this sometimes improves Perl compatibility, and sometimes does not.
814    
815    32. A number of tests that were in test 2 because Perl did things differently
816        have been moved to test 1, because either Perl or PCRE has changed, and
817        these tests are now compatible.
818    
819    32. Backtracking control verbs are now handled in the same way in JIT and
820        interpreter.
821    
822    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
823        contained a forward subroutine reference caused a compile error.
824    
825    34. Auto-detect and optimize limited repetitions in JIT.
826    
827    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
828        blocking (*UTF) etc.
829    
830    36. In the interpreter, maximizing pattern repetitions for characters and
831        character types now use tail recursion, which reduces stack usage.
832    
833    37. The value of the max lookbehind was not correctly preserved if a compiled
834        and saved regex was reloaded on a host of different endianness.
835    
836    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
837        of the compiled pattern block, expand the flags field from 16 to 32 bits
838        because it was almost full.
839    
840    39. Try madvise first before posix_madvise.
841    
842    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
843        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
844        original change locked out all matches of empty strings. This has been
845        changed so that one match of an empty string per line is recognized.
846        Subsequent searches on the same line (for colouring or for --only-matching,
847        for example) do not recognize empty strings.
848    
849    41. Applied a user patch to fix a number of spelling mistakes in comments.
850    
851    42. Data lines longer than 65536 caused pcretest to crash.
852    
853    43. Clarified the data type for length and startoffset arguments for pcre_exec
854        and pcre_dfa_exec in the function-specific man pages, where they were
855        explicitly stated to be in bytes, never having been updated. I also added
856        some clarification to the pcreapi man page.
857    
858    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
859        a segmentation fault.
860    
861    
862    Version 8.32 30-November-2012
863    -----------------------------
864    
865    1.  Improved JIT compiler optimizations for first character search and single
866        character iterators.
867    
868    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
869        Patch by Daniel Richard G.
870    
871    3.  Single character iterator optimizations in the JIT compiler.
872    
873    4.  Improved JIT compiler optimizations for character ranges.
874    
875    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
876        Reported by Giuseppe D'Angelo.
877    
878    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
879        of a line, was being set incorrectly in cases where .* appeared inside
880        atomic brackets at the start of a pattern, or where there was a subsequent
881        *PRUNE or *SKIP.
882    
883    7.  Improved instruction cache flush for POWER/PowerPC.
884        Patch by Daniel Richard G.
885    
886    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
887        grep:
888    
889        (a) There is now no limit to the number of patterns to be matched.
890    
891        (b) An error is given if a pattern is too long.
892    
893        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
894            are now supported.
895    
896        (d) --exclude-from and --include-from (multiple use) have been added.
897    
898        (e) Exclusions and inclusions now apply to all files and directories, not
899            just to those obtained from scanning a directory recursively.
900    
901        (f) Multiple uses of -f and --file-list are now supported.
902    
903        (g) In a Windows environment, the default for -d has been changed from
904            "read" (the GNU grep default) to "skip", because otherwise the presence
905            of a directory in the file list provokes an error.
906    
907        (h) The documentation has been revised and clarified in places.
908    
909    9.  Improve the matching speed of capturing brackets.
910    
911    10. Changed the meaning of \X so that it now matches a Unicode extended
912        grapheme cluster.
913    
914    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
915        out POSIX threads when JIT support is configured.
916    
917    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
918    
919    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
920        structure could go wrong in environments where size_t is not the same size
921        as int.
922    
923    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
924        set.
925    
926    15. The EBCDIC support had decayed; later updates to the code had included
927        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
928        general tidy up of EBCDIC-related issues, and the documentation was also
929        not quite right. There is now a test that can be run on ASCII systems to
930        check some of the EBCDIC-related things (but is it not a full test).
931    
932    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
933        in a small tidy to the code.
934    
935    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
936    
937    18. If the --only-matching (-o) option in pcregrep is specified multiple
938        times, each one causes appropriate output. For example, -o1 -o2 outputs the
939        substrings matched by the 1st and 2nd capturing parentheses. A separating
940        string can be specified by --om-separator (default empty).
941    
942    19. Improving the first n character searches.
943    
944    20. Turn case lists for horizontal and vertical white space into macros so that
945        they are defined only once.
946    
947    21. This set of changes together give more compatible Unicode case-folding
948        behaviour for characters that have more than one other case when UCP
949        support is available.
950    
951        (a) The Unicode property table now has offsets into a new table of sets of
952            three or more characters that are case-equivalent. The MultiStage2.py
953            script that generates these tables (the pcre_ucd.c file) now scans
954            CaseFolding.txt instead of UnicodeData.txt for character case
955            information.
956    
957        (b) The code for adding characters or ranges of characters to a character
958            class has been abstracted into a generalized function that also handles
959            case-independence. In UTF-mode with UCP support, this uses the new data
960            to handle characters with more than one other case.
961    
962        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
963            whose other case is greater than 256 are now correctly matched
964            caselessly. Previously, the high codepoint matched the low one, but not
965            vice versa.
966    
967        (d) The processing of \h, \H, \v, and \ in character classes now makes use
968            of the new class addition function, using character lists defined as
969            macros alongside the case definitions of 20 above.
970    
971        (e) Caseless back references now work with characters that have more than
972            one other case.
973    
974        (f) General caseless matching of characters with more than one other case
975            is supported.
976    
977    22. Unicode character properties were updated from Unicode 6.2.0
978    
979    23. Improved CMake support under Windows. Patch by Daniel Richard G.
980    
981    24. Add support for 32-bit character strings, and UTF-32
982    
983    25. Major JIT compiler update (code refactoring and bugfixing).
984        Experimental Sparc 32 support is added.
985    
986    26. Applied a modified version of Daniel Richard G's patch to create
987        pcre.h.generic and config.h.generic by "make" instead of in the
988        PrepareRelease script.
989    
990    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
991        pcre_compile.c when checking for a zero character.
992    
993    28. Introducing a native interface for JIT. Through this interface, the compiled
994        machine code can be directly executed. The purpose of this interface is to
995        provide fast pattern matching, so several sanity checks are not performed.
996        However, feature tests are still performed. The new interface provides
997        1.4x speedup compared to the old one.
998    
999    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
1000        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
1001        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
1002    
1003    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
1004        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
1005        the "old" RFC 2279). Instead, it ended up passing a negative length to
1006        pcre_exec().
1007    
1008    31. Add support for GCC's visibility feature to hide internal functions.
1009    
1010    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
1011        "unknown -C option" after outputting 0 or 1.
1012    
1013    33. There is now support for generating a code coverage report for the test
1014        suite in environments where gcc is the compiler and lcov is installed. This
1015        is mainly for the benefit of the developers.
1016    
1017    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
1018        unaddressable using valgrind annotations, allowing valgrind to detect
1019        invalid memory accesses. This is mainly for the benefit of the developers.
1020    
1021    25. (*UTF) can now be used to start a pattern in any of the three libraries.
1022    
1023    26. Give configure error if --enable-cpp but no C++ compiler found.
1024    
1025    
1026    Version 8.31 06-July-2012
1027    -------------------------
1028    
1029    1.  Fixing a wrong JIT test case and some compiler warnings.
1030    
1031    2.  Removed a bashism from the RunTest script.
1032    
1033    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
1034        to unsigned type, result still unsigned" that was given by an MS compiler
1035        on encountering the code "-sizeof(xxx)".
1036    
1037    4.  Partial matching support is added to the JIT compiler.
1038    
1039    5.  Fixed several bugs concerned with partial matching of items that consist
1040        of more than one character:
1041    
1042        (a) /^(..)\1/ did not partially match "aba" because checking references was
1043            done on an "all or nothing" basis. This also applied to repeated
1044            references.
1045    
1046        (b) \R did not give a hard partial match if \r was found at the end of the
1047            subject.
1048    
1049        (c) \X did not give a hard partial match after matching one or more
1050            characters at the end of the subject.
1051    
1052        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1053            a partial match for the string "\r".
1054    
1055        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1056            a partial match for a CR character at the end of the subject string.
1057    
1058    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1059        running pcretest, the text "(JIT)" added to the output whenever JIT is
1060        actually used to run the match.
1061    
1062    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1063        or /S+[+] with a digit between 1 and 7.
1064    
1065    8.  OP_NOT now supports any UTF character not just single-byte ones.
1066    
1067    9.  (*MARK) control verb is now supported by the JIT compiler.
1068    
1069    10. The command "./RunTest list" lists the available tests without actually
1070        running any of them. (Because I keep forgetting what they all are.)
1071    
1072    11. Add PCRE_INFO_MAXLOOKBEHIND.
1073    
1074    12. Applied a (slightly modified) user-supplied patch that improves performance
1075        when the heap is used for recursion (compiled with --disable-stack-for-
1076        recursion). Instead of malloc and free for each heap frame each time a
1077        logical recursion happens, frames are retained on a chain and re-used where
1078        possible. This sometimes gives as much as 30% improvement.
1079    
1080    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1081        call.
1082    
1083    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1084    
1085    15. It is now possible to link pcretest with libedit as an alternative to
1086        libreadline.
1087    
1088    16. (*COMMIT) control verb is now supported by the JIT compiler.
1089    
1090    17. The Unicode data tables have been updated to Unicode 6.1.0.
1091    
1092    18. Added --file-list option to pcregrep.
1093    
1094    19. Added binary file support to pcregrep, including the -a, --binary-files,
1095        -I, and --text options.
1096    
1097    20. The madvise function is renamed for posix_madvise for QNX compatibility
1098        reasons. Fixed by Giuseppe D'Angelo.
1099    
1100    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1101    
1102    22. Changed the option for creating symbolic links for 16-bit man pages from
1103        -s to -sf so that re-installing does not cause issues.
1104    
1105    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1106    
1107    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1108        matches in certain environments (the workspace was not being correctly
1109        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1110        some of the workspace data at the beginning of a restart.
1111    
1112    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1113        was not doing so when it should - probably a typo introduced by SVN 528
1114        (change 8.10/14).
1115    
1116    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1117        \w+ when the character tables indicated that \x{c4} was a word character.
1118        There were several related cases, all because the tests for doing a table
1119        lookup were testing for characters less than 127 instead of 255.
1120    
1121    27. If a pattern contains capturing parentheses that are not used in a match,
1122        their slots in the ovector are set to -1. For those that are higher than
1123        any matched groups, this happens at the end of processing. In the case when
1124        there were back references that the ovector was too small to contain
1125        (causing temporary malloc'd memory to be used during matching), and the
1126        highest capturing number was not used, memory off the end of the ovector
1127        was incorrectly being set to -1. (It was using the size of the temporary
1128        memory instead of the true size.)
1129    
1130    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1131        ovector size, it uses memory at the end of the block that it has got.
1132    
1133    29. Check for an overlong MARK name and give an error at compile time. The
1134        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1135    
1136    30. JIT compiler update.
1137    
1138    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1139        Rill for the patch.
1140    
1141    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1142    
1143    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1144    
1145    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1146        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1147        was enabled.
1148    
1149    35. Improve JIT code generation for greedy plus quantifier.
1150    
1151    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1152        1 to "aa" instead of to an empty string. The bug affected repeated groups
1153        that could potentially match an empty string.
1154    
1155    37. Optimizing single character iterators in JIT.
1156    
1157    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1158        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1159        codepoints that are too big for the mode are faulted, and in a UTF mode,
1160        disallowed codepoints are also faulted.
1161    
1162    39. If PCRE was compiled with UTF support, in three places in the DFA
1163        matcher there was code that should only have been obeyed in UTF mode, but
1164        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1165        processing when bytes with values greater than 127 were present. In 16-bit
1166        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1167        both cases the values are those that cannot be the first data item in a UTF
1168        character. The three items that might have provoked this were recursions,
1169        possessively repeated groups, and atomic groups.
1170    
1171    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1172        and pcregrep, because some OS require shared objects to be explicitly
1173        passed to ld, causing the link step to fail if they are not.
1174    
1175    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1176        mode, patterns that started with \h* or \R* might be incorrectly matched.
1177    
1178    
1179    Version 8.30 04-February-2012
1180    -----------------------------
1181    
1182    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1183        name is defined in ctype.h.
1184    
1185    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1186        only in quite long subpatterns.
1187    
1188    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1189        since it was replaced by pcre_fullinfo() in February 2000.
1190    
1191    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1192        match a (*MARK), and the match failed at the start of the subject, a
1193        reference to memory before the start of the subject could occur. This bug
1194        was introduced by fix 17 of release 8.21.
1195    
1196    5.  A reference to an unset group with zero minimum repetition was giving
1197        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1198        /(another)?(\1?)test/ matched against "hello world test". This bug was
1199        introduced in release 8.13.
1200    
1201    6.  Add support for 16-bit character strings (a large amount of work involving
1202        many changes and refactorings).
1203    
1204    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1205        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1206        from a file.
1207    
1208    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1209        rounding is not applied in this particular case).
1210    
1211    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1212        if they appear, or are escaped, in patterns.
1213    
1214    10. Get rid of a number of -Wunused-but-set-variable warnings.
1215    
1216    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1217        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1218        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1219        also returns the mark "x". This bug applied to capturing parentheses,
1220        non-capturing parentheses, and atomic parentheses. It also applied to some
1221        assertions.
1222    
1223    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1224        information out of configure.ac instead of relying on pcre.h.generic, which
1225        is not stored in the repository.
1226    
1227    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1228        -lreadline.
1229    
1230    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1231    
1232    15. Applied Graycode's patch to put the top-level frame on the stack rather
1233        than the heap when not using the stack for recursion. This gives a
1234        performance improvement in many cases when recursion is not deep.
1235    
1236    16. Experimental code added to "pcretest -C" to output the stack frame size.
1237    
1238    
1239    Version 8.21 12-Dec-2011
1240    ------------------------
1241    
1242  1.  Updating the JIT compiler.  1.  Updating the JIT compiler.
1243    
# Line 13  Version 8.21 Line 1248  Version 8.21
1248      PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before      PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1249      calling _pcre_jit_exec. Some extra comments are added.      calling _pcre_jit_exec. Some extra comments are added.
1250    
1251  4.  Mark settings inside atomic groups that do not contain any capturing  4.  (*MARK) settings inside atomic groups that do not contain any capturing
1252      parentheses, for example, (?>a(*:m)), were not being passed out. This bug      parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1253      was introduced by change 18 for 8.20.      was introduced by change 18 for 8.20.
1254    
# Line 22  Version 8.21 Line 1257  Version 8.21
1257    
1258  6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were  6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1259      erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.      erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1260      This bug was probably introduced by change 9 of 8.13.      This bug was probably introduced by change 9 of 8.13.
1261    
1262  7.  While fixing 6 above, I noticed that a number of other items were being  7.  While fixing 6 above, I noticed that a number of other items were being
1263      incorrectly rejected as "not fixed length". This arose partly because newer      incorrectly rejected as "not fixed length". This arose partly because newer
1264      opcodes had not been added to the fixed-length checking code. I have (a)      opcodes had not been added to the fixed-length checking code. I have (a)
1265      corrected the bug and added tests for these items, and (b) arranged for an      corrected the bug and added tests for these items, and (b) arranged for an
1266      error to occur if an unknown opcode is encountered while checking for fixed      error to occur if an unknown opcode is encountered while checking for fixed
1267      length instead of just assuming "not fixed length". The items that were      length instead of just assuming "not fixed length". The items that were
1268      rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),      rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1269      (*THEN), \h, \H, \v, \V, and single character negative classes with fixed      (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1270      repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.      repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1271    
1272  8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was  8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1273      being incorrectly compiled and would have given unpredicatble results.      being incorrectly compiled and would have given unpredicatble results.
1274    
1275  9.  A possessively repeated subpattern with minimum repeat count greater than  9.  A possessively repeated subpattern with minimum repeat count greater than
1276      one behaved incorrectly. For example, (A){2,}+ behaved as if it was      one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1277      (A)(A)++ which meant that, after a subsequent mismatch, backtracking into      (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1278      the first (A) could occur when it should not.      the first (A) could occur when it should not.
1279    
1280  10. Add a cast and remove a redundant test from the code.  10. Add a cast and remove a redundant test from the code.
1281    
1282  11. JIT should use pcre_malloc/pcre_free for allocation.  11. JIT should use pcre_malloc/pcre_free for allocation.
1283    
1284  12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems  12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1285      best practice nowadays, and helps with cross-compiling. (If the exec_prefix      best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1286      is anything other than /usr, -L is still shown).      is anything other than /usr, -L is still shown).
1287    
1288    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1289    
1290    14. Perl does not support \N without a following name in a [] class; PCRE now
1291        also gives an error.
1292    
1293    15. If a forward reference was repeated with an upper limit of around 2000,
1294        it caused the error "internal error: overran compiling workspace". The
1295        maximum number of forward references (including repeats) was limited by the
1296        internal workspace, and dependent on the LINK_SIZE. The code has been
1297        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1298        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1299        of around 200,000 forward references. While doing this, I also speeded up
1300        the filling in of repeated forward references.
1301    
1302    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1303        incorrectly expecting the subject to contain another "a" after the start.
1304    
1305    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1306        in the match, the SKIP should be ignored. This was not happening; instead
1307        the SKIP was being treated as NOMATCH. For patterns such as
1308        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1309        tested.
1310    
1311    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1312        now much more compatible with Perl, in particular in cases where the result
1313        is a non-match for a non-anchored pattern. For example, if
1314        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1315        "m", where previously it did not return a name. A side effect of this
1316        change is that for partial matches, the last encountered mark name is
1317        returned, as for non matches. A number of tests that were previously not
1318        Perl-compatible have been moved into the Perl-compatible test files. The
1319        refactoring has had the pleasing side effect of removing one argument from
1320        the match() function, thus reducing its stack requirements.
1321    
1322    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1323        subsequent uses of /S (without +) incorrectly behaved like /S+.
1324    
1325    21. Retrieve executable code size support for the JIT compiler and fixing
1326        some warnings.
1327    
1328    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1329        not work when the shorter character appeared right at the end of the
1330        subject string.
1331    
1332    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1333        systems.
1334    
1335    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1336        output it when the /M option is used in pcretest.
1337    
1338    25. The CheckMan script was not being included in the distribution. Also, added
1339        an explicit "perl" to run Perl scripts from the PrepareRelease script
1340        because this is reportedly needed in Windows.
1341    
1342    26. If study data was being save in a file and studying had not found a set of
1343        "starts with" bytes for the pattern, the data written to the file (though
1344        never used) was taken from uninitialized memory and so caused valgrind to
1345        complain.
1346    
1347    27. Updated RunTest.bat as provided by Sheri Pierce.
1348    
1349    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1350    
1351    29. Computation of memory usage for the table of capturing group names was
1352        giving an unnecessarily large value.
1353    
1354    
1355  Version 8.20 21-Oct-2011  Version 8.20 21-Oct-2011
# Line 1065  Version 7.9 11-Apr-09 Line 2366  Version 7.9 11-Apr-09
2366  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
2367      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
2368      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2369      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2370        8.33.]
2371    
2372  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
2373      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 1308  Version 7.7 07-May-08 Line 2610  Version 7.7 07-May-08
2610      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2611      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2612      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2613      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2614      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2615      was subsequently set up correctly.)      was subsequently set up correctly.)
2616    

Legend:
Removed from v.753  
changed lines
  Added in v.1582

  ViewVC Help
Powered by ViewVC 1.1.5