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

Diff of /code/trunk/ChangeLog

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

revision 600 by ph10, Mon May 9 08:54:11 2011 UTC revision 1613 by ph10, Fri Nov 27 17:41:04 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 8.13 30-Apr-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 8.39 xx-xxxxxx-201x
8    ---------------------------
9    
10    1.  If PCRE_AUTO_CALLOUT was set on a pattern that had a (?# comment between
11        an item and its qualifier (for example, A(?#comment)?B) pcre_compile()
12        misbehaved. This bug was found by the LLVM fuzzer.
13    
14    2.  Similar to the above, if an isolated \E was present between an item and its
15        qualifier when PCRE_AUTO_CALLOUT was set, pcre_compile() misbehaved. This
16        bug was found by the LLVM fuzzer.
17    
18    3.  Further to 8.38/46, negated classes such as [^[:^ascii:]\d] were also not
19        working correctly in UCP mode.
20    
21    
22    Version 8.38 23-November-2015
23    -----------------------------
24    
25    1.  If a group that contained a recursive back reference also contained a
26        forward reference subroutine call followed by a non-forward-reference
27        subroutine call, for example /.((?2)(?R)\1)()/, pcre2_compile() failed to
28        compile correct code, leading to undefined behaviour or an internally
29        detected error. This bug was discovered by the LLVM fuzzer.
30    
31    2.  Quantification of certain items (e.g. atomic back references) could cause
32        incorrect code to be compiled when recursive forward references were
33        involved. For example, in this pattern: /(?1)()((((((\1++))\x85)+)|))/.
34        This bug was discovered by the LLVM fuzzer.
35    
36    3.  A repeated conditional group whose condition was a reference by name caused
37        a buffer overflow if there was more than one group with the given name.
38        This bug was discovered by the LLVM fuzzer.
39    
40    4.  A recursive back reference by name within a group that had the same name as
41        another group caused a buffer overflow. For example:
42        /(?J)(?'d'(?'d'\g{d}))/. This bug was discovered by the LLVM fuzzer.
43    
44    5.  A forward reference by name to a group whose number is the same as the
45        current group, for example in this pattern: /(?|(\k'Pm')|(?'Pm'))/, caused
46        a buffer overflow at compile time. This bug was discovered by the LLVM
47        fuzzer.
48    
49    6.  A lookbehind assertion within a set of mutually recursive subpatterns could
50        provoke a buffer overflow. This bug was discovered by the LLVM fuzzer.
51    
52    7.  Another buffer overflow bug involved duplicate named groups with a
53        reference between their definition, with a group that reset capture
54        numbers, for example: /(?J:(?|(?'R')(\k'R')|((?'R'))))/. This has been
55        fixed by always allowing for more memory, even if not needed. (A proper fix
56        is implemented in PCRE2, but it involves more refactoring.)
57    
58    8.  There was no check for integer overflow in subroutine calls such as (?123).
59    
60    9.  The table entry for \l in EBCDIC environments was incorrect, leading to its
61        being treated as a literal 'l' instead of causing an error.
62    
63    10. There was a buffer overflow if pcre_exec() was called with an ovector of
64        size 1. This bug was found by american fuzzy lop.
65    
66    11. If a non-capturing group containing a conditional group that could match
67        an empty string was repeated, it was not identified as matching an empty
68        string itself. For example: /^(?:(?(1)x|)+)+$()/.
69    
70    12. In an EBCDIC environment, pcretest was mishandling the escape sequences
71        \a and \e in test subject lines.
72    
73    13. In an EBCDIC environment, \a in a pattern was converted to the ASCII
74        instead of the EBCDIC value.
75    
76    14. The handling of \c in an EBCDIC environment has been revised so that it is
77        now compatible with the specification in Perl's perlebcdic page.
78    
79    15. The EBCDIC character 0x41 is a non-breaking space, equivalent to 0xa0 in
80        ASCII/Unicode. This has now been added to the list of characters that are
81        recognized as white space in EBCDIC.
82    
83    16. When PCRE was compiled without UCP support, the use of \p and \P gave an
84        error (correctly) when used outside a class, but did not give an error
85        within a class.
86    
87    17. \h within a class was incorrectly compiled in EBCDIC environments.
88    
89    18. A pattern with an unmatched closing parenthesis that contained a backward
90        assertion which itself contained a forward reference caused buffer
91        overflow. And example pattern is: /(?=di(?<=(?1))|(?=(.))))/.
92    
93    19. JIT should return with error when the compiled pattern requires more stack
94        space than the maximum.
95    
96    20. A possessively repeated conditional group that could match an empty string,
97        for example, /(?(R))*+/, was incorrectly compiled.
98    
99    21. Fix infinite recursion in the JIT compiler when certain patterns such as
100        /(?:|a|){100}x/ are analysed.
101    
102    22. Some patterns with character classes involving [: and \\ were incorrectly
103        compiled and could cause reading from uninitialized memory or an incorrect
104        error diagnosis.
105    
106    23. Pathological patterns containing many nested occurrences of [: caused
107        pcre_compile() to run for a very long time.
108    
109    24. A conditional group with only one branch has an implicit empty alternative
110        branch and must therefore be treated as potentially matching an empty
111        string.
112    
113    25. If (?R was followed by - or + incorrect behaviour happened instead of a
114        diagnostic.
115    
116    26. Arrange to give up on finding the minimum matching length for overly
117        complex patterns.
118    
119    27. Similar to (4) above: in a pattern with duplicated named groups and an
120        occurrence of (?| it is possible for an apparently non-recursive back
121        reference to become recursive if a later named group with the relevant
122        number is encountered. This could lead to a buffer overflow. Wen Guanxing
123        from Venustech ADLAB discovered this bug.
124    
125    28. If pcregrep was given the -q option with -c or -l, or when handling a
126        binary file, it incorrectly wrote output to stdout.
127    
128    29. The JIT compiler did not restore the control verb head in case of *THEN
129        control verbs. This issue was found by Karl Skomski with a custom LLVM
130        fuzzer.
131    
132    30. Error messages for syntax errors following \g and \k were giving inaccurate
133        offsets in the pattern.
134    
135    31. Added a check for integer overflow in conditions (?(<digits>) and
136        (?(R<digits>). This omission was discovered by Karl Skomski with the LLVM
137        fuzzer.
138    
139    32. Handling recursive references such as (?2) when the reference is to a group
140        later in the pattern uses code that is very hacked about and error-prone.
141        It has been re-written for PCRE2. Here in PCRE1, a check has been added to
142        give an internal error if it is obvious that compiling has gone wrong.
143    
144    33. The JIT compiler should not check repeats after a {0,1} repeat byte code.
145        This issue was found by Karl Skomski with a custom LLVM fuzzer.
146    
147    34. The JIT compiler should restore the control chain for empty possessive
148        repeats. This issue was found by Karl Skomski with a custom LLVM fuzzer.
149    
150    35. Match limit check added to JIT recursion. This issue was found by Karl
151        Skomski with a custom LLVM fuzzer.
152    
153    36. Yet another case similar to 27 above has been circumvented by an
154        unconditional allocation of extra memory. This issue is fixed "properly" in
155        PCRE2 by refactoring the way references are handled. Wen Guanxing
156        from Venustech ADLAB discovered this bug.
157    
158    37. Fix two assertion fails in JIT. These issues were found by Karl Skomski
159        with a custom LLVM fuzzer.
160    
161    38. Fixed a corner case of range optimization in JIT.
162    
163    39. An incorrect error "overran compiling workspace" was given if there were
164        exactly enough group forward references such that the last one extended
165        into the workspace safety margin. The next one would have expanded the
166        workspace. The test for overflow was not including the safety margin.
167    
168    40. A match limit issue is fixed in JIT which was found by Karl Skomski
169        with a custom LLVM fuzzer.
170    
171    41. Remove the use of /dev/null in testdata/testinput2, because it doesn't
172        work under Windows. (Why has it taken so long for anyone to notice?)
173    
174    42. In a character class such as [\W\p{Any}] where both a negative-type escape
175        ("not a word character") and a property escape were present, the property
176        escape was being ignored.
177    
178    43. Fix crash caused by very long (*MARK) or (*THEN) names.
179    
180    44. A sequence such as [[:punct:]b] that is, a POSIX character class followed
181        by a single ASCII character in a class item, was incorrectly compiled in
182        UCP mode. The POSIX class got lost, but only if the single character
183        followed it.
184    
185    45. [:punct:] in UCP mode was matching some characters in the range 128-255
186        that should not have been matched.
187    
188    46. If [:^ascii:] or [:^xdigit:] or [:^cntrl:] are present in a non-negated
189        class, all characters with code points greater than 255 are in the class.
190        When a Unicode property was also in the class (if PCRE_UCP is set, escapes
191        such as \w are turned into Unicode properties), wide characters were not
192        correctly handled, and could fail to match.
193    
194    
195    Version 8.37 28-April-2015
196    --------------------------
197    
198    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
199        for those parentheses to be closed with whatever has been captured so far.
200        However, it was failing to mark any other groups between the hightest
201        capture so far and the currrent group as "unset". Thus, the ovector for
202        those groups contained whatever was previously there. An example is the
203        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
204    
205    2.  If an assertion condition was quantified with a minimum of zero (an odd
206        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
207    
208    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
209        unrecognized modifier, a crash could occur.
210    
211    4.  An attempt to do global matching in pcretest with a zero-length ovector
212        caused a crash.
213    
214    5.  Fixed a memory leak during matching that could occur for a subpattern
215        subroutine call (recursive or otherwise) if the number of captured groups
216        that had to be saved was greater than ten.
217    
218    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
219        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
220        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
221    
222    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
223        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
224    
225    8. When a pattern is compiled, it remembers the highest back reference so that
226       when matching, if the ovector is too small, extra memory can be obtained to
227       use instead. A conditional subpattern whose condition is a check on a
228       capture having happened, such as, for example in the pattern
229       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
230       setting the highest backreference number. This mattered only if pcre_exec()
231       was called with an ovector that was too small to hold the capture, and there
232       was no other kind of back reference (a situation which is probably quite
233       rare). The effect of the bug was that the condition was always treated as
234       FALSE when the capture could not be consulted, leading to a incorrect
235       behaviour by pcre_exec(). This bug has been fixed.
236    
237    9. A reference to a duplicated named group (either a back reference or a test
238       for being set in a conditional) that occurred in a part of the pattern where
239       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
240       to be incorrectly calculated, leading to overwriting.
241    
242    10. A mutually recursive set of back references such as (\2)(\1) caused a
243        segfault at study time (while trying to find the minimum matching length).
244        The infinite loop is now broken (with the minimum length unset, that is,
245        zero).
246    
247    11. If an assertion that was used as a condition was quantified with a minimum
248        of zero, matching went wrong. In particular, if the whole group had
249        unlimited repetition and could match an empty string, a segfault was
250        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
251        assertions to be quantified, but not if they are being used as conditions,
252        so the above pattern is faulted by Perl. PCRE has now been changed so that
253        it also rejects such patterns.
254    
255    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
256        failed to allow the zero-repeat case if pcre2_exec() was called with an
257        ovector too small to capture the group.
258    
259    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
260        Red Hat Product Security:
261    
262        (a) A crash if /K and /F were both set with the option to save the compiled
263        pattern.
264    
265        (b) Another crash if the option to print captured substrings in a callout
266        was combined with setting a null ovector, for example \O\C+ as a subject
267        string.
268    
269    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
270        forward reference repeated a large (but limited) number of times within a
271        repeated outer group that has a zero minimum quantifier, caused incorrect
272        code to be compiled, leading to the error "internal error:
273        previously-checked referenced subpattern not found" when an incorrect
274        memory address was read. This bug was reported as "heap overflow",
275        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
276        CVE-2015-2325.
277    
278    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
279        call within a group that also contained a recursive back reference caused
280        incorrect code to be compiled. This bug was reported as "heap overflow",
281        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
282        number CVE-2015-2326.
283    
284    24. Computing the size of the JIT read-only data in advance has been a source
285        of various issues, and new ones are still appear unfortunately. To fix
286        existing and future issues, size computation is eliminated from the code,
287        and replaced by on-demand memory allocation.
288    
289    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
290        adjacent to the end of the range, and the range contained characters with
291        more than one other case, caused incorrect behaviour when compiled in UTF
292        mode. In that example, the range a-j was left out of the class.
293    
294    26. Fix JIT compilation of conditional blocks, which assertion
295        is converted to (*FAIL). E.g: /(?(?!))/.
296    
297    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
298        discovered by the LLVM fuzzer.
299    
300    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
301        when this assertion was used as a condition, for example (?(?!)a|b). In
302        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
303        error about an unsupported item.
304    
305    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
306        possessification code could take exponential time to complete. A recursion
307        depth limit of 1000 has been imposed to limit the resources used by this
308        optimization.
309    
310    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
311        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
312        because \S ensures they are all in the class. The code for doing this was
313        interacting badly with the code for computing the amount of space needed to
314        compile the pattern, leading to a buffer overflow. This bug was discovered
315        by the LLVM fuzzer.
316    
317    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
318        other kinds of group caused stack overflow at compile time. This bug was
319        discovered by the LLVM fuzzer.
320    
321    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
322        between a subroutine call and its quantifier was incorrectly compiled,
323        leading to buffer overflow or other errors. This bug was discovered by the
324        LLVM fuzzer.
325    
326    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
327        assertion after (?(. The code was failing to check the character after
328        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
329        was discovered by the LLVM fuzzer.
330    
331    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
332        a fixed maximum following a group that contains a subroutine reference was
333        incorrectly compiled and could trigger buffer overflow. This bug was
334        discovered by the LLVM fuzzer.
335    
336    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
337        caused a stack overflow instead of the diagnosis of a non-fixed length
338        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
339    
340    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
341        (e.g. /(?<=\Ka)/) could make pcregrep loop.
342    
343    37. There was a similar problem to 36 in pcretest for global matches.
344    
345    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
346        and a subsequent item in the pattern caused a non-match, backtracking over
347        the repeated \X did not stop, but carried on past the start of the subject,
348        causing reference to random memory and/or a segfault. There were also some
349        other cases where backtracking after \C could crash. This set of bugs was
350        discovered by the LLVM fuzzer.
351    
352    39. The function for finding the minimum length of a matching string could take
353        a very long time if mutual recursion was present many times in a pattern,
354        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
355        method has been implemented. This infelicity was discovered by the LLVM
356        fuzzer.
357    
358    40. Static linking against the PCRE library using the pkg-config module was
359        failing on missing pthread symbols.
360    
361    
362    Version 8.36 26-September-2014
363    ------------------------------
364    
365    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
366        -Wmissing-field-initializers and -Wunused-parameter.
367    
368    2.  The tests for quantifiers being too big (greater than 65535) were being
369        applied after reading the number, and stupidly assuming that integer
370        overflow would give a negative number. The tests are now applied as the
371        numbers are read.
372    
373    3.  Tidy code in pcre_exec.c where two branches that used to be different are
374        now the same.
375    
376    4.  The JIT compiler did not generate match limit checks for certain
377        bracketed expressions with quantifiers. This may lead to exponential
378        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
379        issue should be resolved now.
380    
381    5.  Fixed an issue, which occures when nested alternatives are optimized
382        with table jumps.
383    
384    6.  Inserted two casts and changed some ints to size_t in the light of some
385        reported 64-bit compiler warnings (Bugzilla 1477).
386    
387    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
388        an empty string, which sometimes were behaving incorrectly in the
389        interpreter (though correctly in the JIT matcher). This pcretest input is
390        an example:
391    
392          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
393          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
394    
395        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
396        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
397        for an empty string was breaking the inner loop and carrying on at a lower
398        level, when possessive repeated groups should always return to a higher
399        level as they have no backtrack points in them. The empty string test now
400        occurs at the outer level.
401    
402    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
403        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
404    
405    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
406        doesn't).
407    
408    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
409        to hold a pattern to be incorrectly computed (too small) when there were
410        named back references to duplicated names. This could cause "internal
411        error: code overflow" or "double free or corruption" or other memory
412        handling errors.
413    
414    11. When named subpatterns had the same prefixes, back references could be
415        confused. For example, in this pattern:
416    
417          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
418    
419        the reference to 'Name' was incorrectly treated as a reference to a
420        duplicate name.
421    
422    12. A pattern such as /^s?c/mi8 where the optional character has more than
423        one "other case" was incorrectly compiled such that it would only try to
424        match starting at "c".
425    
426    13. When a pattern starting with \s was studied, VT was not included in the
427        list of possible starting characters; this should have been part of the
428        8.34/18 patch.
429    
430    14. If a character class started [\Qx]... where x is any character, the class
431        was incorrectly terminated at the ].
432    
433    15. If a pattern that started with a caseless match for a character with more
434        than one "other case" was studied, PCRE did not set up the starting code
435        unit bit map for the list of possible characters. Now it does. This is an
436        optimization improvement, not a bug fix.
437    
438    16. The Unicode data tables have been updated to Unicode 7.0.0.
439    
440    17. Fixed a number of memory leaks in pcregrep.
441    
442    18. Avoid a compiler warning (from some compilers) for a function call with
443        a cast that removes "const" from an lvalue by using an intermediate
444        variable (to which the compiler does not object).
445    
446    19. Incorrect code was compiled if a group that contained an internal recursive
447        back reference was optional (had quantifier with a minimum of zero). This
448        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
449        caused segmentation faults because of stack overflows at compile time.
450    
451    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
452        group that is quantified with an indefinite repeat, caused a compile-time
453        loop which used up all the system stack and provoked a segmentation fault.
454        This was not the same bug as 19 above.
455    
456    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
457        Patch by Mike Frysinger.
458    
459    
460    Version 8.35 04-April-2014
461    --------------------------
462    
463    1.  A new flag is set, when property checks are present in an XCLASS.
464        When this flag is not set, PCRE can perform certain optimizations
465        such as studying these XCLASS-es.
466    
467    2.  The auto-possessification of character sets were improved: a normal
468        and an extended character set can be compared now. Furthermore
469        the JIT compiler optimizes more character set checks.
470    
471    3.  Got rid of some compiler warnings for potentially uninitialized variables
472        that show up only when compiled with -O2.
473    
474    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
475        of a match later then the end of the match. The pcretest program was not
476        handling the case sensibly - it was outputting from the start to the next
477        binary zero. It now reports this situation in a message, and outputs the
478        text from the end to the start.
479    
480    5.  Fast forward search is improved in JIT. Instead of the first three
481        characters, any three characters with fixed position can be searched.
482        Search order: first, last, middle.
483    
484    6.  Improve character range checks in JIT. Characters are read by an inprecise
485        function now, which returns with an unknown value if the character code is
486        above a certain threshold (e.g: 256). The only limitation is that the value
487        must be bigger than the threshold as well. This function is useful when
488        the characters above the threshold are handled in the same way.
489    
490    7.  The macros whose names start with RAWUCHAR are placeholders for a future
491        mode in which only the bottom 21 bits of 32-bit data items are used. To
492        make this more memorable for those maintaining the code, the names have
493        been changed to start with UCHAR21, and an extensive comment has been added
494        to their definition.
495    
496    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
497        to the export list in Makefile.am (they were accidentally omitted from the
498        8.34 tarball).
499    
500    9.  The informational output from pcretest used the phrase "starting byte set"
501        which is inappropriate for the 16-bit and 32-bit libraries. As the output
502        for "first char" and "need char" really means "non-UTF-char", I've changed
503        "byte" to "char", and slightly reworded the output. The documentation about
504        these values has also been (I hope) clarified.
505    
506    10. Another JIT related optimization: use table jumps for selecting the correct
507        backtracking path, when more than four alternatives are present inside a
508        bracket.
509    
510    11. Empty match is not possible, when the minimum length is greater than zero,
511        and there is no \K in the pattern. JIT should avoid empty match checks in
512        such cases.
513    
514    12. In a caseless character class with UCP support, when a character with more
515        than one alternative case was not the first character of a range, not all
516        the alternative cases were added to the class. For example, s and \x{17f}
517        are both alternative cases for S: the class [RST] was handled correctly,
518        but [R-T] was not.
519    
520    13. The configure.ac file always checked for pthread support when JIT was
521        enabled. This is not used in Windows, so I have put this test inside a
522        check for the presence of windows.h (which was already tested for).
523    
524    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
525        The algorithm provides a way to skip certain starting offsets, and usually
526        faster than linear prefix searches.
527    
528    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
529        as for 'fr_FR' and 'french'. For some reason, however, it then used the
530        Windows-specific input and output files, which have 'french' screwed in.
531        So this could never have worked. One of the problems with locales is that
532        they aren't always the same. I have now updated RunTest so that it checks
533        the output of the locale test (test 3) against three different output
534        files, and it allows the test to pass if any one of them matches. With luck
535        this should make the test pass on some versions of Solaris where it was
536        failing. Because of the uncertainty, the script did not used to stop if
537        test 3 failed; it now does. If further versions of a French locale ever
538        come to light, they can now easily be added.
539    
540    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
541        there was a message during ./configure, but it did not stop. This now
542        provokes an error. The invalid example in README has been corrected.
543        If a value less than the minimum is given, the minimum value has always
544        been used, but now a warning is given.
545    
546    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
547        was a bug in the test system, which is now fixed. Also, the list of various
548        configurations that are tested for each release did not have one with both
549        16/32 bits and --enable-bar-anycrlf. It now does.
550    
551    18. pcretest was missing "-C bsr" for displaying the \R default setting.
552    
553    19. Little endian PowerPC systems are supported now by the JIT compiler.
554    
555    20. The fast forward newline mechanism could enter to an infinite loop on
556        certain invalid UTF-8 input. Although we don't support these cases
557        this issue can be fixed by a performance optimization.
558    
559    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
560        not take account if existing stack usage. There is now a new global
561        variable called pcre_stack_guard that can be set to point to an external
562        function to check stack availability. It is called at the start of
563        processing every parenthesized group.
564    
565    22. A typo in the code meant that in ungreedy mode the max/min qualifier
566        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
567        not match "ab".
568    
569    23. When UTF was disabled, the JIT program reported some incorrect compile
570        errors. These messages are silenced now.
571    
572    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
573        compiler.
574    
575    25. Change all the temporary files used in RunGrepTest to be different to those
576        used by RunTest so that the tests can be run simultaneously, for example by
577        "make -j check".
578    
579    
580    Version 8.34 15-December-2013
581    -----------------------------
582    
583    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
584        executable memory. Patch inspired by Carsten Klein.
585    
586    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
587        this macro is never tested and has no effect, because the work to support
588        coverage involves only compiling and linking options and special targets in
589        the Makefile. The comment in config.h implied that defining the macro would
590        enable coverage support, which is totally false. There was also support for
591        setting this macro in the CMake files (my fault, I just copied it from
592        configure). SUPPORT_GCOV has now been removed.
593    
594    3.  Make a small performance improvement in strlen16() and strlen32() in
595        pcretest.
596    
597    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
598        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
599        these cases). There was also one in pcretest.c.
600    
601    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
602    
603    6.  In UTF mode, the code for checking whether a group could match an empty
604        string (which is used for indefinitely repeated groups to allow for
605        breaking an infinite loop) was broken when the group contained a repeated
606        negated single-character class with a character that occupied more than one
607        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
608        UTF-8 mode). The effect was undefined: the group might or might not be
609        deemed as matching an empty string, or the program might have crashed.
610    
611    7.  The code for checking whether a group could match an empty string was not
612        recognizing that \h, \H, \v, \V, and \R must match a character.
613    
614    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
615        an empty string. If it can, pcretest shows this in its information output.
616    
617    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
618        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
619        matched by pcre_exec() without using JIT:
620    
621        (a) If the rest of the pattern did not match after a maximal run of
622            grapheme clusters, the code for backing up to try with fewer of them
623            did not always back up over a full grapheme when characters that do not
624            have the modifier quality were involved, e.g. Hangul syllables.
625    
626        (b) If the match point in a subject started with modifier character, and
627            there was no match, the code could incorrectly back up beyond the match
628            point, and potentially beyond the first character in the subject,
629            leading to a segfault or an incorrect match result.
630    
631    10. A conditional group with an assertion condition could lead to PCRE
632        recording an incorrect first data item for a match if no other first data
633        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
634        first data item, and therefore matched "ca" after "c" instead of at the
635        start.
636    
637    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
638        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
639    
640    12. The source of pcregrep now includes z/OS-specific code so that it can be
641        compiled for z/OS as part of the special z/OS distribution.
642    
643    13. Added the -T and -TM options to pcretest.
644    
645    14. The code in pcre_compile.c for creating the table of named capturing groups
646        has been refactored. Instead of creating the table dynamically during the
647        actual compiling pass, the information is remembered during the pre-compile
648        pass (on the stack unless there are more than 20 named groups, in which
649        case malloc() is used) and the whole table is created before the actual
650        compile happens. This has simplified the code (it is now nearly 150 lines
651        shorter) and prepared the way for better handling of references to groups
652        with duplicate names.
653    
654    15. A back reference to a named subpattern when there is more than one of the
655        same name now checks them in the order in which they appear in the pattern.
656        The first one that is set is used for the reference. Previously only the
657        first one was inspected. This change makes PCRE more compatible with Perl.
658    
659    16. Unicode character properties were updated from Unicode 6.3.0.
660    
661    17. The compile-time code for auto-possessification has been refactored, based
662        on a patch by Zoltan Herczeg. It now happens after instead of during
663        compilation. The code is cleaner, and more cases are handled. The option
664        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
665        options in pcretest are provided to set it. It can also be set by
666        (*NO_AUTO_POSSESS) at the start of a pattern.
667    
668    18. The character VT has been added to the default ("C" locale) set of
669        characters that match \s and are generally treated as white space,
670        following this same change in Perl 5.18. There is now no difference between
671        "Perl space" and "POSIX space". Whether VT is treated as white space in
672        other locales depends on the locale.
673    
674    19. The code for checking named groups as conditions, either for being set or
675        for being recursed, has been refactored (this is related to 14 and 15
676        above). Processing unduplicated named groups should now be as fast at
677        numerical groups, and processing duplicated groups should be faster than
678        before.
679    
680    20. Two patches to the CMake build system, by Alexander Barkov:
681    
682          (1) Replace the "source" command by "." in CMakeLists.txt because
683              "source" is a bash-ism.
684    
685          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
686              without these the CMake build does not work on Solaris.
687    
688    21. Perl has changed its handling of \8 and \9. If there is no previously
689        encountered capturing group of those numbers, they are treated as the
690        literal characters 8 and 9 instead of a binary zero followed by the
691        literals. PCRE now does the same.
692    
693    22. Following Perl, added \o{} to specify codepoints in octal, making it
694        possible to specify values greater than 0777 and also making them
695        unambiguous.
696    
697    23. Perl now gives an error for missing closing braces after \x{... instead of
698        treating the string as literal. PCRE now does the same.
699    
700    24. RunTest used to grumble if an inappropriate test was selected explicitly,
701        but just skip it when running all tests. This make it awkward to run ranges
702        of tests when one of them was inappropriate. Now it just skips any
703        inappropriate tests, as it always did when running all tests.
704    
705    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
706        character types such as \d or \w, too many callouts were inserted, and the
707        data that they returned was rubbish.
708    
709    26. In UCP mode, \s was not matching two of the characters that Perl matches,
710        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
711        were matched by \h. The code has now been refactored so that the lists of
712        the horizontal and vertical whitespace characters used for \h and \v (which
713        are defined only in one place) are now also used for \s.
714    
715    27. Add JIT support for the 64 bit TileGX architecture.
716        Patch by Jiong Wang (Tilera Corporation).
717    
718    28. Possessive quantifiers for classes (both explicit and automatically
719        generated) now use special opcodes instead of wrapping in ONCE brackets.
720    
721    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
722        (because it's meaningless), this was not happening when PCRE_CASELESS was
723        set. Not wrong, but inefficient.
724    
725    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
726        properties for \w, \d, etc) is present in a test regex. Otherwise if the
727        test contains no characters greater than 255, Perl doesn't realise it
728        should be using Unicode semantics.
729    
730    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
731        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
732        does in Unicode mode.
733    
734    32. Added the "forbid" facility to pcretest so that putting tests into the
735        wrong test files can sometimes be quickly detected.
736    
737    33. There is now a limit (default 250) on the depth of nesting of parentheses.
738        This limit is imposed to control the amount of system stack used at compile
739        time. It can be changed at build time by --with-parens-nest-limit=xxx or
740        the equivalent in CMake.
741    
742    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
743        errors. Perl warns for these when in warning mode, but PCRE has no facility
744        for giving warnings.
745    
746    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
747        However, this was not working for (?!) because it is optimized to (*FAIL),
748        for which PCRE does not allow quantifiers. The optimization is now disabled
749        when a quantifier follows (?!). I can't see any use for this, but it makes
750        things uniform.
751    
752    36. Perl no longer allows group names to start with digits, so I have made this
753        change also in PCRE. It simplifies the code a bit.
754    
755    37. In extended mode, Perl ignores spaces before a + that indicates a
756        possessive quantifier. PCRE allowed a space before the quantifier, but not
757        before the possessive +. It now does.
758    
759    38. The use of \K (reset reported match start) within a repeated possessive
760        group such as (a\Kb)*+ was not working.
761    
762    40. Document that the same character tables must be used at compile time and
763        run time, and that the facility to pass tables to pcre_exec() and
764        pcre_dfa_exec() is for use only with saved/restored patterns.
765    
766    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
767        features for Builds with MSVC:
768    
769        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
770           builds). The .rc files can be used to set FileDescription and many other
771           attributes.
772    
773        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
774           This allows higher-level build scripts which want .pdb files to avoid
775           hard-coding the exact files needed."
776    
777    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
778        mean "start of word" and "end of word", respectively, as a transition aid.
779    
780    43. A minimizing repeat of a class containing codepoints greater than 255 in
781        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
782        compiled to use the heap for recursion.
783    
784    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
785        is configured.
786    
787    
788    Version 8.33 28-May-2013
789    ------------------------
790    
791    1.  Added 'U' to some constants that are compared to unsigned integers, to
792        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
793        variables that are added to signed variables, to ensure the result is
794        signed and can be negated.
795    
796    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
797        CMake config files.
798    
799    3.  Revise the creation of config.h.generic so that all boolean macros are
800        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
801        overriding via -D on the command line possible.
802    
803    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
804        to unsigned int is reported to make a quite noticeable speed difference in
805        a specific Windows environment. Testing on Linux did also appear to show
806        some benefit (and it is clearly not harmful). Also fixed the definition of
807        Xop which should be unsigned.
808    
809    5.  Related to (4), changing the definition of the intermediate variable cc
810        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
811        improvements.
812    
813    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
814        unnecessary spaces.
815    
816    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
817        and later.
818    
819    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
820    
821    9.  Optimizing fast_forward_start_bits in JIT.
822    
823    10. Adding support for callouts in JIT, and fixing some issues revealed
824        during this work. Namely:
825    
826        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
827    
828        (b) Minimum length was not checked before the matching is started.
829    
830    11. The value of capture_last that is passed to callouts was incorrect in some
831        cases when there was a capture on one path that was subsequently abandoned
832        after a backtrack. Also, the capture_last value is now reset after a
833        recursion, since all captures are also reset in this case.
834    
835    12. The interpreter no longer returns the "too many substrings" error in the
836        case when an overflowing capture is in a branch that is subsequently
837        abandoned after a backtrack.
838    
839    13. In the pathological case when an offset vector of size 2 is used, pcretest
840        now prints out the matched string after a yield of 0 or 1.
841    
842    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
843        Only supported by the JIT compiler at the moment.
844    
845    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
846    
847    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
848        offset of the starting point of the matching process, provided the offsets
849        vector is large enough.
850    
851    17. The \A escape now records a lookbehind value of 1, though its execution
852        does not actually inspect the previous character. This is to ensure that,
853        in partial multi-segment matching, at least one character from the old
854        segment is retained when a new segment is processed. Otherwise, if there
855        are no lookbehinds in the pattern, \A might match incorrectly at the start
856        of a new segment.
857    
858    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
859    
860    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
861        gives some modest performance improvement in 8-bit mode.
862    
863    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
864        be expressed in certain programming languages using Universal Character
865        Names.
866    
867    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
868        which points out that "non characters" are not "characters that may not
869        appear in Unicode strings" but rather "characters that are reserved for
870        internal use and have only local meaning".
871    
872    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
873        there was a conditional group that depended on an assertion, if the
874        assertion was false, the callout that immediately followed the alternation
875        in the condition was skipped when pcre_exec() was used for matching.
876    
877    23. Allow an explicit callout to be inserted before an assertion that is the
878        condition for a conditional group, for compatibility with automatic
879        callouts, which always insert a callout at this point.
880    
881    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
882        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
883    
884    25. (*PRUNE) is now supported by the JIT compiler.
885    
886    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
887    
888    27. Fix the case where there are two or more SKIPs with arguments that may be
889        ignored.
890    
891    28. (*SKIP) is now supported by the JIT compiler.
892    
893    29. (*THEN) is now supported by the JIT compiler.
894    
895    30. Update RunTest with additional test selector options.
896    
897    31. The way PCRE handles backtracking verbs has been changed in two ways.
898    
899        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
900        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
901        backtracking. In some cases this makes it more Perl-compatible, but Perl's
902        rather obscure rules do not always do the same thing.
903    
904        (2) Previously, backtracking verbs were confined within assertions. This is
905        no longer the case for positive assertions, except for (*ACCEPT). Again,
906        this sometimes improves Perl compatibility, and sometimes does not.
907    
908    32. A number of tests that were in test 2 because Perl did things differently
909        have been moved to test 1, because either Perl or PCRE has changed, and
910        these tests are now compatible.
911    
912    32. Backtracking control verbs are now handled in the same way in JIT and
913        interpreter.
914    
915    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
916        contained a forward subroutine reference caused a compile error.
917    
918    34. Auto-detect and optimize limited repetitions in JIT.
919    
920    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
921        blocking (*UTF) etc.
922    
923    36. In the interpreter, maximizing pattern repetitions for characters and
924        character types now use tail recursion, which reduces stack usage.
925    
926    37. The value of the max lookbehind was not correctly preserved if a compiled
927        and saved regex was reloaded on a host of different endianness.
928    
929    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
930        of the compiled pattern block, expand the flags field from 16 to 32 bits
931        because it was almost full.
932    
933    39. Try madvise first before posix_madvise.
934    
935    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
936        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
937        original change locked out all matches of empty strings. This has been
938        changed so that one match of an empty string per line is recognized.
939        Subsequent searches on the same line (for colouring or for --only-matching,
940        for example) do not recognize empty strings.
941    
942    41. Applied a user patch to fix a number of spelling mistakes in comments.
943    
944    42. Data lines longer than 65536 caused pcretest to crash.
945    
946    43. Clarified the data type for length and startoffset arguments for pcre_exec
947        and pcre_dfa_exec in the function-specific man pages, where they were
948        explicitly stated to be in bytes, never having been updated. I also added
949        some clarification to the pcreapi man page.
950    
951    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
952        a segmentation fault.
953    
954    
955    Version 8.32 30-November-2012
956    -----------------------------
957    
958    1.  Improved JIT compiler optimizations for first character search and single
959        character iterators.
960    
961    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
962        Patch by Daniel Richard G.
963    
964    3.  Single character iterator optimizations in the JIT compiler.
965    
966    4.  Improved JIT compiler optimizations for character ranges.
967    
968    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
969        Reported by Giuseppe D'Angelo.
970    
971    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
972        of a line, was being set incorrectly in cases where .* appeared inside
973        atomic brackets at the start of a pattern, or where there was a subsequent
974        *PRUNE or *SKIP.
975    
976    7.  Improved instruction cache flush for POWER/PowerPC.
977        Patch by Daniel Richard G.
978    
979    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
980        grep:
981    
982        (a) There is now no limit to the number of patterns to be matched.
983    
984        (b) An error is given if a pattern is too long.
985    
986        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
987            are now supported.
988    
989        (d) --exclude-from and --include-from (multiple use) have been added.
990    
991        (e) Exclusions and inclusions now apply to all files and directories, not
992            just to those obtained from scanning a directory recursively.
993    
994        (f) Multiple uses of -f and --file-list are now supported.
995    
996        (g) In a Windows environment, the default for -d has been changed from
997            "read" (the GNU grep default) to "skip", because otherwise the presence
998            of a directory in the file list provokes an error.
999    
1000        (h) The documentation has been revised and clarified in places.
1001    
1002    9.  Improve the matching speed of capturing brackets.
1003    
1004    10. Changed the meaning of \X so that it now matches a Unicode extended
1005        grapheme cluster.
1006    
1007    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
1008        out POSIX threads when JIT support is configured.
1009    
1010    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
1011    
1012    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
1013        structure could go wrong in environments where size_t is not the same size
1014        as int.
1015    
1016    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
1017        set.
1018    
1019    15. The EBCDIC support had decayed; later updates to the code had included
1020        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
1021        general tidy up of EBCDIC-related issues, and the documentation was also
1022        not quite right. There is now a test that can be run on ASCII systems to
1023        check some of the EBCDIC-related things (but is it not a full test).
1024    
1025    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
1026        in a small tidy to the code.
1027    
1028    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
1029    
1030    18. If the --only-matching (-o) option in pcregrep is specified multiple
1031        times, each one causes appropriate output. For example, -o1 -o2 outputs the
1032        substrings matched by the 1st and 2nd capturing parentheses. A separating
1033        string can be specified by --om-separator (default empty).
1034    
1035    19. Improving the first n character searches.
1036    
1037    20. Turn case lists for horizontal and vertical white space into macros so that
1038        they are defined only once.
1039    
1040    21. This set of changes together give more compatible Unicode case-folding
1041        behaviour for characters that have more than one other case when UCP
1042        support is available.
1043    
1044        (a) The Unicode property table now has offsets into a new table of sets of
1045            three or more characters that are case-equivalent. The MultiStage2.py
1046            script that generates these tables (the pcre_ucd.c file) now scans
1047            CaseFolding.txt instead of UnicodeData.txt for character case
1048            information.
1049    
1050        (b) The code for adding characters or ranges of characters to a character
1051            class has been abstracted into a generalized function that also handles
1052            case-independence. In UTF-mode with UCP support, this uses the new data
1053            to handle characters with more than one other case.
1054    
1055        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
1056            whose other case is greater than 256 are now correctly matched
1057            caselessly. Previously, the high codepoint matched the low one, but not
1058            vice versa.
1059    
1060        (d) The processing of \h, \H, \v, and \ in character classes now makes use
1061            of the new class addition function, using character lists defined as
1062            macros alongside the case definitions of 20 above.
1063    
1064        (e) Caseless back references now work with characters that have more than
1065            one other case.
1066    
1067        (f) General caseless matching of characters with more than one other case
1068            is supported.
1069    
1070    22. Unicode character properties were updated from Unicode 6.2.0
1071    
1072    23. Improved CMake support under Windows. Patch by Daniel Richard G.
1073    
1074    24. Add support for 32-bit character strings, and UTF-32
1075    
1076    25. Major JIT compiler update (code refactoring and bugfixing).
1077        Experimental Sparc 32 support is added.
1078    
1079    26. Applied a modified version of Daniel Richard G's patch to create
1080        pcre.h.generic and config.h.generic by "make" instead of in the
1081        PrepareRelease script.
1082    
1083    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
1084        pcre_compile.c when checking for a zero character.
1085    
1086    28. Introducing a native interface for JIT. Through this interface, the compiled
1087        machine code can be directly executed. The purpose of this interface is to
1088        provide fast pattern matching, so several sanity checks are not performed.
1089        However, feature tests are still performed. The new interface provides
1090        1.4x speedup compared to the old one.
1091    
1092    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
1093        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
1094        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
1095    
1096    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
1097        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
1098        the "old" RFC 2279). Instead, it ended up passing a negative length to
1099        pcre_exec().
1100    
1101    31. Add support for GCC's visibility feature to hide internal functions.
1102    
1103    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
1104        "unknown -C option" after outputting 0 or 1.
1105    
1106    33. There is now support for generating a code coverage report for the test
1107        suite in environments where gcc is the compiler and lcov is installed. This
1108        is mainly for the benefit of the developers.
1109    
1110    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
1111        unaddressable using valgrind annotations, allowing valgrind to detect
1112        invalid memory accesses. This is mainly for the benefit of the developers.
1113    
1114    25. (*UTF) can now be used to start a pattern in any of the three libraries.
1115    
1116    26. Give configure error if --enable-cpp but no C++ compiler found.
1117    
1118    
1119    Version 8.31 06-July-2012
1120    -------------------------
1121    
1122    1.  Fixing a wrong JIT test case and some compiler warnings.
1123    
1124    2.  Removed a bashism from the RunTest script.
1125    
1126    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
1127        to unsigned type, result still unsigned" that was given by an MS compiler
1128        on encountering the code "-sizeof(xxx)".
1129    
1130    4.  Partial matching support is added to the JIT compiler.
1131    
1132    5.  Fixed several bugs concerned with partial matching of items that consist
1133        of more than one character:
1134    
1135        (a) /^(..)\1/ did not partially match "aba" because checking references was
1136            done on an "all or nothing" basis. This also applied to repeated
1137            references.
1138    
1139        (b) \R did not give a hard partial match if \r was found at the end of the
1140            subject.
1141    
1142        (c) \X did not give a hard partial match after matching one or more
1143            characters at the end of the subject.
1144    
1145        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1146            a partial match for the string "\r".
1147    
1148        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1149            a partial match for a CR character at the end of the subject string.
1150    
1151    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1152        running pcretest, the text "(JIT)" added to the output whenever JIT is
1153        actually used to run the match.
1154    
1155    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1156        or /S+[+] with a digit between 1 and 7.
1157    
1158    8.  OP_NOT now supports any UTF character not just single-byte ones.
1159    
1160    9.  (*MARK) control verb is now supported by the JIT compiler.
1161    
1162    10. The command "./RunTest list" lists the available tests without actually
1163        running any of them. (Because I keep forgetting what they all are.)
1164    
1165    11. Add PCRE_INFO_MAXLOOKBEHIND.
1166    
1167    12. Applied a (slightly modified) user-supplied patch that improves performance
1168        when the heap is used for recursion (compiled with --disable-stack-for-
1169        recursion). Instead of malloc and free for each heap frame each time a
1170        logical recursion happens, frames are retained on a chain and re-used where
1171        possible. This sometimes gives as much as 30% improvement.
1172    
1173    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1174        call.
1175    
1176    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1177    
1178    15. It is now possible to link pcretest with libedit as an alternative to
1179        libreadline.
1180    
1181    16. (*COMMIT) control verb is now supported by the JIT compiler.
1182    
1183    17. The Unicode data tables have been updated to Unicode 6.1.0.
1184    
1185    18. Added --file-list option to pcregrep.
1186    
1187    19. Added binary file support to pcregrep, including the -a, --binary-files,
1188        -I, and --text options.
1189    
1190    20. The madvise function is renamed for posix_madvise for QNX compatibility
1191        reasons. Fixed by Giuseppe D'Angelo.
1192    
1193    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1194    
1195    22. Changed the option for creating symbolic links for 16-bit man pages from
1196        -s to -sf so that re-installing does not cause issues.
1197    
1198    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1199    
1200    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1201        matches in certain environments (the workspace was not being correctly
1202        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1203        some of the workspace data at the beginning of a restart.
1204    
1205    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1206        was not doing so when it should - probably a typo introduced by SVN 528
1207        (change 8.10/14).
1208    
1209    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1210        \w+ when the character tables indicated that \x{c4} was a word character.
1211        There were several related cases, all because the tests for doing a table
1212        lookup were testing for characters less than 127 instead of 255.
1213    
1214    27. If a pattern contains capturing parentheses that are not used in a match,
1215        their slots in the ovector are set to -1. For those that are higher than
1216        any matched groups, this happens at the end of processing. In the case when
1217        there were back references that the ovector was too small to contain
1218        (causing temporary malloc'd memory to be used during matching), and the
1219        highest capturing number was not used, memory off the end of the ovector
1220        was incorrectly being set to -1. (It was using the size of the temporary
1221        memory instead of the true size.)
1222    
1223    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1224        ovector size, it uses memory at the end of the block that it has got.
1225    
1226    29. Check for an overlong MARK name and give an error at compile time. The
1227        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1228    
1229    30. JIT compiler update.
1230    
1231    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1232        Rill for the patch.
1233    
1234    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1235    
1236    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1237    
1238    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1239        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1240        was enabled.
1241    
1242    35. Improve JIT code generation for greedy plus quantifier.
1243    
1244    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1245        1 to "aa" instead of to an empty string. The bug affected repeated groups
1246        that could potentially match an empty string.
1247    
1248    37. Optimizing single character iterators in JIT.
1249    
1250    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1251        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1252        codepoints that are too big for the mode are faulted, and in a UTF mode,
1253        disallowed codepoints are also faulted.
1254    
1255    39. If PCRE was compiled with UTF support, in three places in the DFA
1256        matcher there was code that should only have been obeyed in UTF mode, but
1257        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1258        processing when bytes with values greater than 127 were present. In 16-bit
1259        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1260        both cases the values are those that cannot be the first data item in a UTF
1261        character. The three items that might have provoked this were recursions,
1262        possessively repeated groups, and atomic groups.
1263    
1264    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1265        and pcregrep, because some OS require shared objects to be explicitly
1266        passed to ld, causing the link step to fail if they are not.
1267    
1268    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1269        mode, patterns that started with \h* or \R* might be incorrectly matched.
1270    
1271    
1272    Version 8.30 04-February-2012
1273    -----------------------------
1274    
1275    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1276        name is defined in ctype.h.
1277    
1278    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1279        only in quite long subpatterns.
1280    
1281    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1282        since it was replaced by pcre_fullinfo() in February 2000.
1283    
1284    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1285        match a (*MARK), and the match failed at the start of the subject, a
1286        reference to memory before the start of the subject could occur. This bug
1287        was introduced by fix 17 of release 8.21.
1288    
1289    5.  A reference to an unset group with zero minimum repetition was giving
1290        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1291        /(another)?(\1?)test/ matched against "hello world test". This bug was
1292        introduced in release 8.13.
1293    
1294    6.  Add support for 16-bit character strings (a large amount of work involving
1295        many changes and refactorings).
1296    
1297    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1298        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1299        from a file.
1300    
1301    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1302        rounding is not applied in this particular case).
1303    
1304    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1305        if they appear, or are escaped, in patterns.
1306    
1307    10. Get rid of a number of -Wunused-but-set-variable warnings.
1308    
1309    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1310        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1311        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1312        also returns the mark "x". This bug applied to capturing parentheses,
1313        non-capturing parentheses, and atomic parentheses. It also applied to some
1314        assertions.
1315    
1316    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1317        information out of configure.ac instead of relying on pcre.h.generic, which
1318        is not stored in the repository.
1319    
1320    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1321        -lreadline.
1322    
1323    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1324    
1325    15. Applied Graycode's patch to put the top-level frame on the stack rather
1326        than the heap when not using the stack for recursion. This gives a
1327        performance improvement in many cases when recursion is not deep.
1328    
1329    16. Experimental code added to "pcretest -C" to output the stack frame size.
1330    
1331    
1332    Version 8.21 12-Dec-2011
1333    ------------------------
1334    
1335    1.  Updating the JIT compiler.
1336    
1337    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1338        are added as well.
1339    
1340    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1341        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1342        calling _pcre_jit_exec. Some extra comments are added.
1343    
1344    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1345        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1346        was introduced by change 18 for 8.20.
1347    
1348    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1349        ECMA-262 standard.
1350    
1351    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1352        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1353        This bug was probably introduced by change 9 of 8.13.
1354    
1355    7.  While fixing 6 above, I noticed that a number of other items were being
1356        incorrectly rejected as "not fixed length". This arose partly because newer
1357        opcodes had not been added to the fixed-length checking code. I have (a)
1358        corrected the bug and added tests for these items, and (b) arranged for an
1359        error to occur if an unknown opcode is encountered while checking for fixed
1360        length instead of just assuming "not fixed length". The items that were
1361        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1362        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1363        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1364    
1365    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1366        being incorrectly compiled and would have given unpredicatble results.
1367    
1368    9.  A possessively repeated subpattern with minimum repeat count greater than
1369        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1370        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1371        the first (A) could occur when it should not.
1372    
1373    10. Add a cast and remove a redundant test from the code.
1374    
1375    11. JIT should use pcre_malloc/pcre_free for allocation.
1376    
1377    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1378        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1379        is anything other than /usr, -L is still shown).
1380    
1381    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1382    
1383    14. Perl does not support \N without a following name in a [] class; PCRE now
1384        also gives an error.
1385    
1386    15. If a forward reference was repeated with an upper limit of around 2000,
1387        it caused the error "internal error: overran compiling workspace". The
1388        maximum number of forward references (including repeats) was limited by the
1389        internal workspace, and dependent on the LINK_SIZE. The code has been
1390        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1391        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1392        of around 200,000 forward references. While doing this, I also speeded up
1393        the filling in of repeated forward references.
1394    
1395    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1396        incorrectly expecting the subject to contain another "a" after the start.
1397    
1398    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1399        in the match, the SKIP should be ignored. This was not happening; instead
1400        the SKIP was being treated as NOMATCH. For patterns such as
1401        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1402        tested.
1403    
1404    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1405        now much more compatible with Perl, in particular in cases where the result
1406        is a non-match for a non-anchored pattern. For example, if
1407        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1408        "m", where previously it did not return a name. A side effect of this
1409        change is that for partial matches, the last encountered mark name is
1410        returned, as for non matches. A number of tests that were previously not
1411        Perl-compatible have been moved into the Perl-compatible test files. The
1412        refactoring has had the pleasing side effect of removing one argument from
1413        the match() function, thus reducing its stack requirements.
1414    
1415    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1416        subsequent uses of /S (without +) incorrectly behaved like /S+.
1417    
1418    21. Retrieve executable code size support for the JIT compiler and fixing
1419        some warnings.
1420    
1421    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1422        not work when the shorter character appeared right at the end of the
1423        subject string.
1424    
1425    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1426        systems.
1427    
1428    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1429        output it when the /M option is used in pcretest.
1430    
1431    25. The CheckMan script was not being included in the distribution. Also, added
1432        an explicit "perl" to run Perl scripts from the PrepareRelease script
1433        because this is reportedly needed in Windows.
1434    
1435    26. If study data was being save in a file and studying had not found a set of
1436        "starts with" bytes for the pattern, the data written to the file (though
1437        never used) was taken from uninitialized memory and so caused valgrind to
1438        complain.
1439    
1440    27. Updated RunTest.bat as provided by Sheri Pierce.
1441    
1442    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1443    
1444    29. Computation of memory usage for the table of capturing group names was
1445        giving an unnecessarily large value.
1446    
1447    
1448    Version 8.20 21-Oct-2011
1449    ------------------------
1450    
1451    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1452        a POSIX class. After further experiments with Perl, which convinced me that
1453        Perl has bugs and confusions, a closing square bracket is no longer allowed
1454        in a POSIX name. This bug also affected patterns with classes that started
1455        with full stops.
1456    
1457    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1458        captured substring, but while checking the failing first alternative,
1459        substring 1 is temporarily captured. If the output vector supplied to
1460        pcre_exec() was not big enough for this capture, the yield of the function
1461        was still zero ("insufficient space for captured substrings"). This cannot
1462        be totally fixed without adding another stack variable, which seems a lot
1463        of expense for a edge case. However, I have improved the situation in cases
1464        such as /(a)(b)x|abc/ matched against "abc", where the return code
1465        indicates that fewer than the maximum number of slots in the ovector have
1466        been set.
1467    
1468    3.  Related to (2) above: when there are more back references in a pattern than
1469        slots in the output vector, pcre_exec() uses temporary memory during
1470        matching, and copies in the captures as far as possible afterwards. It was
1471        using the entire output vector, but this conflicts with the specification
1472        that only 2/3 is used for passing back captured substrings. Now it uses
1473        only the first 2/3, for compatibility. This is, of course, another edge
1474        case.
1475    
1476    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1477        main code base, and can be used by building with --enable-jit. When this is
1478        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1479        runtime --no-jit option is given.
1480    
1481    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1482        ovector, the return from the function was zero, implying that there were
1483        other matches that did not fit. The correct "exactly full" value is now
1484        returned.
1485    
1486    6.  If a subpattern that was called recursively or as a subroutine contained
1487        (*PRUNE) or any other control that caused it to give a non-standard return,
1488        invalid errors such as "Error -26 (nested recursion at the same subject
1489        position)" or even infinite loops could occur.
1490    
1491    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1492        computing the minimum length on reaching *ACCEPT, and so ended up with the
1493        wrong value of 1 rather than 0. Further investigation indicates that
1494        computing a minimum subject length in the presence of *ACCEPT is difficult
1495        (think back references, subroutine calls), and so I have changed the code
1496        so that no minimum is registered for a pattern that contains *ACCEPT.
1497    
1498    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1499        it was not handled as intended. [But see 16 below.]
1500    
1501    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1502        Sheri Pierce.
1503    
1504    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1505        the first byte in a match must be "a".
1506    
1507    11. Change 17 for 8.13 increased the recursion depth for patterns like
1508        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1509        pattern contains any instances of (*THEN). If it does not, the old
1510        optimizations are restored. It would be nice to do this on a per-group
1511        basis, but at the moment that is not feasible.
1512    
1513    12. In some environments, the output of pcretest -C is CRLF terminated. This
1514        broke RunTest's code that checks for the link size. A single white space
1515        character after the value is now allowed for.
1516    
1517    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1518        For "fr", it uses the Windows-specific input and output files.
1519    
1520    14. If (*THEN) appeared in a group that was called recursively or as a
1521        subroutine, it did not work as intended. [But see next item.]
1522    
1523    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1524        pattern fragments (but not containing any | characters). If A and B are
1525        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1526        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1527        D. In other words, Perl considers parentheses that do not contain any |
1528        characters to be part of a surrounding alternative, whereas PCRE was
1529        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1530        differently. PCRE now behaves in the same way as Perl, except in the case
1531        of subroutine/recursion calls such as (?1) which have in any case always
1532        been different (but PCRE had them first :-).
1533    
1534    16. Related to 15 above: Perl does not treat the | in a conditional group as
1535        creating alternatives. Such a group is treated in the same way as an
1536        ordinary group without any | characters when processing (*THEN). PCRE has
1537        been changed to match Perl's behaviour.
1538    
1539    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1540        RunGrepTest script failed.
1541    
1542    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1543        inevitable for groups that contain captures, but it can lead to a lot of
1544        stack use in large patterns. The old behaviour has been restored for atomic
1545        groups that do not contain any capturing parentheses.
1546    
1547    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1548        suppress the check for a minimum subject length at run time. (If it was
1549        given to pcre_exec() or pcre_dfa_exec() it did work.)
1550    
1551    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1552        fail to work when decoding hex characters in data strings in EBCDIC
1553        environments.
1554    
1555    21. It appears that in at least one Mac OS environment, the isxdigit() function
1556        is implemented as a macro that evaluates to its argument more than once,
1557        contravening the C 90 Standard (I haven't checked a later standard). There
1558        was an instance in pcretest which caused it to go wrong when processing
1559        \x{...} escapes in subject strings. The has been rewritten to avoid using
1560        things like p++ in the argument of isxdigit().
1561    
1562    
1563    Version 8.13 16-Aug-2011
1564  ------------------------  ------------------------
1565    
1566  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 20  Version 8.13 30-Apr-2011 Line 1579  Version 8.13 30-Apr-2011
1579      code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a      code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1580      2-byte code at the end of the subject (it thought there wasn't enough data      2-byte code at the end of the subject (it thought there wasn't enough data
1581      left).      left).
1582    
1583  5.  Comprehensive information about what went wrong is now returned by  5.  Comprehensive information about what went wrong is now returned by
1584      pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long      pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1585      as the output vector has at least 2 elements. The offset of the start of      as the output vector has at least 2 elements. The offset of the start of
1586      the failing character and a reason code are placed in the vector.      the failing character and a reason code are placed in the vector.
1587    
1588  6.  When the UTF-8 string check fails for pcre_compile(), the offset that is  6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1589      now returned is for the first byte of the failing character, instead of the      now returned is for the first byte of the failing character, instead of the
1590      last byte inspected. This is an incompatible change, but I hope it is small      last byte inspected. This is an incompatible change, but I hope it is small
1591      enough not to be a problem. It makes the returned offset consistent with      enough not to be a problem. It makes the returned offset consistent with
1592      pcre_exec() and pcre_dfa_exec().      pcre_exec() and pcre_dfa_exec().
1593    
1594  7.  pcretest now gives a text phrase as well as the error number when  7.  pcretest now gives a text phrase as well as the error number when
1595      pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check      pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1596      failure, the offset and reason code are output.      failure, the offset and reason code are output.
1597    
1598  8.  When \R was used with a maximizing quantifier it failed to skip backwards  8.  When \R was used with a maximizing quantifier it failed to skip backwards
1599      over a \r\n pair if the subsequent match failed. Instead, it just skipped      over a \r\n pair if the subsequent match failed. Instead, it just skipped
1600      back over a single character (\n). This seems wrong (because it treated the      back over a single character (\n). This seems wrong (because it treated the
1601      two characters as a single entity when going forwards), conflicts with the      two characters as a single entity when going forwards), conflicts with the
1602      documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the      documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1603      behaviour of \R* different to (\R)*, which also seems wrong. The behaviour      behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1604      has been changed.      has been changed.
1605    
1606    9.  Some internal refactoring has changed the processing so that the handling
1607        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1608        time (the PCRE_DOTALL option was changed this way some time ago: version
1609        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1610        which was always a bit of a fudge. It also means that there is one less
1611        argument for the match() function, which reduces its stack requirements
1612        slightly. This change also fixes an incompatibility with Perl: the pattern
1613        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1614    
1615    10. More internal refactoring has drastically reduced the number of recursive
1616        calls to match() for possessively repeated groups such as (abc)++ when
1617        using pcre_exec().
1618    
1619    11. While implementing 10, a number of bugs in the handling of groups were
1620        discovered and fixed:
1621    
1622        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1623        (a|)*(?1) gave a compile-time internal error.
1624        ((a|)+)+  did not notice that the outer group could match an empty string.
1625        (^a|^)+   was not marked as anchored.
1626        (.*a|.*)+ was not marked as matching at start or after a newline.
1627    
1628    12. Yet more internal refactoring has removed another argument from the match()
1629        function. Special calls to this function are now indicated by setting a
1630        value in a variable in the "match data" data block.
1631    
1632    13. Be more explicit in pcre_study() instead of relying on "default" for
1633        opcodes that mean there is no starting character; this means that when new
1634        ones are added and accidentally left out of pcre_study(), testing should
1635        pick them up.
1636    
1637    14. The -s option of pcretest has been documented for ages as being an old
1638        synonym of -m (show memory usage). I have changed it to mean "force study
1639        for every regex", that is, assume /S for every regex. This is similar to -i
1640        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1641        using it. It makes it easier to run collections of tests with and without
1642        study enabled, and thereby test pcre_study() more easily. All the standard
1643        tests are now run with and without -s (but some patterns can be marked as
1644        "never study" - see 20 below).
1645    
1646    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1647        restoration of the capturing data to the outer values was not happening
1648        correctly.
1649    
1650    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1651        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1652        pattern had matched an empty string, and so incorrectly returned a no
1653        match.
1654    
1655    17. There was optimizing code for the last branch of non-capturing parentheses,
1656        and also for the obeyed branch of a conditional subexpression, which used
1657        tail recursion to cut down on stack usage. Unfortunately, now that there is
1658        the possibility of (*THEN) occurring in these branches, tail recursion is
1659        no longer possible because the return has to be checked for (*THEN). These
1660        two optimizations have therefore been removed. [But see 8.20/11 above.]
1661    
1662    18. If a pattern containing \R was studied, it was assumed that \R always
1663        matched two bytes, thus causing the minimum subject length to be
1664        incorrectly computed because \R can also match just one byte.
1665    
1666    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1667        was incorrectly computed.
1668    
1669    20. If /S is present twice on a test pattern in pcretest input, it now
1670        *disables* studying, thereby overriding the use of -s on the command line
1671        (see 14 above). This is necessary for one or two tests to keep the output
1672        identical in both cases.
1673    
1674    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1675        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1676    
1677    22. When an atomic group that contained a capturing parenthesis was
1678        successfully matched, but the branch in which it appeared failed, the
1679        capturing was not being forgotten if a higher numbered group was later
1680        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1681        group 1 to "a", when in fact it should be unset. This applied to multi-
1682        branched capturing and non-capturing groups, repeated or not, and also to
1683        positive assertions (capturing in negative assertions does not happen
1684        in PCRE) and also to nested atomic groups.
1685    
1686    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1687        subject after a captured substring, to make it easier to tell which of a
1688        number of identical substrings has been captured.
1689    
1690    24. The way atomic groups are processed by pcre_exec() has been changed so that
1691        if they are repeated, backtracking one repetition now resets captured
1692        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1693        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1694        "aaa". Previously, it would have been "a". As part of this code
1695        refactoring, the way recursive calls are handled has also been changed.
1696    
1697    25. If an assertion condition captured any substrings, they were not passed
1698        back unless some other capturing happened later. For example, if
1699        (?(?=(a))a) was matched against "a", no capturing was returned.
1700    
1701    26. When studying a pattern that contained subroutine calls or assertions,
1702        the code for finding the minimum length of a possible match was handling
1703        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1704        group 1 called group 2 while simultaneously a separate group 2 called group
1705        1). A stack overflow occurred in this case. I have fixed this by limiting
1706        the recursion depth to 10.
1707    
1708    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1709        Fortmann. This supports explicit test numbers on the command line, and has
1710        argument validation and error reporting.
1711    
1712    28. An instance of \X with an unlimited repeat could fail if at any point the
1713        first character it looked at was a mark character.
1714    
1715    29. Some minor code refactoring concerning Unicode properties and scripts
1716        should reduce the stack requirement of match() slightly.
1717    
1718    30. Added the '=' option to pcretest to check the setting of unused capturing
1719        slots at the end of the pattern, which are documented as being -1, but are
1720        not included in the return count.
1721    
1722    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1723        compiled something random. Now it gives a compile-time error (as does
1724        Perl).
1725    
1726    32. A *MARK encountered during the processing of a positive assertion is now
1727        recorded and passed back (compatible with Perl).
1728    
1729    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1730        had alternative anchored branches, the search for a second match in a line
1731        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1732        matched the line "0102" twice. The same bug affected patterns that started
1733        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1734        twice.
1735    
1736    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1737        does, and because of capturing effects, quantifying parenthesized
1738        assertions may at times be useful. Quantifiers are now allowed for
1739        parenthesized assertions.
1740    
1741    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1742    
1743    36. \g was being checked for fancy things in a character class, when it should
1744        just be a literal "g".
1745    
1746    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1747        appearance of a nested POSIX class supersedes an apparent external class.
1748        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1749        unescaped square brackets may also appear as part of class names. For
1750        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1751        more like Perl. (But see 8.20/1 above.)
1752    
1753    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1754        was because it thought it was \N{name}, which is not supported).
1755    
1756    39. Add minix to OS list not supporting the -S option in pcretest.
1757    
1758    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1759        cannot analyze patterns in sufficient detail to catch mutual recursions
1760        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1761        subgroup is called recursively as a subpattern for a second time at the
1762        same position in the subject string. In previous releases this might have
1763        been caught by the recursion limit, or it might have run out of stack.
1764    
1765    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1766        happen only once. PCRE was, however incorrectly giving a compile time error
1767        "recursive call could loop indefinitely" because it cannot analyze the
1768        pattern in sufficient detail. The compile time test no longer happens when
1769        PCRE is compiling a conditional subpattern, but actual runaway loops are
1770        now caught at runtime (see 40 above).
1771    
1772    42. It seems that Perl allows any characters other than a closing parenthesis
1773        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1774        has been changed to be the same.
1775    
1776    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1777        as not to get warnings when autogen.sh is called. Also changed
1778        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1779    
1780    44. To help people who use pcregrep to scan files containing exceedingly long
1781        lines, the following changes have been made:
1782    
1783        (a) The default value of the buffer size parameter has been increased from
1784            8K to 20K. (The actual buffer used is three times this size.)
1785    
1786        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1787            PCRE is built.
1788    
1789        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1790            to be set at run time.
1791    
1792        (d) Numerical values in pcregrep options can be followed by K or M, for
1793            example --buffer-size=50K.
1794    
1795        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1796            given and the return code is set to 2.
1797    
1798    45. Add a pointer to the latest mark to the callout data block.
1799    
1800    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1801        partial match of an empty string instead of no match. This was specific to
1802        the use of ".".
1803    
1804    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1805        complete match instead of a partial match. This bug was dependent on both
1806        the PCRE_UTF8 and PCRE_DOTALL options being set.
1807    
1808    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1809        starting byte set, because \b was not being ignored.
1810    
1811    
1812  Version 8.12 15-Jan-2011  Version 8.12 15-Jan-2011
1813  ------------------------  ------------------------
# Line 695  Version 7.9 11-Apr-09 Line 2459  Version 7.9 11-Apr-09
2459  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
2460      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
2461      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2462      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2463        8.33.]
2464    
2465  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
2466      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 938  Version 7.7 07-May-08 Line 2703  Version 7.7 07-May-08
2703      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2704      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2705      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2706      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2707      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2708      was subsequently set up correctly.)      was subsequently set up correctly.)
2709    

Legend:
Removed from v.600  
changed lines
  Added in v.1613

  ViewVC Help
Powered by ViewVC 1.1.5