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

Diff of /code/trunk/ChangeLog

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

revision 657 by ph10, Mon Aug 15 17:39:09 2011 UTC revision 1564 by ph10, Tue Jun 9 16:33:27 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 8.13 02-Aug-2011  Note that the PCRE 8.xx series (PCRE1) is now in a bugfix-only state. All
5    development is happening in the PCRE2 10.xx series.
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    
49    Version 8.37 28-April-2015
50    --------------------------
51    
52    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
53        for those parentheses to be closed with whatever has been captured so far.
54        However, it was failing to mark any other groups between the hightest
55        capture so far and the currrent group as "unset". Thus, the ovector for
56        those groups contained whatever was previously there. An example is the
57        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
58    
59    2.  If an assertion condition was quantified with a minimum of zero (an odd
60        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
61    
62    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
63        unrecognized modifier, a crash could occur.
64    
65    4.  An attempt to do global matching in pcretest with a zero-length ovector
66        caused a crash.
67    
68    5.  Fixed a memory leak during matching that could occur for a subpattern
69        subroutine call (recursive or otherwise) if the number of captured groups
70        that had to be saved was greater than ten.
71    
72    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
73        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
74        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
75    
76    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
77        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
78    
79    8. When a pattern is compiled, it remembers the highest back reference so that
80       when matching, if the ovector is too small, extra memory can be obtained to
81       use instead. A conditional subpattern whose condition is a check on a
82       capture having happened, such as, for example in the pattern
83       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
84       setting the highest backreference number. This mattered only if pcre_exec()
85       was called with an ovector that was too small to hold the capture, and there
86       was no other kind of back reference (a situation which is probably quite
87       rare). The effect of the bug was that the condition was always treated as
88       FALSE when the capture could not be consulted, leading to a incorrect
89       behaviour by pcre_exec(). This bug has been fixed.
90    
91    9. A reference to a duplicated named group (either a back reference or a test
92       for being set in a conditional) that occurred in a part of the pattern where
93       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
94       to be incorrectly calculated, leading to overwriting.
95    
96    10. A mutually recursive set of back references such as (\2)(\1) caused a
97        segfault at study time (while trying to find the minimum matching length).
98        The infinite loop is now broken (with the minimum length unset, that is,
99        zero).
100    
101    11. If an assertion that was used as a condition was quantified with a minimum
102        of zero, matching went wrong. In particular, if the whole group had
103        unlimited repetition and could match an empty string, a segfault was
104        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
105        assertions to be quantified, but not if they are being used as conditions,
106        so the above pattern is faulted by Perl. PCRE has now been changed so that
107        it also rejects such patterns.
108    
109    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
110        failed to allow the zero-repeat case if pcre2_exec() was called with an
111        ovector too small to capture the group.
112    
113    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
114        Red Hat Product Security:
115    
116        (a) A crash if /K and /F were both set with the option to save the compiled
117        pattern.
118    
119        (b) Another crash if the option to print captured substrings in a callout
120        was combined with setting a null ovector, for example \O\C+ as a subject
121        string.
122    
123    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
124        forward reference repeated a large (but limited) number of times within a
125        repeated outer group that has a zero minimum quantifier, caused incorrect
126        code to be compiled, leading to the error "internal error:
127        previously-checked referenced subpattern not found" when an incorrect
128        memory address was read. This bug was reported as "heap overflow",
129        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
130        CVE-2015-2325.
131    
132    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
133        call within a group that also contained a recursive back reference caused
134        incorrect code to be compiled. This bug was reported as "heap overflow",
135        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
136        number CVE-2015-2326.
137    
138    24. Computing the size of the JIT read-only data in advance has been a source
139        of various issues, and new ones are still appear unfortunately. To fix
140        existing and future issues, size computation is eliminated from the code,
141        and replaced by on-demand memory allocation.
142    
143    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
144        adjacent to the end of the range, and the range contained characters with
145        more than one other case, caused incorrect behaviour when compiled in UTF
146        mode. In that example, the range a-j was left out of the class.
147    
148    26. Fix JIT compilation of conditional blocks, which assertion
149        is converted to (*FAIL). E.g: /(?(?!))/.
150    
151    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
152        discovered by the LLVM fuzzer.
153    
154    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
155        when this assertion was used as a condition, for example (?(?!)a|b). In
156        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
157        error about an unsupported item.
158    
159    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
160        possessification code could take exponential time to complete. A recursion
161        depth limit of 1000 has been imposed to limit the resources used by this
162        optimization.
163    
164    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
165        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
166        because \S ensures they are all in the class. The code for doing this was
167        interacting badly with the code for computing the amount of space needed to
168        compile the pattern, leading to a buffer overflow. This bug was discovered
169        by the LLVM fuzzer.
170    
171    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
172        other kinds of group caused stack overflow at compile time. This bug was
173        discovered by the LLVM fuzzer.
174    
175    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
176        between a subroutine call and its quantifier was incorrectly compiled,
177        leading to buffer overflow or other errors. This bug was discovered by the
178        LLVM fuzzer.
179    
180    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
181        assertion after (?(. The code was failing to check the character after
182        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
183        was discovered by the LLVM fuzzer.
184    
185    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
186        a fixed maximum following a group that contains a subroutine reference was
187        incorrectly compiled and could trigger buffer overflow. This bug was
188        discovered by the LLVM fuzzer.
189    
190    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
191        caused a stack overflow instead of the diagnosis of a non-fixed length
192        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
193    
194    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
195        (e.g. /(?<=\Ka)/) could make pcregrep loop.
196    
197    37. There was a similar problem to 36 in pcretest for global matches.
198    
199    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
200        and a subsequent item in the pattern caused a non-match, backtracking over
201        the repeated \X did not stop, but carried on past the start of the subject,
202        causing reference to random memory and/or a segfault. There were also some
203        other cases where backtracking after \C could crash. This set of bugs was
204        discovered by the LLVM fuzzer.
205    
206    39. The function for finding the minimum length of a matching string could take
207        a very long time if mutual recursion was present many times in a pattern,
208        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
209        method has been implemented. This infelicity was discovered by the LLVM
210        fuzzer.
211    
212    40. Static linking against the PCRE library using the pkg-config module was
213        failing on missing pthread symbols.
214    
215    
216    Version 8.36 26-September-2014
217    ------------------------------
218    
219    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
220        -Wmissing-field-initializers and -Wunused-parameter.
221    
222    2.  The tests for quantifiers being too big (greater than 65535) were being
223        applied after reading the number, and stupidly assuming that integer
224        overflow would give a negative number. The tests are now applied as the
225        numbers are read.
226    
227    3.  Tidy code in pcre_exec.c where two branches that used to be different are
228        now the same.
229    
230    4.  The JIT compiler did not generate match limit checks for certain
231        bracketed expressions with quantifiers. This may lead to exponential
232        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
233        issue should be resolved now.
234    
235    5.  Fixed an issue, which occures when nested alternatives are optimized
236        with table jumps.
237    
238    6.  Inserted two casts and changed some ints to size_t in the light of some
239        reported 64-bit compiler warnings (Bugzilla 1477).
240    
241    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
242        an empty string, which sometimes were behaving incorrectly in the
243        interpreter (though correctly in the JIT matcher). This pcretest input is
244        an example:
245    
246          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
247          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
248    
249        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
250        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
251        for an empty string was breaking the inner loop and carrying on at a lower
252        level, when possessive repeated groups should always return to a higher
253        level as they have no backtrack points in them. The empty string test now
254        occurs at the outer level.
255    
256    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
257        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
258    
259    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
260        doesn't).
261    
262    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
263        to hold a pattern to be incorrectly computed (too small) when there were
264        named back references to duplicated names. This could cause "internal
265        error: code overflow" or "double free or corruption" or other memory
266        handling errors.
267    
268    11. When named subpatterns had the same prefixes, back references could be
269        confused. For example, in this pattern:
270    
271          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
272    
273        the reference to 'Name' was incorrectly treated as a reference to a
274        duplicate name.
275    
276    12. A pattern such as /^s?c/mi8 where the optional character has more than
277        one "other case" was incorrectly compiled such that it would only try to
278        match starting at "c".
279    
280    13. When a pattern starting with \s was studied, VT was not included in the
281        list of possible starting characters; this should have been part of the
282        8.34/18 patch.
283    
284    14. If a character class started [\Qx]... where x is any character, the class
285        was incorrectly terminated at the ].
286    
287    15. If a pattern that started with a caseless match for a character with more
288        than one "other case" was studied, PCRE did not set up the starting code
289        unit bit map for the list of possible characters. Now it does. This is an
290        optimization improvement, not a bug fix.
291    
292    16. The Unicode data tables have been updated to Unicode 7.0.0.
293    
294    17. Fixed a number of memory leaks in pcregrep.
295    
296    18. Avoid a compiler warning (from some compilers) for a function call with
297        a cast that removes "const" from an lvalue by using an intermediate
298        variable (to which the compiler does not object).
299    
300    19. Incorrect code was compiled if a group that contained an internal recursive
301        back reference was optional (had quantifier with a minimum of zero). This
302        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
303        caused segmentation faults because of stack overflows at compile time.
304    
305    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
306        group that is quantified with an indefinite repeat, caused a compile-time
307        loop which used up all the system stack and provoked a segmentation fault.
308        This was not the same bug as 19 above.
309    
310    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
311        Patch by Mike Frysinger.
312    
313    
314    Version 8.35 04-April-2014
315    --------------------------
316    
317    1.  A new flag is set, when property checks are present in an XCLASS.
318        When this flag is not set, PCRE can perform certain optimizations
319        such as studying these XCLASS-es.
320    
321    2.  The auto-possessification of character sets were improved: a normal
322        and an extended character set can be compared now. Furthermore
323        the JIT compiler optimizes more character set checks.
324    
325    3.  Got rid of some compiler warnings for potentially uninitialized variables
326        that show up only when compiled with -O2.
327    
328    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
329        of a match later then the end of the match. The pcretest program was not
330        handling the case sensibly - it was outputting from the start to the next
331        binary zero. It now reports this situation in a message, and outputs the
332        text from the end to the start.
333    
334    5.  Fast forward search is improved in JIT. Instead of the first three
335        characters, any three characters with fixed position can be searched.
336        Search order: first, last, middle.
337    
338    6.  Improve character range checks in JIT. Characters are read by an inprecise
339        function now, which returns with an unknown value if the character code is
340        above a certain threshold (e.g: 256). The only limitation is that the value
341        must be bigger than the threshold as well. This function is useful when
342        the characters above the threshold are handled in the same way.
343    
344    7.  The macros whose names start with RAWUCHAR are placeholders for a future
345        mode in which only the bottom 21 bits of 32-bit data items are used. To
346        make this more memorable for those maintaining the code, the names have
347        been changed to start with UCHAR21, and an extensive comment has been added
348        to their definition.
349    
350    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
351        to the export list in Makefile.am (they were accidentally omitted from the
352        8.34 tarball).
353    
354    9.  The informational output from pcretest used the phrase "starting byte set"
355        which is inappropriate for the 16-bit and 32-bit libraries. As the output
356        for "first char" and "need char" really means "non-UTF-char", I've changed
357        "byte" to "char", and slightly reworded the output. The documentation about
358        these values has also been (I hope) clarified.
359    
360    10. Another JIT related optimization: use table jumps for selecting the correct
361        backtracking path, when more than four alternatives are present inside a
362        bracket.
363    
364    11. Empty match is not possible, when the minimum length is greater than zero,
365        and there is no \K in the pattern. JIT should avoid empty match checks in
366        such cases.
367    
368    12. In a caseless character class with UCP support, when a character with more
369        than one alternative case was not the first character of a range, not all
370        the alternative cases were added to the class. For example, s and \x{17f}
371        are both alternative cases for S: the class [RST] was handled correctly,
372        but [R-T] was not.
373    
374    13. The configure.ac file always checked for pthread support when JIT was
375        enabled. This is not used in Windows, so I have put this test inside a
376        check for the presence of windows.h (which was already tested for).
377    
378    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
379        The algorithm provides a way to skip certain starting offsets, and usually
380        faster than linear prefix searches.
381    
382    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
383        as for 'fr_FR' and 'french'. For some reason, however, it then used the
384        Windows-specific input and output files, which have 'french' screwed in.
385        So this could never have worked. One of the problems with locales is that
386        they aren't always the same. I have now updated RunTest so that it checks
387        the output of the locale test (test 3) against three different output
388        files, and it allows the test to pass if any one of them matches. With luck
389        this should make the test pass on some versions of Solaris where it was
390        failing. Because of the uncertainty, the script did not used to stop if
391        test 3 failed; it now does. If further versions of a French locale ever
392        come to light, they can now easily be added.
393    
394    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
395        there was a message during ./configure, but it did not stop. This now
396        provokes an error. The invalid example in README has been corrected.
397        If a value less than the minimum is given, the minimum value has always
398        been used, but now a warning is given.
399    
400    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
401        was a bug in the test system, which is now fixed. Also, the list of various
402        configurations that are tested for each release did not have one with both
403        16/32 bits and --enable-bar-anycrlf. It now does.
404    
405    18. pcretest was missing "-C bsr" for displaying the \R default setting.
406    
407    19. Little endian PowerPC systems are supported now by the JIT compiler.
408    
409    20. The fast forward newline mechanism could enter to an infinite loop on
410        certain invalid UTF-8 input. Although we don't support these cases
411        this issue can be fixed by a performance optimization.
412    
413    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
414        not take account if existing stack usage. There is now a new global
415        variable called pcre_stack_guard that can be set to point to an external
416        function to check stack availability. It is called at the start of
417        processing every parenthesized group.
418    
419    22. A typo in the code meant that in ungreedy mode the max/min qualifier
420        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
421        not match "ab".
422    
423    23. When UTF was disabled, the JIT program reported some incorrect compile
424        errors. These messages are silenced now.
425    
426    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
427        compiler.
428    
429    25. Change all the temporary files used in RunGrepTest to be different to those
430        used by RunTest so that the tests can be run simultaneously, for example by
431        "make -j check".
432    
433    
434    Version 8.34 15-December-2013
435    -----------------------------
436    
437    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
438        executable memory. Patch inspired by Carsten Klein.
439    
440    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
441        this macro is never tested and has no effect, because the work to support
442        coverage involves only compiling and linking options and special targets in
443        the Makefile. The comment in config.h implied that defining the macro would
444        enable coverage support, which is totally false. There was also support for
445        setting this macro in the CMake files (my fault, I just copied it from
446        configure). SUPPORT_GCOV has now been removed.
447    
448    3.  Make a small performance improvement in strlen16() and strlen32() in
449        pcretest.
450    
451    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
452        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
453        these cases). There was also one in pcretest.c.
454    
455    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
456    
457    6.  In UTF mode, the code for checking whether a group could match an empty
458        string (which is used for indefinitely repeated groups to allow for
459        breaking an infinite loop) was broken when the group contained a repeated
460        negated single-character class with a character that occupied more than one
461        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
462        UTF-8 mode). The effect was undefined: the group might or might not be
463        deemed as matching an empty string, or the program might have crashed.
464    
465    7.  The code for checking whether a group could match an empty string was not
466        recognizing that \h, \H, \v, \V, and \R must match a character.
467    
468    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
469        an empty string. If it can, pcretest shows this in its information output.
470    
471    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
472        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
473        matched by pcre_exec() without using JIT:
474    
475        (a) If the rest of the pattern did not match after a maximal run of
476            grapheme clusters, the code for backing up to try with fewer of them
477            did not always back up over a full grapheme when characters that do not
478            have the modifier quality were involved, e.g. Hangul syllables.
479    
480        (b) If the match point in a subject started with modifier character, and
481            there was no match, the code could incorrectly back up beyond the match
482            point, and potentially beyond the first character in the subject,
483            leading to a segfault or an incorrect match result.
484    
485    10. A conditional group with an assertion condition could lead to PCRE
486        recording an incorrect first data item for a match if no other first data
487        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
488        first data item, and therefore matched "ca" after "c" instead of at the
489        start.
490    
491    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
492        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
493    
494    12. The source of pcregrep now includes z/OS-specific code so that it can be
495        compiled for z/OS as part of the special z/OS distribution.
496    
497    13. Added the -T and -TM options to pcretest.
498    
499    14. The code in pcre_compile.c for creating the table of named capturing groups
500        has been refactored. Instead of creating the table dynamically during the
501        actual compiling pass, the information is remembered during the pre-compile
502        pass (on the stack unless there are more than 20 named groups, in which
503        case malloc() is used) and the whole table is created before the actual
504        compile happens. This has simplified the code (it is now nearly 150 lines
505        shorter) and prepared the way for better handling of references to groups
506        with duplicate names.
507    
508    15. A back reference to a named subpattern when there is more than one of the
509        same name now checks them in the order in which they appear in the pattern.
510        The first one that is set is used for the reference. Previously only the
511        first one was inspected. This change makes PCRE more compatible with Perl.
512    
513    16. Unicode character properties were updated from Unicode 6.3.0.
514    
515    17. The compile-time code for auto-possessification has been refactored, based
516        on a patch by Zoltan Herczeg. It now happens after instead of during
517        compilation. The code is cleaner, and more cases are handled. The option
518        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
519        options in pcretest are provided to set it. It can also be set by
520        (*NO_AUTO_POSSESS) at the start of a pattern.
521    
522    18. The character VT has been added to the default ("C" locale) set of
523        characters that match \s and are generally treated as white space,
524        following this same change in Perl 5.18. There is now no difference between
525        "Perl space" and "POSIX space". Whether VT is treated as white space in
526        other locales depends on the locale.
527    
528    19. The code for checking named groups as conditions, either for being set or
529        for being recursed, has been refactored (this is related to 14 and 15
530        above). Processing unduplicated named groups should now be as fast at
531        numerical groups, and processing duplicated groups should be faster than
532        before.
533    
534    20. Two patches to the CMake build system, by Alexander Barkov:
535    
536          (1) Replace the "source" command by "." in CMakeLists.txt because
537              "source" is a bash-ism.
538    
539          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
540              without these the CMake build does not work on Solaris.
541    
542    21. Perl has changed its handling of \8 and \9. If there is no previously
543        encountered capturing group of those numbers, they are treated as the
544        literal characters 8 and 9 instead of a binary zero followed by the
545        literals. PCRE now does the same.
546    
547    22. Following Perl, added \o{} to specify codepoints in octal, making it
548        possible to specify values greater than 0777 and also making them
549        unambiguous.
550    
551    23. Perl now gives an error for missing closing braces after \x{... instead of
552        treating the string as literal. PCRE now does the same.
553    
554    24. RunTest used to grumble if an inappropriate test was selected explicitly,
555        but just skip it when running all tests. This make it awkward to run ranges
556        of tests when one of them was inappropriate. Now it just skips any
557        inappropriate tests, as it always did when running all tests.
558    
559    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
560        character types such as \d or \w, too many callouts were inserted, and the
561        data that they returned was rubbish.
562    
563    26. In UCP mode, \s was not matching two of the characters that Perl matches,
564        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
565        were matched by \h. The code has now been refactored so that the lists of
566        the horizontal and vertical whitespace characters used for \h and \v (which
567        are defined only in one place) are now also used for \s.
568    
569    27. Add JIT support for the 64 bit TileGX architecture.
570        Patch by Jiong Wang (Tilera Corporation).
571    
572    28. Possessive quantifiers for classes (both explicit and automatically
573        generated) now use special opcodes instead of wrapping in ONCE brackets.
574    
575    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
576        (because it's meaningless), this was not happening when PCRE_CASELESS was
577        set. Not wrong, but inefficient.
578    
579    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
580        properties for \w, \d, etc) is present in a test regex. Otherwise if the
581        test contains no characters greater than 255, Perl doesn't realise it
582        should be using Unicode semantics.
583    
584    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
585        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
586        does in Unicode mode.
587    
588    32. Added the "forbid" facility to pcretest so that putting tests into the
589        wrong test files can sometimes be quickly detected.
590    
591    33. There is now a limit (default 250) on the depth of nesting of parentheses.
592        This limit is imposed to control the amount of system stack used at compile
593        time. It can be changed at build time by --with-parens-nest-limit=xxx or
594        the equivalent in CMake.
595    
596    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
597        errors. Perl warns for these when in warning mode, but PCRE has no facility
598        for giving warnings.
599    
600    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
601        However, this was not working for (?!) because it is optimized to (*FAIL),
602        for which PCRE does not allow quantifiers. The optimization is now disabled
603        when a quantifier follows (?!). I can't see any use for this, but it makes
604        things uniform.
605    
606    36. Perl no longer allows group names to start with digits, so I have made this
607        change also in PCRE. It simplifies the code a bit.
608    
609    37. In extended mode, Perl ignores spaces before a + that indicates a
610        possessive quantifier. PCRE allowed a space before the quantifier, but not
611        before the possessive +. It now does.
612    
613    38. The use of \K (reset reported match start) within a repeated possessive
614        group such as (a\Kb)*+ was not working.
615    
616    40. Document that the same character tables must be used at compile time and
617        run time, and that the facility to pass tables to pcre_exec() and
618        pcre_dfa_exec() is for use only with saved/restored patterns.
619    
620    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
621        features for Builds with MSVC:
622    
623        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
624           builds). The .rc files can be used to set FileDescription and many other
625           attributes.
626    
627        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
628           This allows higher-level build scripts which want .pdb files to avoid
629           hard-coding the exact files needed."
630    
631    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
632        mean "start of word" and "end of word", respectively, as a transition aid.
633    
634    43. A minimizing repeat of a class containing codepoints greater than 255 in
635        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
636        compiled to use the heap for recursion.
637    
638    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
639        is configured.
640    
641    
642    Version 8.33 28-May-2013
643    ------------------------
644    
645    1.  Added 'U' to some constants that are compared to unsigned integers, to
646        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
647        variables that are added to signed variables, to ensure the result is
648        signed and can be negated.
649    
650    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
651        CMake config files.
652    
653    3.  Revise the creation of config.h.generic so that all boolean macros are
654        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
655        overriding via -D on the command line possible.
656    
657    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
658        to unsigned int is reported to make a quite noticeable speed difference in
659        a specific Windows environment. Testing on Linux did also appear to show
660        some benefit (and it is clearly not harmful). Also fixed the definition of
661        Xop which should be unsigned.
662    
663    5.  Related to (4), changing the definition of the intermediate variable cc
664        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
665        improvements.
666    
667    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
668        unnecessary spaces.
669    
670    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
671        and later.
672    
673    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
674    
675    9.  Optimizing fast_forward_start_bits in JIT.
676    
677    10. Adding support for callouts in JIT, and fixing some issues revealed
678        during this work. Namely:
679    
680        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
681    
682        (b) Minimum length was not checked before the matching is started.
683    
684    11. The value of capture_last that is passed to callouts was incorrect in some
685        cases when there was a capture on one path that was subsequently abandoned
686        after a backtrack. Also, the capture_last value is now reset after a
687        recursion, since all captures are also reset in this case.
688    
689    12. The interpreter no longer returns the "too many substrings" error in the
690        case when an overflowing capture is in a branch that is subsequently
691        abandoned after a backtrack.
692    
693    13. In the pathological case when an offset vector of size 2 is used, pcretest
694        now prints out the matched string after a yield of 0 or 1.
695    
696    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
697        Only supported by the JIT compiler at the moment.
698    
699    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
700    
701    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
702        offset of the starting point of the matching process, provided the offsets
703        vector is large enough.
704    
705    17. The \A escape now records a lookbehind value of 1, though its execution
706        does not actually inspect the previous character. This is to ensure that,
707        in partial multi-segment matching, at least one character from the old
708        segment is retained when a new segment is processed. Otherwise, if there
709        are no lookbehinds in the pattern, \A might match incorrectly at the start
710        of a new segment.
711    
712    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
713    
714    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
715        gives some modest performance improvement in 8-bit mode.
716    
717    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
718        be expressed in certain programming languages using Universal Character
719        Names.
720    
721    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
722        which points out that "non characters" are not "characters that may not
723        appear in Unicode strings" but rather "characters that are reserved for
724        internal use and have only local meaning".
725    
726    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
727        there was a conditional group that depended on an assertion, if the
728        assertion was false, the callout that immediately followed the alternation
729        in the condition was skipped when pcre_exec() was used for matching.
730    
731    23. Allow an explicit callout to be inserted before an assertion that is the
732        condition for a conditional group, for compatibility with automatic
733        callouts, which always insert a callout at this point.
734    
735    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
736        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
737    
738    25. (*PRUNE) is now supported by the JIT compiler.
739    
740    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
741    
742    27. Fix the case where there are two or more SKIPs with arguments that may be
743        ignored.
744    
745    28. (*SKIP) is now supported by the JIT compiler.
746    
747    29. (*THEN) is now supported by the JIT compiler.
748    
749    30. Update RunTest with additional test selector options.
750    
751    31. The way PCRE handles backtracking verbs has been changed in two ways.
752    
753        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
754        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
755        backtracking. In some cases this makes it more Perl-compatible, but Perl's
756        rather obscure rules do not always do the same thing.
757    
758        (2) Previously, backtracking verbs were confined within assertions. This is
759        no longer the case for positive assertions, except for (*ACCEPT). Again,
760        this sometimes improves Perl compatibility, and sometimes does not.
761    
762    32. A number of tests that were in test 2 because Perl did things differently
763        have been moved to test 1, because either Perl or PCRE has changed, and
764        these tests are now compatible.
765    
766    32. Backtracking control verbs are now handled in the same way in JIT and
767        interpreter.
768    
769    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
770        contained a forward subroutine reference caused a compile error.
771    
772    34. Auto-detect and optimize limited repetitions in JIT.
773    
774    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
775        blocking (*UTF) etc.
776    
777    36. In the interpreter, maximizing pattern repetitions for characters and
778        character types now use tail recursion, which reduces stack usage.
779    
780    37. The value of the max lookbehind was not correctly preserved if a compiled
781        and saved regex was reloaded on a host of different endianness.
782    
783    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
784        of the compiled pattern block, expand the flags field from 16 to 32 bits
785        because it was almost full.
786    
787    39. Try madvise first before posix_madvise.
788    
789    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
790        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
791        original change locked out all matches of empty strings. This has been
792        changed so that one match of an empty string per line is recognized.
793        Subsequent searches on the same line (for colouring or for --only-matching,
794        for example) do not recognize empty strings.
795    
796    41. Applied a user patch to fix a number of spelling mistakes in comments.
797    
798    42. Data lines longer than 65536 caused pcretest to crash.
799    
800    43. Clarified the data type for length and startoffset arguments for pcre_exec
801        and pcre_dfa_exec in the function-specific man pages, where they were
802        explicitly stated to be in bytes, never having been updated. I also added
803        some clarification to the pcreapi man page.
804    
805    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
806        a segmentation fault.
807    
808    
809    Version 8.32 30-November-2012
810    -----------------------------
811    
812    1.  Improved JIT compiler optimizations for first character search and single
813        character iterators.
814    
815    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
816        Patch by Daniel Richard G.
817    
818    3.  Single character iterator optimizations in the JIT compiler.
819    
820    4.  Improved JIT compiler optimizations for character ranges.
821    
822    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
823        Reported by Giuseppe D'Angelo.
824    
825    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
826        of a line, was being set incorrectly in cases where .* appeared inside
827        atomic brackets at the start of a pattern, or where there was a subsequent
828        *PRUNE or *SKIP.
829    
830    7.  Improved instruction cache flush for POWER/PowerPC.
831        Patch by Daniel Richard G.
832    
833    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
834        grep:
835    
836        (a) There is now no limit to the number of patterns to be matched.
837    
838        (b) An error is given if a pattern is too long.
839    
840        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
841            are now supported.
842    
843        (d) --exclude-from and --include-from (multiple use) have been added.
844    
845        (e) Exclusions and inclusions now apply to all files and directories, not
846            just to those obtained from scanning a directory recursively.
847    
848        (f) Multiple uses of -f and --file-list are now supported.
849    
850        (g) In a Windows environment, the default for -d has been changed from
851            "read" (the GNU grep default) to "skip", because otherwise the presence
852            of a directory in the file list provokes an error.
853    
854        (h) The documentation has been revised and clarified in places.
855    
856    9.  Improve the matching speed of capturing brackets.
857    
858    10. Changed the meaning of \X so that it now matches a Unicode extended
859        grapheme cluster.
860    
861    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
862        out POSIX threads when JIT support is configured.
863    
864    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
865    
866    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
867        structure could go wrong in environments where size_t is not the same size
868        as int.
869    
870    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
871        set.
872    
873    15. The EBCDIC support had decayed; later updates to the code had included
874        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
875        general tidy up of EBCDIC-related issues, and the documentation was also
876        not quite right. There is now a test that can be run on ASCII systems to
877        check some of the EBCDIC-related things (but is it not a full test).
878    
879    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
880        in a small tidy to the code.
881    
882    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
883    
884    18. If the --only-matching (-o) option in pcregrep is specified multiple
885        times, each one causes appropriate output. For example, -o1 -o2 outputs the
886        substrings matched by the 1st and 2nd capturing parentheses. A separating
887        string can be specified by --om-separator (default empty).
888    
889    19. Improving the first n character searches.
890    
891    20. Turn case lists for horizontal and vertical white space into macros so that
892        they are defined only once.
893    
894    21. This set of changes together give more compatible Unicode case-folding
895        behaviour for characters that have more than one other case when UCP
896        support is available.
897    
898        (a) The Unicode property table now has offsets into a new table of sets of
899            three or more characters that are case-equivalent. The MultiStage2.py
900            script that generates these tables (the pcre_ucd.c file) now scans
901            CaseFolding.txt instead of UnicodeData.txt for character case
902            information.
903    
904        (b) The code for adding characters or ranges of characters to a character
905            class has been abstracted into a generalized function that also handles
906            case-independence. In UTF-mode with UCP support, this uses the new data
907            to handle characters with more than one other case.
908    
909        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
910            whose other case is greater than 256 are now correctly matched
911            caselessly. Previously, the high codepoint matched the low one, but not
912            vice versa.
913    
914        (d) The processing of \h, \H, \v, and \ in character classes now makes use
915            of the new class addition function, using character lists defined as
916            macros alongside the case definitions of 20 above.
917    
918        (e) Caseless back references now work with characters that have more than
919            one other case.
920    
921        (f) General caseless matching of characters with more than one other case
922            is supported.
923    
924    22. Unicode character properties were updated from Unicode 6.2.0
925    
926    23. Improved CMake support under Windows. Patch by Daniel Richard G.
927    
928    24. Add support for 32-bit character strings, and UTF-32
929    
930    25. Major JIT compiler update (code refactoring and bugfixing).
931        Experimental Sparc 32 support is added.
932    
933    26. Applied a modified version of Daniel Richard G's patch to create
934        pcre.h.generic and config.h.generic by "make" instead of in the
935        PrepareRelease script.
936    
937    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
938        pcre_compile.c when checking for a zero character.
939    
940    28. Introducing a native interface for JIT. Through this interface, the compiled
941        machine code can be directly executed. The purpose of this interface is to
942        provide fast pattern matching, so several sanity checks are not performed.
943        However, feature tests are still performed. The new interface provides
944        1.4x speedup compared to the old one.
945    
946    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
947        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
948        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
949    
950    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
951        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
952        the "old" RFC 2279). Instead, it ended up passing a negative length to
953        pcre_exec().
954    
955    31. Add support for GCC's visibility feature to hide internal functions.
956    
957    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
958        "unknown -C option" after outputting 0 or 1.
959    
960    33. There is now support for generating a code coverage report for the test
961        suite in environments where gcc is the compiler and lcov is installed. This
962        is mainly for the benefit of the developers.
963    
964    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
965        unaddressable using valgrind annotations, allowing valgrind to detect
966        invalid memory accesses. This is mainly for the benefit of the developers.
967    
968    25. (*UTF) can now be used to start a pattern in any of the three libraries.
969    
970    26. Give configure error if --enable-cpp but no C++ compiler found.
971    
972    
973    Version 8.31 06-July-2012
974    -------------------------
975    
976    1.  Fixing a wrong JIT test case and some compiler warnings.
977    
978    2.  Removed a bashism from the RunTest script.
979    
980    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
981        to unsigned type, result still unsigned" that was given by an MS compiler
982        on encountering the code "-sizeof(xxx)".
983    
984    4.  Partial matching support is added to the JIT compiler.
985    
986    5.  Fixed several bugs concerned with partial matching of items that consist
987        of more than one character:
988    
989        (a) /^(..)\1/ did not partially match "aba" because checking references was
990            done on an "all or nothing" basis. This also applied to repeated
991            references.
992    
993        (b) \R did not give a hard partial match if \r was found at the end of the
994            subject.
995    
996        (c) \X did not give a hard partial match after matching one or more
997            characters at the end of the subject.
998    
999        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1000            a partial match for the string "\r".
1001    
1002        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1003            a partial match for a CR character at the end of the subject string.
1004    
1005    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1006        running pcretest, the text "(JIT)" added to the output whenever JIT is
1007        actually used to run the match.
1008    
1009    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1010        or /S+[+] with a digit between 1 and 7.
1011    
1012    8.  OP_NOT now supports any UTF character not just single-byte ones.
1013    
1014    9.  (*MARK) control verb is now supported by the JIT compiler.
1015    
1016    10. The command "./RunTest list" lists the available tests without actually
1017        running any of them. (Because I keep forgetting what they all are.)
1018    
1019    11. Add PCRE_INFO_MAXLOOKBEHIND.
1020    
1021    12. Applied a (slightly modified) user-supplied patch that improves performance
1022        when the heap is used for recursion (compiled with --disable-stack-for-
1023        recursion). Instead of malloc and free for each heap frame each time a
1024        logical recursion happens, frames are retained on a chain and re-used where
1025        possible. This sometimes gives as much as 30% improvement.
1026    
1027    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1028        call.
1029    
1030    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1031    
1032    15. It is now possible to link pcretest with libedit as an alternative to
1033        libreadline.
1034    
1035    16. (*COMMIT) control verb is now supported by the JIT compiler.
1036    
1037    17. The Unicode data tables have been updated to Unicode 6.1.0.
1038    
1039    18. Added --file-list option to pcregrep.
1040    
1041    19. Added binary file support to pcregrep, including the -a, --binary-files,
1042        -I, and --text options.
1043    
1044    20. The madvise function is renamed for posix_madvise for QNX compatibility
1045        reasons. Fixed by Giuseppe D'Angelo.
1046    
1047    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1048    
1049    22. Changed the option for creating symbolic links for 16-bit man pages from
1050        -s to -sf so that re-installing does not cause issues.
1051    
1052    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1053    
1054    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1055        matches in certain environments (the workspace was not being correctly
1056        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1057        some of the workspace data at the beginning of a restart.
1058    
1059    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1060        was not doing so when it should - probably a typo introduced by SVN 528
1061        (change 8.10/14).
1062    
1063    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1064        \w+ when the character tables indicated that \x{c4} was a word character.
1065        There were several related cases, all because the tests for doing a table
1066        lookup were testing for characters less than 127 instead of 255.
1067    
1068    27. If a pattern contains capturing parentheses that are not used in a match,
1069        their slots in the ovector are set to -1. For those that are higher than
1070        any matched groups, this happens at the end of processing. In the case when
1071        there were back references that the ovector was too small to contain
1072        (causing temporary malloc'd memory to be used during matching), and the
1073        highest capturing number was not used, memory off the end of the ovector
1074        was incorrectly being set to -1. (It was using the size of the temporary
1075        memory instead of the true size.)
1076    
1077    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1078        ovector size, it uses memory at the end of the block that it has got.
1079    
1080    29. Check for an overlong MARK name and give an error at compile time. The
1081        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1082    
1083    30. JIT compiler update.
1084    
1085    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1086        Rill for the patch.
1087    
1088    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1089    
1090    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1091    
1092    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1093        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1094        was enabled.
1095    
1096    35. Improve JIT code generation for greedy plus quantifier.
1097    
1098    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1099        1 to "aa" instead of to an empty string. The bug affected repeated groups
1100        that could potentially match an empty string.
1101    
1102    37. Optimizing single character iterators in JIT.
1103    
1104    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1105        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1106        codepoints that are too big for the mode are faulted, and in a UTF mode,
1107        disallowed codepoints are also faulted.
1108    
1109    39. If PCRE was compiled with UTF support, in three places in the DFA
1110        matcher there was code that should only have been obeyed in UTF mode, but
1111        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1112        processing when bytes with values greater than 127 were present. In 16-bit
1113        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1114        both cases the values are those that cannot be the first data item in a UTF
1115        character. The three items that might have provoked this were recursions,
1116        possessively repeated groups, and atomic groups.
1117    
1118    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1119        and pcregrep, because some OS require shared objects to be explicitly
1120        passed to ld, causing the link step to fail if they are not.
1121    
1122    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1123        mode, patterns that started with \h* or \R* might be incorrectly matched.
1124    
1125    
1126    Version 8.30 04-February-2012
1127    -----------------------------
1128    
1129    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1130        name is defined in ctype.h.
1131    
1132    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1133        only in quite long subpatterns.
1134    
1135    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1136        since it was replaced by pcre_fullinfo() in February 2000.
1137    
1138    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1139        match a (*MARK), and the match failed at the start of the subject, a
1140        reference to memory before the start of the subject could occur. This bug
1141        was introduced by fix 17 of release 8.21.
1142    
1143    5.  A reference to an unset group with zero minimum repetition was giving
1144        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1145        /(another)?(\1?)test/ matched against "hello world test". This bug was
1146        introduced in release 8.13.
1147    
1148    6.  Add support for 16-bit character strings (a large amount of work involving
1149        many changes and refactorings).
1150    
1151    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1152        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1153        from a file.
1154    
1155    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1156        rounding is not applied in this particular case).
1157    
1158    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1159        if they appear, or are escaped, in patterns.
1160    
1161    10. Get rid of a number of -Wunused-but-set-variable warnings.
1162    
1163    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1164        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1165        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1166        also returns the mark "x". This bug applied to capturing parentheses,
1167        non-capturing parentheses, and atomic parentheses. It also applied to some
1168        assertions.
1169    
1170    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1171        information out of configure.ac instead of relying on pcre.h.generic, which
1172        is not stored in the repository.
1173    
1174    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1175        -lreadline.
1176    
1177    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1178    
1179    15. Applied Graycode's patch to put the top-level frame on the stack rather
1180        than the heap when not using the stack for recursion. This gives a
1181        performance improvement in many cases when recursion is not deep.
1182    
1183    16. Experimental code added to "pcretest -C" to output the stack frame size.
1184    
1185    
1186    Version 8.21 12-Dec-2011
1187    ------------------------
1188    
1189    1.  Updating the JIT compiler.
1190    
1191    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1192        are added as well.
1193    
1194    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1195        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1196        calling _pcre_jit_exec. Some extra comments are added.
1197    
1198    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1199        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1200        was introduced by change 18 for 8.20.
1201    
1202    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1203        ECMA-262 standard.
1204    
1205    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1206        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1207        This bug was probably introduced by change 9 of 8.13.
1208    
1209    7.  While fixing 6 above, I noticed that a number of other items were being
1210        incorrectly rejected as "not fixed length". This arose partly because newer
1211        opcodes had not been added to the fixed-length checking code. I have (a)
1212        corrected the bug and added tests for these items, and (b) arranged for an
1213        error to occur if an unknown opcode is encountered while checking for fixed
1214        length instead of just assuming "not fixed length". The items that were
1215        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1216        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1217        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1218    
1219    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1220        being incorrectly compiled and would have given unpredicatble results.
1221    
1222    9.  A possessively repeated subpattern with minimum repeat count greater than
1223        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1224        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1225        the first (A) could occur when it should not.
1226    
1227    10. Add a cast and remove a redundant test from the code.
1228    
1229    11. JIT should use pcre_malloc/pcre_free for allocation.
1230    
1231    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1232        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1233        is anything other than /usr, -L is still shown).
1234    
1235    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1236    
1237    14. Perl does not support \N without a following name in a [] class; PCRE now
1238        also gives an error.
1239    
1240    15. If a forward reference was repeated with an upper limit of around 2000,
1241        it caused the error "internal error: overran compiling workspace". The
1242        maximum number of forward references (including repeats) was limited by the
1243        internal workspace, and dependent on the LINK_SIZE. The code has been
1244        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1245        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1246        of around 200,000 forward references. While doing this, I also speeded up
1247        the filling in of repeated forward references.
1248    
1249    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1250        incorrectly expecting the subject to contain another "a" after the start.
1251    
1252    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1253        in the match, the SKIP should be ignored. This was not happening; instead
1254        the SKIP was being treated as NOMATCH. For patterns such as
1255        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1256        tested.
1257    
1258    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1259        now much more compatible with Perl, in particular in cases where the result
1260        is a non-match for a non-anchored pattern. For example, if
1261        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1262        "m", where previously it did not return a name. A side effect of this
1263        change is that for partial matches, the last encountered mark name is
1264        returned, as for non matches. A number of tests that were previously not
1265        Perl-compatible have been moved into the Perl-compatible test files. The
1266        refactoring has had the pleasing side effect of removing one argument from
1267        the match() function, thus reducing its stack requirements.
1268    
1269    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1270        subsequent uses of /S (without +) incorrectly behaved like /S+.
1271    
1272    21. Retrieve executable code size support for the JIT compiler and fixing
1273        some warnings.
1274    
1275    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1276        not work when the shorter character appeared right at the end of the
1277        subject string.
1278    
1279    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1280        systems.
1281    
1282    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1283        output it when the /M option is used in pcretest.
1284    
1285    25. The CheckMan script was not being included in the distribution. Also, added
1286        an explicit "perl" to run Perl scripts from the PrepareRelease script
1287        because this is reportedly needed in Windows.
1288    
1289    26. If study data was being save in a file and studying had not found a set of
1290        "starts with" bytes for the pattern, the data written to the file (though
1291        never used) was taken from uninitialized memory and so caused valgrind to
1292        complain.
1293    
1294    27. Updated RunTest.bat as provided by Sheri Pierce.
1295    
1296    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1297    
1298    29. Computation of memory usage for the table of capturing group names was
1299        giving an unnecessarily large value.
1300    
1301    
1302    Version 8.20 21-Oct-2011
1303    ------------------------
1304    
1305    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1306        a POSIX class. After further experiments with Perl, which convinced me that
1307        Perl has bugs and confusions, a closing square bracket is no longer allowed
1308        in a POSIX name. This bug also affected patterns with classes that started
1309        with full stops.
1310    
1311    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1312        captured substring, but while checking the failing first alternative,
1313        substring 1 is temporarily captured. If the output vector supplied to
1314        pcre_exec() was not big enough for this capture, the yield of the function
1315        was still zero ("insufficient space for captured substrings"). This cannot
1316        be totally fixed without adding another stack variable, which seems a lot
1317        of expense for a edge case. However, I have improved the situation in cases
1318        such as /(a)(b)x|abc/ matched against "abc", where the return code
1319        indicates that fewer than the maximum number of slots in the ovector have
1320        been set.
1321    
1322    3.  Related to (2) above: when there are more back references in a pattern than
1323        slots in the output vector, pcre_exec() uses temporary memory during
1324        matching, and copies in the captures as far as possible afterwards. It was
1325        using the entire output vector, but this conflicts with the specification
1326        that only 2/3 is used for passing back captured substrings. Now it uses
1327        only the first 2/3, for compatibility. This is, of course, another edge
1328        case.
1329    
1330    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1331        main code base, and can be used by building with --enable-jit. When this is
1332        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1333        runtime --no-jit option is given.
1334    
1335    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1336        ovector, the return from the function was zero, implying that there were
1337        other matches that did not fit. The correct "exactly full" value is now
1338        returned.
1339    
1340    6.  If a subpattern that was called recursively or as a subroutine contained
1341        (*PRUNE) or any other control that caused it to give a non-standard return,
1342        invalid errors such as "Error -26 (nested recursion at the same subject
1343        position)" or even infinite loops could occur.
1344    
1345    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1346        computing the minimum length on reaching *ACCEPT, and so ended up with the
1347        wrong value of 1 rather than 0. Further investigation indicates that
1348        computing a minimum subject length in the presence of *ACCEPT is difficult
1349        (think back references, subroutine calls), and so I have changed the code
1350        so that no minimum is registered for a pattern that contains *ACCEPT.
1351    
1352    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1353        it was not handled as intended. [But see 16 below.]
1354    
1355    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1356        Sheri Pierce.
1357    
1358    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1359        the first byte in a match must be "a".
1360    
1361    11. Change 17 for 8.13 increased the recursion depth for patterns like
1362        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1363        pattern contains any instances of (*THEN). If it does not, the old
1364        optimizations are restored. It would be nice to do this on a per-group
1365        basis, but at the moment that is not feasible.
1366    
1367    12. In some environments, the output of pcretest -C is CRLF terminated. This
1368        broke RunTest's code that checks for the link size. A single white space
1369        character after the value is now allowed for.
1370    
1371    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1372        For "fr", it uses the Windows-specific input and output files.
1373    
1374    14. If (*THEN) appeared in a group that was called recursively or as a
1375        subroutine, it did not work as intended. [But see next item.]
1376    
1377    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1378        pattern fragments (but not containing any | characters). If A and B are
1379        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1380        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1381        D. In other words, Perl considers parentheses that do not contain any |
1382        characters to be part of a surrounding alternative, whereas PCRE was
1383        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1384        differently. PCRE now behaves in the same way as Perl, except in the case
1385        of subroutine/recursion calls such as (?1) which have in any case always
1386        been different (but PCRE had them first :-).
1387    
1388    16. Related to 15 above: Perl does not treat the | in a conditional group as
1389        creating alternatives. Such a group is treated in the same way as an
1390        ordinary group without any | characters when processing (*THEN). PCRE has
1391        been changed to match Perl's behaviour.
1392    
1393    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1394        RunGrepTest script failed.
1395    
1396    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1397        inevitable for groups that contain captures, but it can lead to a lot of
1398        stack use in large patterns. The old behaviour has been restored for atomic
1399        groups that do not contain any capturing parentheses.
1400    
1401    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1402        suppress the check for a minimum subject length at run time. (If it was
1403        given to pcre_exec() or pcre_dfa_exec() it did work.)
1404    
1405    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1406        fail to work when decoding hex characters in data strings in EBCDIC
1407        environments.
1408    
1409    21. It appears that in at least one Mac OS environment, the isxdigit() function
1410        is implemented as a macro that evaluates to its argument more than once,
1411        contravening the C 90 Standard (I haven't checked a later standard). There
1412        was an instance in pcretest which caused it to go wrong when processing
1413        \x{...} escapes in subject strings. The has been rewritten to avoid using
1414        things like p++ in the argument of isxdigit().
1415    
1416    
1417    Version 8.13 16-Aug-2011
1418  ------------------------  ------------------------
1419    
1420  1.  The Unicode data tables have been updated to Unicode 6.0.0.  1.  The Unicode data tables have been updated to Unicode 6.0.0.
# Line 98  Version 8.13 02-Aug-2011 Line 1511  Version 8.13 02-Aug-2011
1511      tail recursion to cut down on stack usage. Unfortunately, now that there is      tail recursion to cut down on stack usage. Unfortunately, now that there is
1512      the possibility of (*THEN) occurring in these branches, tail recursion is      the possibility of (*THEN) occurring in these branches, tail recursion is
1513      no longer possible because the return has to be checked for (*THEN). These      no longer possible because the return has to be checked for (*THEN). These
1514      two optimizations have therefore been removed.      two optimizations have therefore been removed. [But see 8.20/11 above.]
1515    
1516  18. If a pattern containing \R was studied, it was assumed that \R always  18. If a pattern containing \R was studied, it was assumed that \R always
1517      matched two bytes, thus causing the minimum subject length to be      matched two bytes, thus causing the minimum subject length to be
# Line 189  Version 8.13 02-Aug-2011 Line 1602  Version 8.13 02-Aug-2011
1602      For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,      For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1603      unescaped square brackets may also appear as part of class names. For      unescaped square brackets may also appear as part of class names. For
1604      example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves      example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1605      more like Perl.      more like Perl. (But see 8.20/1 above.)
1606    
1607  38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this  38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1608      was because it thought it was \N{name}, which is not supported).      was because it thought it was \N{name}, which is not supported).
# Line 224  Version 8.13 02-Aug-2011 Line 1637  Version 8.13 02-Aug-2011
1637      (a) The default value of the buffer size parameter has been increased from      (a) The default value of the buffer size parameter has been increased from
1638          8K to 20K. (The actual buffer used is three times this size.)          8K to 20K. (The actual buffer used is three times this size.)
1639    
1640      (b) The default can be changed by ./configure --with-pcregrep-bufsiz when      (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1641          PCRE is built.          PCRE is built.
1642    
1643      (c) A --buffer-size=n option has been added to pcregrep, to allow the size      (c) A --buffer-size=n option has been added to pcregrep, to allow the size
# Line 245  Version 8.13 02-Aug-2011 Line 1658  Version 8.13 02-Aug-2011
1658  47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a  47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1659      complete match instead of a partial match. This bug was dependent on both      complete match instead of a partial match. This bug was dependent on both
1660      the PCRE_UTF8 and PCRE_DOTALL options being set.      the PCRE_UTF8 and PCRE_DOTALL options being set.
1661    
1662  48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the  48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1663      starting byte set, because \b was not being ignored.      starting byte set, because \b was not being ignored.
1664    
1665    
1666  Version 8.12 15-Jan-2011  Version 8.12 15-Jan-2011
# Line 900  Version 7.9 11-Apr-09 Line 2313  Version 7.9 11-Apr-09
2313  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
2314      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
2315      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2316      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2317        8.33.]
2318    
2319  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
2320      start or after a newline", because the conditional assertion was not being      start or after a newline", because the conditional assertion was not being
# Line 1143  Version 7.7 07-May-08 Line 2557  Version 7.7 07-May-08
2557      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2558      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2559      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2560      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2561      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2562      was subsequently set up correctly.)      was subsequently set up correctly.)
2563    

Legend:
Removed from v.657  
changed lines
  Added in v.1564

  ViewVC Help
Powered by ViewVC 1.1.5