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

Diff of /code/trunk/ChangeLog

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

revision 389 by ph10, Sun Mar 15 18:24:05 2009 UTC revision 1600 by ph10, Wed Sep 2 08:54:57 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.9 xx-xxx-09  Note that the PCRE 8.xx series (PCRE1) is now in a bugfix-only state. All
5    development is happening in the PCRE2 10.xx series.
6    
7    Version 8.38 xx-xxx-xxxx
8    ------------------------
9    
10    1.  If a group that contained a recursive back reference also contained a
11        forward reference subroutine call followed by a non-forward-reference
12        subroutine call, for example /.((?2)(?R)\1)()/, pcre2_compile() failed to
13        compile correct code, leading to undefined behaviour or an internally
14        detected error. This bug was discovered by the LLVM fuzzer.
15    
16    2.  Quantification of certain items (e.g. atomic back references) could cause
17        incorrect code to be compiled when recursive forward references were
18        involved. For example, in this pattern: /(?1)()((((((\1++))\x85)+)|))/.
19        This bug was discovered by the LLVM fuzzer.
20    
21    3.  A repeated conditional group whose condition was a reference by name caused
22        a buffer overflow if there was more than one group with the given name.
23        This bug was discovered by the LLVM fuzzer.
24    
25    4.  A recursive back reference by name within a group that had the same name as
26        another group caused a buffer overflow. For example:
27        /(?J)(?'d'(?'d'\g{d}))/. This bug was discovered by the LLVM fuzzer.
28    
29    5.  A forward reference by name to a group whose number is the same as the
30        current group, for example in this pattern: /(?|(\k'Pm')|(?'Pm'))/, caused
31        a buffer overflow at compile time. This bug was discovered by the LLVM
32        fuzzer.
33    
34    6.  A lookbehind assertion within a set of mutually recursive subpatterns could
35        provoke a buffer overflow. This bug was discovered by the LLVM fuzzer.
36    
37    7.  Another buffer overflow bug involved duplicate named groups with a
38        reference between their definition, with a group that reset capture
39        numbers, for example: /(?J:(?|(?'R')(\k'R')|((?'R'))))/. This has been
40        fixed by always allowing for more memory, even if not needed. (A proper fix
41        is implemented in PCRE2, but it involves more refactoring.)
42    
43    8.  There was no check for integer overflow in subroutine calls such as (?123).
44    
45    9.  The table entry for \l in EBCDIC environments was incorrect, leading to its
46        being treated as a literal 'l' instead of causing an error.
47    
48    10. There was a buffer overflow if pcre_exec() was called with an ovector of
49        size 1. This bug was found by american fuzzy lop.
50    
51    11. If a non-capturing group containing a conditional group that could match
52        an empty string was repeated, it was not identified as matching an empty
53        string itself. For example: /^(?:(?(1)x|)+)+$()/.
54    
55    12. In an EBCDIC environment, pcretest was mishandling the escape sequences
56        \a and \e in test subject lines.
57    
58    13. In an EBCDIC environment, \a in a pattern was converted to the ASCII
59        instead of the EBCDIC value.
60    
61    14. The handling of \c in an EBCDIC environment has been revised so that it is
62        now compatible with the specification in Perl's perlebcdic page.
63    
64    15. The EBCDIC character 0x41 is a non-breaking space, equivalent to 0xa0 in
65        ASCII/Unicode. This has now been added to the list of characters that are
66        recognized as white space in EBCDIC.
67    
68    16. When PCRE was compiled without UCP support, the use of \p and \P gave an
69        error (correctly) when used outside a class, but did not give an error
70        within a class.
71    
72    17. \h within a class was incorrectly compiled in EBCDIC environments.
73    
74    18. A pattern with an unmatched closing parenthesis that contained a backward
75        assertion which itself contained a forward reference caused buffer
76        overflow. And example pattern is: /(?=di(?<=(?1))|(?=(.))))/.
77    
78    19. JIT should return with error when the compiled pattern requires more stack
79        space than the maximum.
80    
81    20. A possessively repeated conditional group that could match an empty string,
82        for example, /(?(R))*+/, was incorrectly compiled.
83    
84    21. Fix infinite recursion in the JIT compiler when certain patterns such as
85        /(?:|a|){100}x/ are analysed.
86    
87    22. Some patterns with character classes involving [: and \\ were incorrectly
88        compiled and could cause reading from uninitialized memory or an incorrect
89        error diagnosis.
90    
91    23. Pathological patterns containing many nested occurrences of [: caused
92        pcre_compile() to run for a very long time.
93    
94    24. A conditional group with only one branch has an implicit empty alternative
95        branch and must therefore be treated as potentially matching an empty
96        string.
97    
98    25. If (?R was followed by - or + incorrect behaviour happened instead of a
99        diagnostic.
100    
101    26. Arrange to give up on finding the minimum matching length for overly
102        complex patterns.
103    
104    27. Similar to (4) above: in a pattern with duplicated named groups and an
105        occurrence of (?| it is possible for an apparently non-recursive back
106        reference to become recursive if a later named group with the relevant
107        number is encountered. This could lead to a buffer overflow. Wen Guanxing
108        from Venustech ADLAB discovered this bug.
109    
110    28. If pcregrep was given the -q option with -c or -l, or when handling a
111        binary file, it incorrectly wrote output to stdout.
112    
113    29. The JIT compiler did not restore the control verb head in case of *THEN
114        control verbs. This issue was found by Karl Skomski with a custom LLVM
115        fuzzer.
116    
117    30. Error messages for syntax errors following \g and \k were giving inaccurate
118        offsets in the pattern.
119    
120    31. Added a check for integer overflow in conditions (?(<digits>) and
121        (?(R<digits>). This omission was discovered by Karl Skomski with the LLVM
122        fuzzer.
123    
124    32. Handling recursive references such as (?2) when the reference is to a group
125        later in the pattern uses code that is very hacked about and error-prone.
126        It has been re-written for PCRE2. Here in PCRE1, a check has been added to
127        give an internal error if it is obvious that compiling has gone wrong.
128    
129    33. The JIT compiler should not check repeats after a {0,1} repeat byte code.
130        This issue was found by Karl Skomski with a custom LLVM fuzzer.
131    
132    34. The JIT compiler should restore the control chain for empty possessive
133        repeats. This issue was found by Karl Skomski with a custom LLVM fuzzer.
134    
135    35. Match limit check added to JIT recursion. This issue was found by Karl
136        Skomski with a custom LLVM fuzzer.
137    
138    36. Yet another case similar to 27 above has been circumvented by an
139        unconditional allocation of extra memory. This issue is fixed "properly" in
140        PCRE2 by refactoring the way references are handled. Wen Guanxing
141        from Venustech ADLAB discovered this bug.
142    
143    37. Fix two assertion fails in JIT. These issues were found by Karl Skomski
144        with a custom LLVM fuzzer.
145    
146    38. Fixed a corner case of range optimization in JIT.
147    
148    39. An incorrect error "overran compiling workspace" was given if there were
149        exactly enough group forward references such that the last one extended
150        into the workspace safety margin. The next one would have expanded the
151        workspace. The test for overflow was not including the safety margin.
152    
153    40. A match limit issue is fixed in JIT which was found by Karl Skomski
154        with a custom LLVM fuzzer.
155    
156    41. Remove the use of /dev/null in testdata/testinput2, because it doesn't
157        work under Windows. (Why has it taken so long for anyone to notice?)
158    
159    
160    Version 8.37 28-April-2015
161    --------------------------
162    
163    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
164        for those parentheses to be closed with whatever has been captured so far.
165        However, it was failing to mark any other groups between the hightest
166        capture so far and the currrent group as "unset". Thus, the ovector for
167        those groups contained whatever was previously there. An example is the
168        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
169    
170    2.  If an assertion condition was quantified with a minimum of zero (an odd
171        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
172    
173    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
174        unrecognized modifier, a crash could occur.
175    
176    4.  An attempt to do global matching in pcretest with a zero-length ovector
177        caused a crash.
178    
179    5.  Fixed a memory leak during matching that could occur for a subpattern
180        subroutine call (recursive or otherwise) if the number of captured groups
181        that had to be saved was greater than ten.
182    
183    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
184        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
185        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
186    
187    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
188        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
189    
190    8. When a pattern is compiled, it remembers the highest back reference so that
191       when matching, if the ovector is too small, extra memory can be obtained to
192       use instead. A conditional subpattern whose condition is a check on a
193       capture having happened, such as, for example in the pattern
194       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
195       setting the highest backreference number. This mattered only if pcre_exec()
196       was called with an ovector that was too small to hold the capture, and there
197       was no other kind of back reference (a situation which is probably quite
198       rare). The effect of the bug was that the condition was always treated as
199       FALSE when the capture could not be consulted, leading to a incorrect
200       behaviour by pcre_exec(). This bug has been fixed.
201    
202    9. A reference to a duplicated named group (either a back reference or a test
203       for being set in a conditional) that occurred in a part of the pattern where
204       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
205       to be incorrectly calculated, leading to overwriting.
206    
207    10. A mutually recursive set of back references such as (\2)(\1) caused a
208        segfault at study time (while trying to find the minimum matching length).
209        The infinite loop is now broken (with the minimum length unset, that is,
210        zero).
211    
212    11. If an assertion that was used as a condition was quantified with a minimum
213        of zero, matching went wrong. In particular, if the whole group had
214        unlimited repetition and could match an empty string, a segfault was
215        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
216        assertions to be quantified, but not if they are being used as conditions,
217        so the above pattern is faulted by Perl. PCRE has now been changed so that
218        it also rejects such patterns.
219    
220    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
221        failed to allow the zero-repeat case if pcre2_exec() was called with an
222        ovector too small to capture the group.
223    
224    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
225        Red Hat Product Security:
226    
227        (a) A crash if /K and /F were both set with the option to save the compiled
228        pattern.
229    
230        (b) Another crash if the option to print captured substrings in a callout
231        was combined with setting a null ovector, for example \O\C+ as a subject
232        string.
233    
234    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
235        forward reference repeated a large (but limited) number of times within a
236        repeated outer group that has a zero minimum quantifier, caused incorrect
237        code to be compiled, leading to the error "internal error:
238        previously-checked referenced subpattern not found" when an incorrect
239        memory address was read. This bug was reported as "heap overflow",
240        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
241        CVE-2015-2325.
242    
243    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
244        call within a group that also contained a recursive back reference caused
245        incorrect code to be compiled. This bug was reported as "heap overflow",
246        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
247        number CVE-2015-2326.
248    
249    24. Computing the size of the JIT read-only data in advance has been a source
250        of various issues, and new ones are still appear unfortunately. To fix
251        existing and future issues, size computation is eliminated from the code,
252        and replaced by on-demand memory allocation.
253    
254    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
255        adjacent to the end of the range, and the range contained characters with
256        more than one other case, caused incorrect behaviour when compiled in UTF
257        mode. In that example, the range a-j was left out of the class.
258    
259    26. Fix JIT compilation of conditional blocks, which assertion
260        is converted to (*FAIL). E.g: /(?(?!))/.
261    
262    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
263        discovered by the LLVM fuzzer.
264    
265    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
266        when this assertion was used as a condition, for example (?(?!)a|b). In
267        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
268        error about an unsupported item.
269    
270    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
271        possessification code could take exponential time to complete. A recursion
272        depth limit of 1000 has been imposed to limit the resources used by this
273        optimization.
274    
275    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
276        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
277        because \S ensures they are all in the class. The code for doing this was
278        interacting badly with the code for computing the amount of space needed to
279        compile the pattern, leading to a buffer overflow. This bug was discovered
280        by the LLVM fuzzer.
281    
282    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
283        other kinds of group caused stack overflow at compile time. This bug was
284        discovered by the LLVM fuzzer.
285    
286    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
287        between a subroutine call and its quantifier was incorrectly compiled,
288        leading to buffer overflow or other errors. This bug was discovered by the
289        LLVM fuzzer.
290    
291    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
292        assertion after (?(. The code was failing to check the character after
293        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
294        was discovered by the LLVM fuzzer.
295    
296    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
297        a fixed maximum following a group that contains a subroutine reference was
298        incorrectly compiled and could trigger buffer overflow. This bug was
299        discovered by the LLVM fuzzer.
300    
301    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
302        caused a stack overflow instead of the diagnosis of a non-fixed length
303        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
304    
305    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
306        (e.g. /(?<=\Ka)/) could make pcregrep loop.
307    
308    37. There was a similar problem to 36 in pcretest for global matches.
309    
310    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
311        and a subsequent item in the pattern caused a non-match, backtracking over
312        the repeated \X did not stop, but carried on past the start of the subject,
313        causing reference to random memory and/or a segfault. There were also some
314        other cases where backtracking after \C could crash. This set of bugs was
315        discovered by the LLVM fuzzer.
316    
317    39. The function for finding the minimum length of a matching string could take
318        a very long time if mutual recursion was present many times in a pattern,
319        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
320        method has been implemented. This infelicity was discovered by the LLVM
321        fuzzer.
322    
323    40. Static linking against the PCRE library using the pkg-config module was
324        failing on missing pthread symbols.
325    
326    
327    Version 8.36 26-September-2014
328    ------------------------------
329    
330    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
331        -Wmissing-field-initializers and -Wunused-parameter.
332    
333    2.  The tests for quantifiers being too big (greater than 65535) were being
334        applied after reading the number, and stupidly assuming that integer
335        overflow would give a negative number. The tests are now applied as the
336        numbers are read.
337    
338    3.  Tidy code in pcre_exec.c where two branches that used to be different are
339        now the same.
340    
341    4.  The JIT compiler did not generate match limit checks for certain
342        bracketed expressions with quantifiers. This may lead to exponential
343        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
344        issue should be resolved now.
345    
346    5.  Fixed an issue, which occures when nested alternatives are optimized
347        with table jumps.
348    
349    6.  Inserted two casts and changed some ints to size_t in the light of some
350        reported 64-bit compiler warnings (Bugzilla 1477).
351    
352    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
353        an empty string, which sometimes were behaving incorrectly in the
354        interpreter (though correctly in the JIT matcher). This pcretest input is
355        an example:
356    
357          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
358          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
359    
360        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
361        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
362        for an empty string was breaking the inner loop and carrying on at a lower
363        level, when possessive repeated groups should always return to a higher
364        level as they have no backtrack points in them. The empty string test now
365        occurs at the outer level.
366    
367    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
368        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
369    
370    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
371        doesn't).
372    
373    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
374        to hold a pattern to be incorrectly computed (too small) when there were
375        named back references to duplicated names. This could cause "internal
376        error: code overflow" or "double free or corruption" or other memory
377        handling errors.
378    
379    11. When named subpatterns had the same prefixes, back references could be
380        confused. For example, in this pattern:
381    
382          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
383    
384        the reference to 'Name' was incorrectly treated as a reference to a
385        duplicate name.
386    
387    12. A pattern such as /^s?c/mi8 where the optional character has more than
388        one "other case" was incorrectly compiled such that it would only try to
389        match starting at "c".
390    
391    13. When a pattern starting with \s was studied, VT was not included in the
392        list of possible starting characters; this should have been part of the
393        8.34/18 patch.
394    
395    14. If a character class started [\Qx]... where x is any character, the class
396        was incorrectly terminated at the ].
397    
398    15. If a pattern that started with a caseless match for a character with more
399        than one "other case" was studied, PCRE did not set up the starting code
400        unit bit map for the list of possible characters. Now it does. This is an
401        optimization improvement, not a bug fix.
402    
403    16. The Unicode data tables have been updated to Unicode 7.0.0.
404    
405    17. Fixed a number of memory leaks in pcregrep.
406    
407    18. Avoid a compiler warning (from some compilers) for a function call with
408        a cast that removes "const" from an lvalue by using an intermediate
409        variable (to which the compiler does not object).
410    
411    19. Incorrect code was compiled if a group that contained an internal recursive
412        back reference was optional (had quantifier with a minimum of zero). This
413        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
414        caused segmentation faults because of stack overflows at compile time.
415    
416    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
417        group that is quantified with an indefinite repeat, caused a compile-time
418        loop which used up all the system stack and provoked a segmentation fault.
419        This was not the same bug as 19 above.
420    
421    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
422        Patch by Mike Frysinger.
423    
424    
425    Version 8.35 04-April-2014
426    --------------------------
427    
428    1.  A new flag is set, when property checks are present in an XCLASS.
429        When this flag is not set, PCRE can perform certain optimizations
430        such as studying these XCLASS-es.
431    
432    2.  The auto-possessification of character sets were improved: a normal
433        and an extended character set can be compared now. Furthermore
434        the JIT compiler optimizes more character set checks.
435    
436    3.  Got rid of some compiler warnings for potentially uninitialized variables
437        that show up only when compiled with -O2.
438    
439    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
440        of a match later then the end of the match. The pcretest program was not
441        handling the case sensibly - it was outputting from the start to the next
442        binary zero. It now reports this situation in a message, and outputs the
443        text from the end to the start.
444    
445    5.  Fast forward search is improved in JIT. Instead of the first three
446        characters, any three characters with fixed position can be searched.
447        Search order: first, last, middle.
448    
449    6.  Improve character range checks in JIT. Characters are read by an inprecise
450        function now, which returns with an unknown value if the character code is
451        above a certain threshold (e.g: 256). The only limitation is that the value
452        must be bigger than the threshold as well. This function is useful when
453        the characters above the threshold are handled in the same way.
454    
455    7.  The macros whose names start with RAWUCHAR are placeholders for a future
456        mode in which only the bottom 21 bits of 32-bit data items are used. To
457        make this more memorable for those maintaining the code, the names have
458        been changed to start with UCHAR21, and an extensive comment has been added
459        to their definition.
460    
461    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
462        to the export list in Makefile.am (they were accidentally omitted from the
463        8.34 tarball).
464    
465    9.  The informational output from pcretest used the phrase "starting byte set"
466        which is inappropriate for the 16-bit and 32-bit libraries. As the output
467        for "first char" and "need char" really means "non-UTF-char", I've changed
468        "byte" to "char", and slightly reworded the output. The documentation about
469        these values has also been (I hope) clarified.
470    
471    10. Another JIT related optimization: use table jumps for selecting the correct
472        backtracking path, when more than four alternatives are present inside a
473        bracket.
474    
475    11. Empty match is not possible, when the minimum length is greater than zero,
476        and there is no \K in the pattern. JIT should avoid empty match checks in
477        such cases.
478    
479    12. In a caseless character class with UCP support, when a character with more
480        than one alternative case was not the first character of a range, not all
481        the alternative cases were added to the class. For example, s and \x{17f}
482        are both alternative cases for S: the class [RST] was handled correctly,
483        but [R-T] was not.
484    
485    13. The configure.ac file always checked for pthread support when JIT was
486        enabled. This is not used in Windows, so I have put this test inside a
487        check for the presence of windows.h (which was already tested for).
488    
489    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
490        The algorithm provides a way to skip certain starting offsets, and usually
491        faster than linear prefix searches.
492    
493    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
494        as for 'fr_FR' and 'french'. For some reason, however, it then used the
495        Windows-specific input and output files, which have 'french' screwed in.
496        So this could never have worked. One of the problems with locales is that
497        they aren't always the same. I have now updated RunTest so that it checks
498        the output of the locale test (test 3) against three different output
499        files, and it allows the test to pass if any one of them matches. With luck
500        this should make the test pass on some versions of Solaris where it was
501        failing. Because of the uncertainty, the script did not used to stop if
502        test 3 failed; it now does. If further versions of a French locale ever
503        come to light, they can now easily be added.
504    
505    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
506        there was a message during ./configure, but it did not stop. This now
507        provokes an error. The invalid example in README has been corrected.
508        If a value less than the minimum is given, the minimum value has always
509        been used, but now a warning is given.
510    
511    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
512        was a bug in the test system, which is now fixed. Also, the list of various
513        configurations that are tested for each release did not have one with both
514        16/32 bits and --enable-bar-anycrlf. It now does.
515    
516    18. pcretest was missing "-C bsr" for displaying the \R default setting.
517    
518    19. Little endian PowerPC systems are supported now by the JIT compiler.
519    
520    20. The fast forward newline mechanism could enter to an infinite loop on
521        certain invalid UTF-8 input. Although we don't support these cases
522        this issue can be fixed by a performance optimization.
523    
524    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
525        not take account if existing stack usage. There is now a new global
526        variable called pcre_stack_guard that can be set to point to an external
527        function to check stack availability. It is called at the start of
528        processing every parenthesized group.
529    
530    22. A typo in the code meant that in ungreedy mode the max/min qualifier
531        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
532        not match "ab".
533    
534    23. When UTF was disabled, the JIT program reported some incorrect compile
535        errors. These messages are silenced now.
536    
537    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
538        compiler.
539    
540    25. Change all the temporary files used in RunGrepTest to be different to those
541        used by RunTest so that the tests can be run simultaneously, for example by
542        "make -j check".
543    
544    
545    Version 8.34 15-December-2013
546    -----------------------------
547    
548    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
549        executable memory. Patch inspired by Carsten Klein.
550    
551    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
552        this macro is never tested and has no effect, because the work to support
553        coverage involves only compiling and linking options and special targets in
554        the Makefile. The comment in config.h implied that defining the macro would
555        enable coverage support, which is totally false. There was also support for
556        setting this macro in the CMake files (my fault, I just copied it from
557        configure). SUPPORT_GCOV has now been removed.
558    
559    3.  Make a small performance improvement in strlen16() and strlen32() in
560        pcretest.
561    
562    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
563        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
564        these cases). There was also one in pcretest.c.
565    
566    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
567    
568    6.  In UTF mode, the code for checking whether a group could match an empty
569        string (which is used for indefinitely repeated groups to allow for
570        breaking an infinite loop) was broken when the group contained a repeated
571        negated single-character class with a character that occupied more than one
572        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
573        UTF-8 mode). The effect was undefined: the group might or might not be
574        deemed as matching an empty string, or the program might have crashed.
575    
576    7.  The code for checking whether a group could match an empty string was not
577        recognizing that \h, \H, \v, \V, and \R must match a character.
578    
579    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
580        an empty string. If it can, pcretest shows this in its information output.
581    
582    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
583        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
584        matched by pcre_exec() without using JIT:
585    
586        (a) If the rest of the pattern did not match after a maximal run of
587            grapheme clusters, the code for backing up to try with fewer of them
588            did not always back up over a full grapheme when characters that do not
589            have the modifier quality were involved, e.g. Hangul syllables.
590    
591        (b) If the match point in a subject started with modifier character, and
592            there was no match, the code could incorrectly back up beyond the match
593            point, and potentially beyond the first character in the subject,
594            leading to a segfault or an incorrect match result.
595    
596    10. A conditional group with an assertion condition could lead to PCRE
597        recording an incorrect first data item for a match if no other first data
598        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
599        first data item, and therefore matched "ca" after "c" instead of at the
600        start.
601    
602    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
603        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
604    
605    12. The source of pcregrep now includes z/OS-specific code so that it can be
606        compiled for z/OS as part of the special z/OS distribution.
607    
608    13. Added the -T and -TM options to pcretest.
609    
610    14. The code in pcre_compile.c for creating the table of named capturing groups
611        has been refactored. Instead of creating the table dynamically during the
612        actual compiling pass, the information is remembered during the pre-compile
613        pass (on the stack unless there are more than 20 named groups, in which
614        case malloc() is used) and the whole table is created before the actual
615        compile happens. This has simplified the code (it is now nearly 150 lines
616        shorter) and prepared the way for better handling of references to groups
617        with duplicate names.
618    
619    15. A back reference to a named subpattern when there is more than one of the
620        same name now checks them in the order in which they appear in the pattern.
621        The first one that is set is used for the reference. Previously only the
622        first one was inspected. This change makes PCRE more compatible with Perl.
623    
624    16. Unicode character properties were updated from Unicode 6.3.0.
625    
626    17. The compile-time code for auto-possessification has been refactored, based
627        on a patch by Zoltan Herczeg. It now happens after instead of during
628        compilation. The code is cleaner, and more cases are handled. The option
629        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
630        options in pcretest are provided to set it. It can also be set by
631        (*NO_AUTO_POSSESS) at the start of a pattern.
632    
633    18. The character VT has been added to the default ("C" locale) set of
634        characters that match \s and are generally treated as white space,
635        following this same change in Perl 5.18. There is now no difference between
636        "Perl space" and "POSIX space". Whether VT is treated as white space in
637        other locales depends on the locale.
638    
639    19. The code for checking named groups as conditions, either for being set or
640        for being recursed, has been refactored (this is related to 14 and 15
641        above). Processing unduplicated named groups should now be as fast at
642        numerical groups, and processing duplicated groups should be faster than
643        before.
644    
645    20. Two patches to the CMake build system, by Alexander Barkov:
646    
647          (1) Replace the "source" command by "." in CMakeLists.txt because
648              "source" is a bash-ism.
649    
650          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
651              without these the CMake build does not work on Solaris.
652    
653    21. Perl has changed its handling of \8 and \9. If there is no previously
654        encountered capturing group of those numbers, they are treated as the
655        literal characters 8 and 9 instead of a binary zero followed by the
656        literals. PCRE now does the same.
657    
658    22. Following Perl, added \o{} to specify codepoints in octal, making it
659        possible to specify values greater than 0777 and also making them
660        unambiguous.
661    
662    23. Perl now gives an error for missing closing braces after \x{... instead of
663        treating the string as literal. PCRE now does the same.
664    
665    24. RunTest used to grumble if an inappropriate test was selected explicitly,
666        but just skip it when running all tests. This make it awkward to run ranges
667        of tests when one of them was inappropriate. Now it just skips any
668        inappropriate tests, as it always did when running all tests.
669    
670    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
671        character types such as \d or \w, too many callouts were inserted, and the
672        data that they returned was rubbish.
673    
674    26. In UCP mode, \s was not matching two of the characters that Perl matches,
675        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
676        were matched by \h. The code has now been refactored so that the lists of
677        the horizontal and vertical whitespace characters used for \h and \v (which
678        are defined only in one place) are now also used for \s.
679    
680    27. Add JIT support for the 64 bit TileGX architecture.
681        Patch by Jiong Wang (Tilera Corporation).
682    
683    28. Possessive quantifiers for classes (both explicit and automatically
684        generated) now use special opcodes instead of wrapping in ONCE brackets.
685    
686    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
687        (because it's meaningless), this was not happening when PCRE_CASELESS was
688        set. Not wrong, but inefficient.
689    
690    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
691        properties for \w, \d, etc) is present in a test regex. Otherwise if the
692        test contains no characters greater than 255, Perl doesn't realise it
693        should be using Unicode semantics.
694    
695    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
696        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
697        does in Unicode mode.
698    
699    32. Added the "forbid" facility to pcretest so that putting tests into the
700        wrong test files can sometimes be quickly detected.
701    
702    33. There is now a limit (default 250) on the depth of nesting of parentheses.
703        This limit is imposed to control the amount of system stack used at compile
704        time. It can be changed at build time by --with-parens-nest-limit=xxx or
705        the equivalent in CMake.
706    
707    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
708        errors. Perl warns for these when in warning mode, but PCRE has no facility
709        for giving warnings.
710    
711    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
712        However, this was not working for (?!) because it is optimized to (*FAIL),
713        for which PCRE does not allow quantifiers. The optimization is now disabled
714        when a quantifier follows (?!). I can't see any use for this, but it makes
715        things uniform.
716    
717    36. Perl no longer allows group names to start with digits, so I have made this
718        change also in PCRE. It simplifies the code a bit.
719    
720    37. In extended mode, Perl ignores spaces before a + that indicates a
721        possessive quantifier. PCRE allowed a space before the quantifier, but not
722        before the possessive +. It now does.
723    
724    38. The use of \K (reset reported match start) within a repeated possessive
725        group such as (a\Kb)*+ was not working.
726    
727    40. Document that the same character tables must be used at compile time and
728        run time, and that the facility to pass tables to pcre_exec() and
729        pcre_dfa_exec() is for use only with saved/restored patterns.
730    
731    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
732        features for Builds with MSVC:
733    
734        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
735           builds). The .rc files can be used to set FileDescription and many other
736           attributes.
737    
738        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
739           This allows higher-level build scripts which want .pdb files to avoid
740           hard-coding the exact files needed."
741    
742    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
743        mean "start of word" and "end of word", respectively, as a transition aid.
744    
745    43. A minimizing repeat of a class containing codepoints greater than 255 in
746        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
747        compiled to use the heap for recursion.
748    
749    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
750        is configured.
751    
752    
753    Version 8.33 28-May-2013
754    ------------------------
755    
756    1.  Added 'U' to some constants that are compared to unsigned integers, to
757        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
758        variables that are added to signed variables, to ensure the result is
759        signed and can be negated.
760    
761    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
762        CMake config files.
763    
764    3.  Revise the creation of config.h.generic so that all boolean macros are
765        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
766        overriding via -D on the command line possible.
767    
768    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
769        to unsigned int is reported to make a quite noticeable speed difference in
770        a specific Windows environment. Testing on Linux did also appear to show
771        some benefit (and it is clearly not harmful). Also fixed the definition of
772        Xop which should be unsigned.
773    
774    5.  Related to (4), changing the definition of the intermediate variable cc
775        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
776        improvements.
777    
778    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
779        unnecessary spaces.
780    
781    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
782        and later.
783    
784    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
785    
786    9.  Optimizing fast_forward_start_bits in JIT.
787    
788    10. Adding support for callouts in JIT, and fixing some issues revealed
789        during this work. Namely:
790    
791        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
792    
793        (b) Minimum length was not checked before the matching is started.
794    
795    11. The value of capture_last that is passed to callouts was incorrect in some
796        cases when there was a capture on one path that was subsequently abandoned
797        after a backtrack. Also, the capture_last value is now reset after a
798        recursion, since all captures are also reset in this case.
799    
800    12. The interpreter no longer returns the "too many substrings" error in the
801        case when an overflowing capture is in a branch that is subsequently
802        abandoned after a backtrack.
803    
804    13. In the pathological case when an offset vector of size 2 is used, pcretest
805        now prints out the matched string after a yield of 0 or 1.
806    
807    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
808        Only supported by the JIT compiler at the moment.
809    
810    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
811    
812    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
813        offset of the starting point of the matching process, provided the offsets
814        vector is large enough.
815    
816    17. The \A escape now records a lookbehind value of 1, though its execution
817        does not actually inspect the previous character. This is to ensure that,
818        in partial multi-segment matching, at least one character from the old
819        segment is retained when a new segment is processed. Otherwise, if there
820        are no lookbehinds in the pattern, \A might match incorrectly at the start
821        of a new segment.
822    
823    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
824    
825    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
826        gives some modest performance improvement in 8-bit mode.
827    
828    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
829        be expressed in certain programming languages using Universal Character
830        Names.
831    
832    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
833        which points out that "non characters" are not "characters that may not
834        appear in Unicode strings" but rather "characters that are reserved for
835        internal use and have only local meaning".
836    
837    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
838        there was a conditional group that depended on an assertion, if the
839        assertion was false, the callout that immediately followed the alternation
840        in the condition was skipped when pcre_exec() was used for matching.
841    
842    23. Allow an explicit callout to be inserted before an assertion that is the
843        condition for a conditional group, for compatibility with automatic
844        callouts, which always insert a callout at this point.
845    
846    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
847        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
848    
849    25. (*PRUNE) is now supported by the JIT compiler.
850    
851    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
852    
853    27. Fix the case where there are two or more SKIPs with arguments that may be
854        ignored.
855    
856    28. (*SKIP) is now supported by the JIT compiler.
857    
858    29. (*THEN) is now supported by the JIT compiler.
859    
860    30. Update RunTest with additional test selector options.
861    
862    31. The way PCRE handles backtracking verbs has been changed in two ways.
863    
864        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
865        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
866        backtracking. In some cases this makes it more Perl-compatible, but Perl's
867        rather obscure rules do not always do the same thing.
868    
869        (2) Previously, backtracking verbs were confined within assertions. This is
870        no longer the case for positive assertions, except for (*ACCEPT). Again,
871        this sometimes improves Perl compatibility, and sometimes does not.
872    
873    32. A number of tests that were in test 2 because Perl did things differently
874        have been moved to test 1, because either Perl or PCRE has changed, and
875        these tests are now compatible.
876    
877    32. Backtracking control verbs are now handled in the same way in JIT and
878        interpreter.
879    
880    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
881        contained a forward subroutine reference caused a compile error.
882    
883    34. Auto-detect and optimize limited repetitions in JIT.
884    
885    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
886        blocking (*UTF) etc.
887    
888    36. In the interpreter, maximizing pattern repetitions for characters and
889        character types now use tail recursion, which reduces stack usage.
890    
891    37. The value of the max lookbehind was not correctly preserved if a compiled
892        and saved regex was reloaded on a host of different endianness.
893    
894    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
895        of the compiled pattern block, expand the flags field from 16 to 32 bits
896        because it was almost full.
897    
898    39. Try madvise first before posix_madvise.
899    
900    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
901        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
902        original change locked out all matches of empty strings. This has been
903        changed so that one match of an empty string per line is recognized.
904        Subsequent searches on the same line (for colouring or for --only-matching,
905        for example) do not recognize empty strings.
906    
907    41. Applied a user patch to fix a number of spelling mistakes in comments.
908    
909    42. Data lines longer than 65536 caused pcretest to crash.
910    
911    43. Clarified the data type for length and startoffset arguments for pcre_exec
912        and pcre_dfa_exec in the function-specific man pages, where they were
913        explicitly stated to be in bytes, never having been updated. I also added
914        some clarification to the pcreapi man page.
915    
916    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
917        a segmentation fault.
918    
919    
920    Version 8.32 30-November-2012
921    -----------------------------
922    
923    1.  Improved JIT compiler optimizations for first character search and single
924        character iterators.
925    
926    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
927        Patch by Daniel Richard G.
928    
929    3.  Single character iterator optimizations in the JIT compiler.
930    
931    4.  Improved JIT compiler optimizations for character ranges.
932    
933    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
934        Reported by Giuseppe D'Angelo.
935    
936    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
937        of a line, was being set incorrectly in cases where .* appeared inside
938        atomic brackets at the start of a pattern, or where there was a subsequent
939        *PRUNE or *SKIP.
940    
941    7.  Improved instruction cache flush for POWER/PowerPC.
942        Patch by Daniel Richard G.
943    
944    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
945        grep:
946    
947        (a) There is now no limit to the number of patterns to be matched.
948    
949        (b) An error is given if a pattern is too long.
950    
951        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
952            are now supported.
953    
954        (d) --exclude-from and --include-from (multiple use) have been added.
955    
956        (e) Exclusions and inclusions now apply to all files and directories, not
957            just to those obtained from scanning a directory recursively.
958    
959        (f) Multiple uses of -f and --file-list are now supported.
960    
961        (g) In a Windows environment, the default for -d has been changed from
962            "read" (the GNU grep default) to "skip", because otherwise the presence
963            of a directory in the file list provokes an error.
964    
965        (h) The documentation has been revised and clarified in places.
966    
967    9.  Improve the matching speed of capturing brackets.
968    
969    10. Changed the meaning of \X so that it now matches a Unicode extended
970        grapheme cluster.
971    
972    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
973        out POSIX threads when JIT support is configured.
974    
975    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
976    
977    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
978        structure could go wrong in environments where size_t is not the same size
979        as int.
980    
981    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
982        set.
983    
984    15. The EBCDIC support had decayed; later updates to the code had included
985        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
986        general tidy up of EBCDIC-related issues, and the documentation was also
987        not quite right. There is now a test that can be run on ASCII systems to
988        check some of the EBCDIC-related things (but is it not a full test).
989    
990    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
991        in a small tidy to the code.
992    
993    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
994    
995    18. If the --only-matching (-o) option in pcregrep is specified multiple
996        times, each one causes appropriate output. For example, -o1 -o2 outputs the
997        substrings matched by the 1st and 2nd capturing parentheses. A separating
998        string can be specified by --om-separator (default empty).
999    
1000    19. Improving the first n character searches.
1001    
1002    20. Turn case lists for horizontal and vertical white space into macros so that
1003        they are defined only once.
1004    
1005    21. This set of changes together give more compatible Unicode case-folding
1006        behaviour for characters that have more than one other case when UCP
1007        support is available.
1008    
1009        (a) The Unicode property table now has offsets into a new table of sets of
1010            three or more characters that are case-equivalent. The MultiStage2.py
1011            script that generates these tables (the pcre_ucd.c file) now scans
1012            CaseFolding.txt instead of UnicodeData.txt for character case
1013            information.
1014    
1015        (b) The code for adding characters or ranges of characters to a character
1016            class has been abstracted into a generalized function that also handles
1017            case-independence. In UTF-mode with UCP support, this uses the new data
1018            to handle characters with more than one other case.
1019    
1020        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
1021            whose other case is greater than 256 are now correctly matched
1022            caselessly. Previously, the high codepoint matched the low one, but not
1023            vice versa.
1024    
1025        (d) The processing of \h, \H, \v, and \ in character classes now makes use
1026            of the new class addition function, using character lists defined as
1027            macros alongside the case definitions of 20 above.
1028    
1029        (e) Caseless back references now work with characters that have more than
1030            one other case.
1031    
1032        (f) General caseless matching of characters with more than one other case
1033            is supported.
1034    
1035    22. Unicode character properties were updated from Unicode 6.2.0
1036    
1037    23. Improved CMake support under Windows. Patch by Daniel Richard G.
1038    
1039    24. Add support for 32-bit character strings, and UTF-32
1040    
1041    25. Major JIT compiler update (code refactoring and bugfixing).
1042        Experimental Sparc 32 support is added.
1043    
1044    26. Applied a modified version of Daniel Richard G's patch to create
1045        pcre.h.generic and config.h.generic by "make" instead of in the
1046        PrepareRelease script.
1047    
1048    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
1049        pcre_compile.c when checking for a zero character.
1050    
1051    28. Introducing a native interface for JIT. Through this interface, the compiled
1052        machine code can be directly executed. The purpose of this interface is to
1053        provide fast pattern matching, so several sanity checks are not performed.
1054        However, feature tests are still performed. The new interface provides
1055        1.4x speedup compared to the old one.
1056    
1057    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
1058        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
1059        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
1060    
1061    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
1062        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
1063        the "old" RFC 2279). Instead, it ended up passing a negative length to
1064        pcre_exec().
1065    
1066    31. Add support for GCC's visibility feature to hide internal functions.
1067    
1068    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
1069        "unknown -C option" after outputting 0 or 1.
1070    
1071    33. There is now support for generating a code coverage report for the test
1072        suite in environments where gcc is the compiler and lcov is installed. This
1073        is mainly for the benefit of the developers.
1074    
1075    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
1076        unaddressable using valgrind annotations, allowing valgrind to detect
1077        invalid memory accesses. This is mainly for the benefit of the developers.
1078    
1079    25. (*UTF) can now be used to start a pattern in any of the three libraries.
1080    
1081    26. Give configure error if --enable-cpp but no C++ compiler found.
1082    
1083    
1084    Version 8.31 06-July-2012
1085    -------------------------
1086    
1087    1.  Fixing a wrong JIT test case and some compiler warnings.
1088    
1089    2.  Removed a bashism from the RunTest script.
1090    
1091    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
1092        to unsigned type, result still unsigned" that was given by an MS compiler
1093        on encountering the code "-sizeof(xxx)".
1094    
1095    4.  Partial matching support is added to the JIT compiler.
1096    
1097    5.  Fixed several bugs concerned with partial matching of items that consist
1098        of more than one character:
1099    
1100        (a) /^(..)\1/ did not partially match "aba" because checking references was
1101            done on an "all or nothing" basis. This also applied to repeated
1102            references.
1103    
1104        (b) \R did not give a hard partial match if \r was found at the end of the
1105            subject.
1106    
1107        (c) \X did not give a hard partial match after matching one or more
1108            characters at the end of the subject.
1109    
1110        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1111            a partial match for the string "\r".
1112    
1113        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1114            a partial match for a CR character at the end of the subject string.
1115    
1116    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1117        running pcretest, the text "(JIT)" added to the output whenever JIT is
1118        actually used to run the match.
1119    
1120    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1121        or /S+[+] with a digit between 1 and 7.
1122    
1123    8.  OP_NOT now supports any UTF character not just single-byte ones.
1124    
1125    9.  (*MARK) control verb is now supported by the JIT compiler.
1126    
1127    10. The command "./RunTest list" lists the available tests without actually
1128        running any of them. (Because I keep forgetting what they all are.)
1129    
1130    11. Add PCRE_INFO_MAXLOOKBEHIND.
1131    
1132    12. Applied a (slightly modified) user-supplied patch that improves performance
1133        when the heap is used for recursion (compiled with --disable-stack-for-
1134        recursion). Instead of malloc and free for each heap frame each time a
1135        logical recursion happens, frames are retained on a chain and re-used where
1136        possible. This sometimes gives as much as 30% improvement.
1137    
1138    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1139        call.
1140    
1141    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1142    
1143    15. It is now possible to link pcretest with libedit as an alternative to
1144        libreadline.
1145    
1146    16. (*COMMIT) control verb is now supported by the JIT compiler.
1147    
1148    17. The Unicode data tables have been updated to Unicode 6.1.0.
1149    
1150    18. Added --file-list option to pcregrep.
1151    
1152    19. Added binary file support to pcregrep, including the -a, --binary-files,
1153        -I, and --text options.
1154    
1155    20. The madvise function is renamed for posix_madvise for QNX compatibility
1156        reasons. Fixed by Giuseppe D'Angelo.
1157    
1158    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1159    
1160    22. Changed the option for creating symbolic links for 16-bit man pages from
1161        -s to -sf so that re-installing does not cause issues.
1162    
1163    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1164    
1165    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1166        matches in certain environments (the workspace was not being correctly
1167        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1168        some of the workspace data at the beginning of a restart.
1169    
1170    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1171        was not doing so when it should - probably a typo introduced by SVN 528
1172        (change 8.10/14).
1173    
1174    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1175        \w+ when the character tables indicated that \x{c4} was a word character.
1176        There were several related cases, all because the tests for doing a table
1177        lookup were testing for characters less than 127 instead of 255.
1178    
1179    27. If a pattern contains capturing parentheses that are not used in a match,
1180        their slots in the ovector are set to -1. For those that are higher than
1181        any matched groups, this happens at the end of processing. In the case when
1182        there were back references that the ovector was too small to contain
1183        (causing temporary malloc'd memory to be used during matching), and the
1184        highest capturing number was not used, memory off the end of the ovector
1185        was incorrectly being set to -1. (It was using the size of the temporary
1186        memory instead of the true size.)
1187    
1188    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1189        ovector size, it uses memory at the end of the block that it has got.
1190    
1191    29. Check for an overlong MARK name and give an error at compile time. The
1192        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1193    
1194    30. JIT compiler update.
1195    
1196    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1197        Rill for the patch.
1198    
1199    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1200    
1201    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1202    
1203    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1204        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1205        was enabled.
1206    
1207    35. Improve JIT code generation for greedy plus quantifier.
1208    
1209    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1210        1 to "aa" instead of to an empty string. The bug affected repeated groups
1211        that could potentially match an empty string.
1212    
1213    37. Optimizing single character iterators in JIT.
1214    
1215    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1216        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1217        codepoints that are too big for the mode are faulted, and in a UTF mode,
1218        disallowed codepoints are also faulted.
1219    
1220    39. If PCRE was compiled with UTF support, in three places in the DFA
1221        matcher there was code that should only have been obeyed in UTF mode, but
1222        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1223        processing when bytes with values greater than 127 were present. In 16-bit
1224        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1225        both cases the values are those that cannot be the first data item in a UTF
1226        character. The three items that might have provoked this were recursions,
1227        possessively repeated groups, and atomic groups.
1228    
1229    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1230        and pcregrep, because some OS require shared objects to be explicitly
1231        passed to ld, causing the link step to fail if they are not.
1232    
1233    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1234        mode, patterns that started with \h* or \R* might be incorrectly matched.
1235    
1236    
1237    Version 8.30 04-February-2012
1238    -----------------------------
1239    
1240    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1241        name is defined in ctype.h.
1242    
1243    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1244        only in quite long subpatterns.
1245    
1246    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1247        since it was replaced by pcre_fullinfo() in February 2000.
1248    
1249    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1250        match a (*MARK), and the match failed at the start of the subject, a
1251        reference to memory before the start of the subject could occur. This bug
1252        was introduced by fix 17 of release 8.21.
1253    
1254    5.  A reference to an unset group with zero minimum repetition was giving
1255        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1256        /(another)?(\1?)test/ matched against "hello world test". This bug was
1257        introduced in release 8.13.
1258    
1259    6.  Add support for 16-bit character strings (a large amount of work involving
1260        many changes and refactorings).
1261    
1262    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1263        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1264        from a file.
1265    
1266    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1267        rounding is not applied in this particular case).
1268    
1269    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1270        if they appear, or are escaped, in patterns.
1271    
1272    10. Get rid of a number of -Wunused-but-set-variable warnings.
1273    
1274    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1275        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1276        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1277        also returns the mark "x". This bug applied to capturing parentheses,
1278        non-capturing parentheses, and atomic parentheses. It also applied to some
1279        assertions.
1280    
1281    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1282        information out of configure.ac instead of relying on pcre.h.generic, which
1283        is not stored in the repository.
1284    
1285    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1286        -lreadline.
1287    
1288    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1289    
1290    15. Applied Graycode's patch to put the top-level frame on the stack rather
1291        than the heap when not using the stack for recursion. This gives a
1292        performance improvement in many cases when recursion is not deep.
1293    
1294    16. Experimental code added to "pcretest -C" to output the stack frame size.
1295    
1296    
1297    Version 8.21 12-Dec-2011
1298    ------------------------
1299    
1300    1.  Updating the JIT compiler.
1301    
1302    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1303        are added as well.
1304    
1305    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1306        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1307        calling _pcre_jit_exec. Some extra comments are added.
1308    
1309    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1310        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1311        was introduced by change 18 for 8.20.
1312    
1313    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1314        ECMA-262 standard.
1315    
1316    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1317        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1318        This bug was probably introduced by change 9 of 8.13.
1319    
1320    7.  While fixing 6 above, I noticed that a number of other items were being
1321        incorrectly rejected as "not fixed length". This arose partly because newer
1322        opcodes had not been added to the fixed-length checking code. I have (a)
1323        corrected the bug and added tests for these items, and (b) arranged for an
1324        error to occur if an unknown opcode is encountered while checking for fixed
1325        length instead of just assuming "not fixed length". The items that were
1326        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1327        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1328        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1329    
1330    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1331        being incorrectly compiled and would have given unpredicatble results.
1332    
1333    9.  A possessively repeated subpattern with minimum repeat count greater than
1334        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1335        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1336        the first (A) could occur when it should not.
1337    
1338    10. Add a cast and remove a redundant test from the code.
1339    
1340    11. JIT should use pcre_malloc/pcre_free for allocation.
1341    
1342    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1343        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1344        is anything other than /usr, -L is still shown).
1345    
1346    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1347    
1348    14. Perl does not support \N without a following name in a [] class; PCRE now
1349        also gives an error.
1350    
1351    15. If a forward reference was repeated with an upper limit of around 2000,
1352        it caused the error "internal error: overran compiling workspace". The
1353        maximum number of forward references (including repeats) was limited by the
1354        internal workspace, and dependent on the LINK_SIZE. The code has been
1355        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1356        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1357        of around 200,000 forward references. While doing this, I also speeded up
1358        the filling in of repeated forward references.
1359    
1360    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1361        incorrectly expecting the subject to contain another "a" after the start.
1362    
1363    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1364        in the match, the SKIP should be ignored. This was not happening; instead
1365        the SKIP was being treated as NOMATCH. For patterns such as
1366        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1367        tested.
1368    
1369    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1370        now much more compatible with Perl, in particular in cases where the result
1371        is a non-match for a non-anchored pattern. For example, if
1372        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1373        "m", where previously it did not return a name. A side effect of this
1374        change is that for partial matches, the last encountered mark name is
1375        returned, as for non matches. A number of tests that were previously not
1376        Perl-compatible have been moved into the Perl-compatible test files. The
1377        refactoring has had the pleasing side effect of removing one argument from
1378        the match() function, thus reducing its stack requirements.
1379    
1380    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1381        subsequent uses of /S (without +) incorrectly behaved like /S+.
1382    
1383    21. Retrieve executable code size support for the JIT compiler and fixing
1384        some warnings.
1385    
1386    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1387        not work when the shorter character appeared right at the end of the
1388        subject string.
1389    
1390    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1391        systems.
1392    
1393    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1394        output it when the /M option is used in pcretest.
1395    
1396    25. The CheckMan script was not being included in the distribution. Also, added
1397        an explicit "perl" to run Perl scripts from the PrepareRelease script
1398        because this is reportedly needed in Windows.
1399    
1400    26. If study data was being save in a file and studying had not found a set of
1401        "starts with" bytes for the pattern, the data written to the file (though
1402        never used) was taken from uninitialized memory and so caused valgrind to
1403        complain.
1404    
1405    27. Updated RunTest.bat as provided by Sheri Pierce.
1406    
1407    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1408    
1409    29. Computation of memory usage for the table of capturing group names was
1410        giving an unnecessarily large value.
1411    
1412    
1413    Version 8.20 21-Oct-2011
1414    ------------------------
1415    
1416    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1417        a POSIX class. After further experiments with Perl, which convinced me that
1418        Perl has bugs and confusions, a closing square bracket is no longer allowed
1419        in a POSIX name. This bug also affected patterns with classes that started
1420        with full stops.
1421    
1422    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1423        captured substring, but while checking the failing first alternative,
1424        substring 1 is temporarily captured. If the output vector supplied to
1425        pcre_exec() was not big enough for this capture, the yield of the function
1426        was still zero ("insufficient space for captured substrings"). This cannot
1427        be totally fixed without adding another stack variable, which seems a lot
1428        of expense for a edge case. However, I have improved the situation in cases
1429        such as /(a)(b)x|abc/ matched against "abc", where the return code
1430        indicates that fewer than the maximum number of slots in the ovector have
1431        been set.
1432    
1433    3.  Related to (2) above: when there are more back references in a pattern than
1434        slots in the output vector, pcre_exec() uses temporary memory during
1435        matching, and copies in the captures as far as possible afterwards. It was
1436        using the entire output vector, but this conflicts with the specification
1437        that only 2/3 is used for passing back captured substrings. Now it uses
1438        only the first 2/3, for compatibility. This is, of course, another edge
1439        case.
1440    
1441    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1442        main code base, and can be used by building with --enable-jit. When this is
1443        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1444        runtime --no-jit option is given.
1445    
1446    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1447        ovector, the return from the function was zero, implying that there were
1448        other matches that did not fit. The correct "exactly full" value is now
1449        returned.
1450    
1451    6.  If a subpattern that was called recursively or as a subroutine contained
1452        (*PRUNE) or any other control that caused it to give a non-standard return,
1453        invalid errors such as "Error -26 (nested recursion at the same subject
1454        position)" or even infinite loops could occur.
1455    
1456    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1457        computing the minimum length on reaching *ACCEPT, and so ended up with the
1458        wrong value of 1 rather than 0. Further investigation indicates that
1459        computing a minimum subject length in the presence of *ACCEPT is difficult
1460        (think back references, subroutine calls), and so I have changed the code
1461        so that no minimum is registered for a pattern that contains *ACCEPT.
1462    
1463    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1464        it was not handled as intended. [But see 16 below.]
1465    
1466    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1467        Sheri Pierce.
1468    
1469    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1470        the first byte in a match must be "a".
1471    
1472    11. Change 17 for 8.13 increased the recursion depth for patterns like
1473        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1474        pattern contains any instances of (*THEN). If it does not, the old
1475        optimizations are restored. It would be nice to do this on a per-group
1476        basis, but at the moment that is not feasible.
1477    
1478    12. In some environments, the output of pcretest -C is CRLF terminated. This
1479        broke RunTest's code that checks for the link size. A single white space
1480        character after the value is now allowed for.
1481    
1482    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1483        For "fr", it uses the Windows-specific input and output files.
1484    
1485    14. If (*THEN) appeared in a group that was called recursively or as a
1486        subroutine, it did not work as intended. [But see next item.]
1487    
1488    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1489        pattern fragments (but not containing any | characters). If A and B are
1490        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1491        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1492        D. In other words, Perl considers parentheses that do not contain any |
1493        characters to be part of a surrounding alternative, whereas PCRE was
1494        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1495        differently. PCRE now behaves in the same way as Perl, except in the case
1496        of subroutine/recursion calls such as (?1) which have in any case always
1497        been different (but PCRE had them first :-).
1498    
1499    16. Related to 15 above: Perl does not treat the | in a conditional group as
1500        creating alternatives. Such a group is treated in the same way as an
1501        ordinary group without any | characters when processing (*THEN). PCRE has
1502        been changed to match Perl's behaviour.
1503    
1504    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1505        RunGrepTest script failed.
1506    
1507    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1508        inevitable for groups that contain captures, but it can lead to a lot of
1509        stack use in large patterns. The old behaviour has been restored for atomic
1510        groups that do not contain any capturing parentheses.
1511    
1512    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1513        suppress the check for a minimum subject length at run time. (If it was
1514        given to pcre_exec() or pcre_dfa_exec() it did work.)
1515    
1516    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1517        fail to work when decoding hex characters in data strings in EBCDIC
1518        environments.
1519    
1520    21. It appears that in at least one Mac OS environment, the isxdigit() function
1521        is implemented as a macro that evaluates to its argument more than once,
1522        contravening the C 90 Standard (I haven't checked a later standard). There
1523        was an instance in pcretest which caused it to go wrong when processing
1524        \x{...} escapes in subject strings. The has been rewritten to avoid using
1525        things like p++ in the argument of isxdigit().
1526    
1527    
1528    Version 8.13 16-Aug-2011
1529    ------------------------
1530    
1531    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1532    
1533    2.  Two minor typos in pcre_internal.h have been fixed.
1534    
1535    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1536        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1537        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1538    
1539    4.  There were a number of related bugs in the code for matching backrefences
1540        caselessly in UTF-8 mode when codes for the characters concerned were
1541        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1542        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1543        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1544        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1545        2-byte code at the end of the subject (it thought there wasn't enough data
1546        left).
1547    
1548    5.  Comprehensive information about what went wrong is now returned by
1549        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1550        as the output vector has at least 2 elements. The offset of the start of
1551        the failing character and a reason code are placed in the vector.
1552    
1553    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1554        now returned is for the first byte of the failing character, instead of the
1555        last byte inspected. This is an incompatible change, but I hope it is small
1556        enough not to be a problem. It makes the returned offset consistent with
1557        pcre_exec() and pcre_dfa_exec().
1558    
1559    7.  pcretest now gives a text phrase as well as the error number when
1560        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1561        failure, the offset and reason code are output.
1562    
1563    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1564        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1565        back over a single character (\n). This seems wrong (because it treated the
1566        two characters as a single entity when going forwards), conflicts with the
1567        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1568        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1569        has been changed.
1570    
1571    9.  Some internal refactoring has changed the processing so that the handling
1572        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1573        time (the PCRE_DOTALL option was changed this way some time ago: version
1574        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1575        which was always a bit of a fudge. It also means that there is one less
1576        argument for the match() function, which reduces its stack requirements
1577        slightly. This change also fixes an incompatibility with Perl: the pattern
1578        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1579    
1580    10. More internal refactoring has drastically reduced the number of recursive
1581        calls to match() for possessively repeated groups such as (abc)++ when
1582        using pcre_exec().
1583    
1584    11. While implementing 10, a number of bugs in the handling of groups were
1585        discovered and fixed:
1586    
1587        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1588        (a|)*(?1) gave a compile-time internal error.
1589        ((a|)+)+  did not notice that the outer group could match an empty string.
1590        (^a|^)+   was not marked as anchored.
1591        (.*a|.*)+ was not marked as matching at start or after a newline.
1592    
1593    12. Yet more internal refactoring has removed another argument from the match()
1594        function. Special calls to this function are now indicated by setting a
1595        value in a variable in the "match data" data block.
1596    
1597    13. Be more explicit in pcre_study() instead of relying on "default" for
1598        opcodes that mean there is no starting character; this means that when new
1599        ones are added and accidentally left out of pcre_study(), testing should
1600        pick them up.
1601    
1602    14. The -s option of pcretest has been documented for ages as being an old
1603        synonym of -m (show memory usage). I have changed it to mean "force study
1604        for every regex", that is, assume /S for every regex. This is similar to -i
1605        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1606        using it. It makes it easier to run collections of tests with and without
1607        study enabled, and thereby test pcre_study() more easily. All the standard
1608        tests are now run with and without -s (but some patterns can be marked as
1609        "never study" - see 20 below).
1610    
1611    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1612        restoration of the capturing data to the outer values was not happening
1613        correctly.
1614    
1615    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1616        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1617        pattern had matched an empty string, and so incorrectly returned a no
1618        match.
1619    
1620    17. There was optimizing code for the last branch of non-capturing parentheses,
1621        and also for the obeyed branch of a conditional subexpression, which used
1622        tail recursion to cut down on stack usage. Unfortunately, now that there is
1623        the possibility of (*THEN) occurring in these branches, tail recursion is
1624        no longer possible because the return has to be checked for (*THEN). These
1625        two optimizations have therefore been removed. [But see 8.20/11 above.]
1626    
1627    18. If a pattern containing \R was studied, it was assumed that \R always
1628        matched two bytes, thus causing the minimum subject length to be
1629        incorrectly computed because \R can also match just one byte.
1630    
1631    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1632        was incorrectly computed.
1633    
1634    20. If /S is present twice on a test pattern in pcretest input, it now
1635        *disables* studying, thereby overriding the use of -s on the command line
1636        (see 14 above). This is necessary for one or two tests to keep the output
1637        identical in both cases.
1638    
1639    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1640        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1641    
1642    22. When an atomic group that contained a capturing parenthesis was
1643        successfully matched, but the branch in which it appeared failed, the
1644        capturing was not being forgotten if a higher numbered group was later
1645        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1646        group 1 to "a", when in fact it should be unset. This applied to multi-
1647        branched capturing and non-capturing groups, repeated or not, and also to
1648        positive assertions (capturing in negative assertions does not happen
1649        in PCRE) and also to nested atomic groups.
1650    
1651    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1652        subject after a captured substring, to make it easier to tell which of a
1653        number of identical substrings has been captured.
1654    
1655    24. The way atomic groups are processed by pcre_exec() has been changed so that
1656        if they are repeated, backtracking one repetition now resets captured
1657        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1658        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1659        "aaa". Previously, it would have been "a". As part of this code
1660        refactoring, the way recursive calls are handled has also been changed.
1661    
1662    25. If an assertion condition captured any substrings, they were not passed
1663        back unless some other capturing happened later. For example, if
1664        (?(?=(a))a) was matched against "a", no capturing was returned.
1665    
1666    26. When studying a pattern that contained subroutine calls or assertions,
1667        the code for finding the minimum length of a possible match was handling
1668        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1669        group 1 called group 2 while simultaneously a separate group 2 called group
1670        1). A stack overflow occurred in this case. I have fixed this by limiting
1671        the recursion depth to 10.
1672    
1673    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1674        Fortmann. This supports explicit test numbers on the command line, and has
1675        argument validation and error reporting.
1676    
1677    28. An instance of \X with an unlimited repeat could fail if at any point the
1678        first character it looked at was a mark character.
1679    
1680    29. Some minor code refactoring concerning Unicode properties and scripts
1681        should reduce the stack requirement of match() slightly.
1682    
1683    30. Added the '=' option to pcretest to check the setting of unused capturing
1684        slots at the end of the pattern, which are documented as being -1, but are
1685        not included in the return count.
1686    
1687    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1688        compiled something random. Now it gives a compile-time error (as does
1689        Perl).
1690    
1691    32. A *MARK encountered during the processing of a positive assertion is now
1692        recorded and passed back (compatible with Perl).
1693    
1694    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1695        had alternative anchored branches, the search for a second match in a line
1696        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1697        matched the line "0102" twice. The same bug affected patterns that started
1698        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1699        twice.
1700    
1701    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1702        does, and because of capturing effects, quantifying parenthesized
1703        assertions may at times be useful. Quantifiers are now allowed for
1704        parenthesized assertions.
1705    
1706    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1707    
1708    36. \g was being checked for fancy things in a character class, when it should
1709        just be a literal "g".
1710    
1711    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1712        appearance of a nested POSIX class supersedes an apparent external class.
1713        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1714        unescaped square brackets may also appear as part of class names. For
1715        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1716        more like Perl. (But see 8.20/1 above.)
1717    
1718    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1719        was because it thought it was \N{name}, which is not supported).
1720    
1721    39. Add minix to OS list not supporting the -S option in pcretest.
1722    
1723    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1724        cannot analyze patterns in sufficient detail to catch mutual recursions
1725        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1726        subgroup is called recursively as a subpattern for a second time at the
1727        same position in the subject string. In previous releases this might have
1728        been caught by the recursion limit, or it might have run out of stack.
1729    
1730    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1731        happen only once. PCRE was, however incorrectly giving a compile time error
1732        "recursive call could loop indefinitely" because it cannot analyze the
1733        pattern in sufficient detail. The compile time test no longer happens when
1734        PCRE is compiling a conditional subpattern, but actual runaway loops are
1735        now caught at runtime (see 40 above).
1736    
1737    42. It seems that Perl allows any characters other than a closing parenthesis
1738        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1739        has been changed to be the same.
1740    
1741    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1742        as not to get warnings when autogen.sh is called. Also changed
1743        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1744    
1745    44. To help people who use pcregrep to scan files containing exceedingly long
1746        lines, the following changes have been made:
1747    
1748        (a) The default value of the buffer size parameter has been increased from
1749            8K to 20K. (The actual buffer used is three times this size.)
1750    
1751        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1752            PCRE is built.
1753    
1754        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1755            to be set at run time.
1756    
1757        (d) Numerical values in pcregrep options can be followed by K or M, for
1758            example --buffer-size=50K.
1759    
1760        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1761            given and the return code is set to 2.
1762    
1763    45. Add a pointer to the latest mark to the callout data block.
1764    
1765    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1766        partial match of an empty string instead of no match. This was specific to
1767        the use of ".".
1768    
1769    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1770        complete match instead of a partial match. This bug was dependent on both
1771        the PCRE_UTF8 and PCRE_DOTALL options being set.
1772    
1773    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1774        starting byte set, because \b was not being ignored.
1775    
1776    
1777    Version 8.12 15-Jan-2011
1778    ------------------------
1779    
1780    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1781        checks for such things as part of the documentation building process.
1782    
1783    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1784        --match-limit and --recursion-limit options (added for 8.11). In
1785        particular, this made one of the standard tests fail. (The integer value
1786        went into the wrong half of a long int.)
1787    
1788    3.  If the --colour option was given to pcregrep with -v (invert match), it
1789        did strange things, either producing crazy output, or crashing. It should,
1790        of course, ignore a request for colour when reporting lines that do not
1791        match.
1792    
1793    4.  Another pcregrep bug caused similar problems if --colour was specified with
1794        -M (multiline) and the pattern match finished with a line ending.
1795    
1796    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1797        matched in multiline mode, the following line was shown as part of the
1798        match. This seems wrong, so I have changed it.
1799    
1800    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1801        the check for further matches in the same line (so they could be coloured)
1802        to overrun the end of the current line. If another match was found, it was
1803        incorrectly shown (and then shown again when found in the next line).
1804    
1805    7.  If pcregrep was compiled under Windows, there was a reference to the
1806        function pcregrep_exit() before it was defined. I am assuming this was
1807        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1808        reported by a user. I've moved the definition above the reference.
1809    
1810    
1811    Version 8.11 10-Dec-2010
1812    ------------------------
1813    
1814    1.  (*THEN) was not working properly if there were untried alternatives prior
1815        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1816        backtracked to try for "b" instead of moving to the next alternative branch
1817        at the same level (in this case, to look for "c"). The Perl documentation
1818        is clear that when (*THEN) is backtracked onto, it goes to the "next
1819        alternative in the innermost enclosing group".
1820    
1821    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1822        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1823        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1824        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1825        (*THEN).
1826    
1827    3.  If \s appeared in a character class, it removed the VT character from
1828        the class, even if it had been included by some previous item, for example
1829        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1830        of \s, but is part of the POSIX "space" class.)
1831    
1832    4.  A partial match never returns an empty string (because you can always
1833        match an empty string at the end of the subject); however the checking for
1834        an empty string was starting at the "start of match" point. This has been
1835        changed to the "earliest inspected character" point, because the returned
1836        data for a partial match starts at this character. This means that, for
1837        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1838        (previously it gave "no match").
1839    
1840    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1841        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1842        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1843        has an implication that the given string is incomplete (because a partial
1844        match is preferred over a full match). For this reason, these items now
1845        give a partial match in this situation. [Aside: previously, the one case
1846        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1847        match rather than a full match, which was wrong by the old rules, but is
1848        now correct.]
1849    
1850    6.  There was a bug in the handling of #-introduced comments, recognized when
1851        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1852        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1853        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1854        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1855        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1856        places in pcre_compile().
1857    
1858    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1859        comments when looking ahead for named forward references to subpatterns,
1860        the only newline sequence it recognized was NL. It now handles newlines
1861        according to the set newline convention.
1862    
1863    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1864        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1865        cater for a lack of strerror(). These oversights have been fixed.
1866    
1867    9.  Added --match-limit and --recursion-limit to pcregrep.
1868    
1869    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1870    
1871    11. When the -o option was used, pcregrep was setting a return code of 1, even
1872        when matches were found, and --line-buffered was not being honoured.
1873    
1874    12. Added an optional parentheses number to the -o and --only-matching options
1875        of pcregrep.
1876    
1877    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1878        can match an empty string. The code to do it in pcretest and pcredemo
1879        needed fixing:
1880    
1881        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1882            only one byte after an empty string match just before CRLF (this case
1883            just got forgotten; "any" and "anycrlf" were OK).
1884    
1885        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1886            mode when an empty string match preceded an ASCII character followed by
1887            a non-ASCII character. (The code for advancing by one character rather
1888            than one byte was nonsense.)
1889    
1890        (c) The pcredemo.c sample program did not have any code at all to handle
1891            the cases when CRLF is a valid newline sequence.
1892    
1893    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1894        as a starting offset was within the subject string. There is now a new
1895        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1896        negative or greater than the length of the string. In order to test this,
1897        pcretest is extended to allow the setting of negative starting offsets.
1898    
1899    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1900        starting offset points to the beginning of a UTF-8 character was
1901        unnecessarily clumsy. I tidied it up.
1902    
1903    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1904        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1905    
1906    17. Nobody had reported that the --include_dir option, which was added in
1907        release 7.7 should have been called --include-dir (hyphen, not underscore)
1908        for compatibility with GNU grep. I have changed it to --include-dir, but
1909        left --include_dir as an undocumented synonym, and the same for
1910        --exclude-dir, though that is not available in GNU grep, at least as of
1911        release 2.5.4.
1912    
1913    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1914        characters from a string of bytes) have been redefined so as not to use
1915        loops, in order to improve performance in some environments. At the same
1916        time, I abstracted some of the common code into auxiliary macros to save
1917        repetition (this should not affect the compiled code).
1918    
1919    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1920        compile-time error is now given if \c is not followed by an ASCII
1921        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1922        different, and any byte value is allowed.)
1923    
1924    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1925        START_OPTIMIZE option, which is now allowed at compile time - but just
1926        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1927        to pcregrep and other applications that have no direct access to PCRE
1928        options. The new /Y option in pcretest sets this option when calling
1929        pcre_compile().
1930    
1931    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1932        back references. Groups containing recursive back references were forced to
1933        be atomic by that change, but in the case of named groups, the amount of
1934        memory required was incorrectly computed, leading to "Failed: internal
1935        error: code overflow". This has been fixed.
1936    
1937    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1938        pcretest.c, to avoid build problems in some Borland environments.
1939    
1940    
1941    Version 8.10 25-Jun-2010
1942    ------------------------
1943    
1944    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1945        THEN.
1946    
1947    2.  (*ACCEPT) was not working when inside an atomic group.
1948    
1949    3.  Inside a character class, \B is treated as a literal by default, but
1950        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1951        causes the error). The code is unchanged, but I tidied the documentation.
1952    
1953    4.  Inside a character class, PCRE always treated \R and \X as literals,
1954        whereas Perl faults them if its -w option is set. I have changed PCRE so
1955        that it faults them when PCRE_EXTRA is set.
1956    
1957    5.  Added support for \N, which always matches any character other than
1958        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1959    
1960    6.  When compiling pcregrep with newer versions of gcc which may have
1961        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1962        declared with attribute warn_unused_result" were given. Just casting the
1963        result to (void) does not stop the warnings; a more elaborate fudge is
1964        needed. I've used a macro to implement this.
1965    
1966    7.  Minor change to pcretest.c to avoid a compiler warning.
1967    
1968    8.  Added four artifical Unicode properties to help with an option to make
1969        \s etc use properties (see next item). The new properties are: Xan
1970        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1971    
1972    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1973        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1974        this option. Modified pcretest to add /W to test this facility. Added
1975        REG_UCP to make it available via the POSIX interface.
1976    
1977    10. Added --line-buffered to pcregrep.
1978    
1979    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1980        studied, and the match started with a letter with a code point greater than
1981        127 whose first byte was different to the first byte of the other case of
1982        the letter, the other case of this starting letter was not recognized
1983        (#976).
1984    
1985    12. If a pattern that was studied started with a repeated Unicode property
1986        test, for example, \p{Nd}+, there was the theoretical possibility of
1987        setting up an incorrect bitmap of starting bytes, but fortunately it could
1988        not have actually happened in practice until change 8 above was made (it
1989        added property types that matched character-matching opcodes).
1990    
1991    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1992        possible starting bytes for non-anchored patterns.
1993    
1994    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1995        \R, and also a number of cases that involve Unicode properties, both
1996        explicit and implicit when PCRE_UCP is set.
1997    
1998    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1999        input, it could crash or give wrong results if characters with values
2000        greater than 0xc0 were present in the subject string. (Detail: it assumed
2001        UTF-8 input when processing these items.)
2002    
2003    16. Added a lot of (int) casts to avoid compiler warnings in systems where
2004        size_t is 64-bit (#991).
2005    
2006    17. Added a check for running out of memory when PCRE is compiled with
2007        --disable-stack-for-recursion (#990).
2008    
2009    18. If the last data line in a file for pcretest does not have a newline on
2010        the end, a newline was missing in the output.
2011    
2012    19. The default pcre_chartables.c file recognizes only ASCII characters (values
2013        less than 128) in its various bitmaps. However, there is a facility for
2014        generating tables according to the current locale when PCRE is compiled. It
2015        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
2016        characters, are recognized by isspace() and therefore were getting set in
2017        these tables, and indeed these tables seem to approximate to ISO 8859. This
2018        caused a problem in UTF-8 mode when pcre_study() was used to create a list
2019        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
2020        which of course cannot start UTF-8 characters. I have changed the code so
2021        that only real ASCII characters (less than 128) and the correct starting
2022        bytes for UTF-8 encodings are set for characters greater than 127 when in
2023        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
2024        altogether.)
2025    
2026    20. Added the /T option to pcretest so as to be able to run tests with non-
2027        standard character tables, thus making it possible to include the tests
2028        used for 19 above in the standard set of tests.
2029    
2030    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
2031        reference to a subpattern the other side of a comment that contains an
2032        opening parenthesis caused either an internal compiling error, or a
2033        reference to the wrong subpattern.
2034    
2035    
2036    Version 8.02 19-Mar-2010
2037    ------------------------
2038    
2039    1.  The Unicode data tables have been updated to Unicode 5.2.0.
2040    
2041    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
2042        configured.
2043    
2044    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
2045        original author of that file, following a query about its status.
2046    
2047    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
2048        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
2049    
2050    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
2051        quantifier applied to a forward-referencing subroutine call, could compile
2052        incorrect code or give the error "internal error: previously-checked
2053        referenced subpattern not found".
2054    
2055    6.  Both MS Visual Studio and Symbian OS have problems with initializing
2056        variables to point to external functions. For these systems, therefore,
2057        pcre_malloc etc. are now initialized to local functions that call the
2058        relevant global functions.
2059    
2060    7.  There were two entries missing in the vectors called coptable and poptable
2061        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
2062        I've fixed the data, and added a kludgy way of testing at compile time that
2063        the lengths are correct (equal to the number of opcodes).
2064    
2065    8.  Following on from 7, I added a similar kludge to check the length of the
2066        eint vector in pcreposix.c.
2067    
2068    9.  Error texts for pcre_compile() are held as one long string to avoid too
2069        much relocation at load time. To find a text, the string is searched,
2070        counting zeros. There was no check for running off the end of the string,
2071        which could happen if a new error number was added without updating the
2072        string.
2073    
2074    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
2075    
2076    11. \K was not working if it appeared in an atomic group or in a group that
2077        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
2078        \K is "not well defined" if used in an assertion. PCRE now accepts it if
2079        the assertion is positive, but not if it is negative.
2080    
2081    12. Change 11 fortuitously reduced the size of the stack frame used in the
2082        "match()" function of pcre_exec.c by one pointer. Forthcoming
2083        implementation of support for (*MARK) will need an extra pointer on the
2084        stack; I have reserved it now, so that the stack frame size does not
2085        decrease.
2086    
2087    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
2088        item in branch that calls a recursion is a subroutine call - as in the
2089        second branch in the above example - was incorrectly given the compile-
2090        time error "recursive call could loop indefinitely" because pcre_compile()
2091        was not correctly checking the subroutine for matching a non-empty string.
2092    
2093    14. The checks for overrunning compiling workspace could trigger after an
2094        overrun had occurred. This is a "should never occur" error, but it can be
2095        triggered by pathological patterns such as hundreds of nested parentheses.
2096        The checks now trigger 100 bytes before the end of the workspace.
2097    
2098    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
2099    
2100    
2101    Version 8.01 19-Jan-2010
2102    ------------------------
2103    
2104    1.  If a pattern contained a conditional subpattern with only one branch (in
2105        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
2106        computed the wrong minimum data length (which is of course zero for such
2107        subpatterns). This could cause incorrect "no match" results.
2108    
2109    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
2110        the pattern is reset in the first branch, pcre_compile() failed with
2111        "internal error: code overflow at offset...". This happened only when
2112        the reset was to the original external option setting. (An optimization
2113        abstracts leading options settings into an external setting, which was the
2114        cause of this.)
2115    
2116    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
2117        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
2118        assertion pattern did not match (meaning that the assertion was true), it
2119        was incorrectly treated as false if the SKIP had been reached during the
2120        matching. This also applied to assertions used as conditions.
2121    
2122    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
2123        assertion subpattern, including such a pattern used as a condition,
2124        unpredictable results occurred, instead of the error return
2125        PCRE_ERROR_DFA_UITEM.
2126    
2127    5.  The C++ GlobalReplace function was not working like Perl for the special
2128        situation when an empty string is matched. It now does the fancy magic
2129        stuff that is necessary.
2130    
2131    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
2132        removed. (These were left over from very, very early versions of PCRE.)
2133    
2134    7.  Some cosmetic changes to the code to make life easier when compiling it
2135        as part of something else:
2136    
2137        (a) Change DEBUG to PCRE_DEBUG.
2138    
2139        (b) In pcre_compile(), rename the member of the "branch_chain" structure
2140            called "current" as "current_branch", to prevent a collision with the
2141            Linux macro when compiled as a kernel module.
2142    
2143        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
2144            prevent a collision with the Linux macro when compiled as a kernel
2145            module.
2146    
2147    8.  In pcre_compile() there are some checks for integer overflows that used to
2148        cast potentially large values to (double). This has been changed to that
2149        when building, a check for int64_t is made, and if it is found, it is used
2150        instead, thus avoiding the use of floating point arithmetic. (There is no
2151        other use of FP in PCRE.) If int64_t is not found, the fallback is to
2152        double.
2153    
2154    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
2155        2005 (difference between two addresses compared to an unsigned value).
2156    
2157    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
2158        custom one, because of the following reported problem in Windows:
2159    
2160          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
2161              under Win32.
2162          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
2163              therefore missing the function definition.
2164          - The compiler thus generates a "C" signature for the test function.
2165          - The linker fails to find the "C" function.
2166          - PCRE fails to configure if asked to do so against libbz2.
2167    
2168    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
2169        messages were output:
2170    
2171          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
2172          rerunning libtoolize, to keep the correct libtool macros in-tree.
2173          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
2174    
2175        I have done both of these things.
2176    
2177    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
2178        most of the time, it *can* run out if it is given a pattern that contains a
2179        runaway infinite recursion. I updated the discussion in the pcrestack man
2180        page.
2181    
2182    13. Now that we have gone to the x.xx style of version numbers, the minor
2183        version may start with zero. Using 08 or 09 is a bad idea because users
2184        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
2185        interpreted as invalid octal numbers. I've updated the previous comment in
2186        configure.ac, and also added a check that gives an error if 08 or 09 are
2187        used.
2188    
2189    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
2190        causing partial matching to fail when the end of the subject matched \W
2191        in a UTF-8 pattern where \W was quantified with a minimum of 3.
2192    
2193    15. There were some discrepancies between the declarations in pcre_internal.h
2194        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
2195        their definitions. The declarations used "const uschar *" and the
2196        definitions used USPTR. Even though USPTR is normally defined as "const
2197        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
2198        reported that: "This difference in casting confuses some C++ compilers, for
2199        example, SunCC recognizes above declarations as different functions and
2200        generates broken code for hbpcre." I have changed the declarations to use
2201        USPTR.
2202    
2203    16. GNU libtool is named differently on some systems. The autogen.sh script now
2204        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
2205        (FreeBSD).
2206    
2207    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
2208        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
2209        comment: "Figure out how to create a longlong from a string: strtoll and
2210        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
2211        instance, but it only takes 2 args instead of 3!"
2212    
2213    18. A subtle bug concerned with back references has been fixed by a change of
2214        specification, with a corresponding code fix. A pattern such as
2215        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
2216        refers, was giving matches when it shouldn't. For example, xa=xaaa would
2217        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
2218        same bug. Such groups have to be quantified to be useful, or contained
2219        inside another quantified group. (If there's no repetition, the reference
2220        can never match.) The problem arises because, having left the group and
2221        moved on to the rest of the pattern, a later failure that backtracks into
2222        the group uses the captured value from the final iteration of the group
2223        rather than the correct earlier one. I have fixed this in PCRE by forcing
2224        any group that contains a reference to itself to be an atomic group; that
2225        is, there cannot be any backtracking into it once it has completed. This is
2226        similar to recursive and subroutine calls.
2227    
2228    
2229    Version 8.00 19-Oct-09
2230    ----------------------
2231    
2232    1.  The table for translating pcre_compile() error codes into POSIX error codes
2233        was out-of-date, and there was no check on the pcre_compile() error code
2234        being within the table. This could lead to an OK return being given in
2235        error.
2236    
2237    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
2238        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
2239        in a Windows environment.
2240    
2241    3.  The pcregrep --count option prints the count for each file even when it is
2242        zero, as does GNU grep. However, pcregrep was also printing all files when
2243        --files-with-matches was added. Now, when both options are given, it prints
2244        counts only for those files that have at least one match. (GNU grep just
2245        prints the file name in this circumstance, but including the count seems
2246        more useful - otherwise, why use --count?) Also ensured that the
2247        combination -clh just lists non-zero counts, with no names.
2248    
2249    4.  The long form of the pcregrep -F option was incorrectly implemented as
2250        --fixed_strings instead of --fixed-strings. This is an incompatible change,
2251        but it seems right to fix it, and I didn't think it was worth preserving
2252        the old behaviour.
2253    
2254    5.  The command line items --regex=pattern and --regexp=pattern were not
2255        recognized by pcregrep, which required --regex pattern or --regexp pattern
2256        (with a space rather than an '='). The man page documented the '=' forms,
2257        which are compatible with GNU grep; these now work.
2258    
2259    6.  No libpcreposix.pc file was created for pkg-config; there was just
2260        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
2261    
2262    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
2263        when UCP support is not needed, by modifying the Python script that
2264        generates it from Unicode data files. This should not matter if the module
2265        is correctly used as a library, but I received one complaint about 50K of
2266        unwanted data. My guess is that the person linked everything into his
2267        program rather than using a library. Anyway, it does no harm.
2268    
2269    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
2270        was a minimum greater than 1 for a wide character in a possessive
2271        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
2272        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
2273        character. Chaos in the form of incorrect output or a compiling loop could
2274        result.
2275    
2276    9.  The restrictions on what a pattern can contain when partial matching is
2277        requested for pcre_exec() have been removed. All patterns can now be
2278        partially matched by this function. In addition, if there are at least two
2279        slots in the offset vector, the offset of the earliest inspected character
2280        for the match and the offset of the end of the subject are set in them when
2281        PCRE_ERROR_PARTIAL is returned.
2282    
2283    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
2284        synonymous with PCRE_PARTIAL, for backwards compatibility, and
2285        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
2286        and may be more useful for multi-segment matching.
2287    
2288    11. Partial matching with pcre_exec() is now more intuitive. A partial match
2289        used to be given if ever the end of the subject was reached; now it is
2290        given only if matching could not proceed because another character was
2291        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
2292        string "Z", which now yields "no match" instead of "partial match". In the
2293        case of pcre_dfa_exec(), "no match" is given if every matching path for the
2294        final character ended with (*FAIL).
2295    
2296    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
2297        if the pattern had a "must contain" character that was already found in the
2298        earlier partial match, unless partial matching was again requested. For
2299        example, with the pattern /dog.(body)?/, the "must contain" character is
2300        "g". If the first part-match was for the string "dog", restarting with
2301        "sbody" failed. This bug has been fixed.
2302    
2303    13. The string returned by pcre_dfa_exec() after a partial match has been
2304        changed so that it starts at the first inspected character rather than the
2305        first character of the match. This makes a difference only if the pattern
2306        starts with a lookbehind assertion or \b or \B (\K is not supported by
2307        pcre_dfa_exec()). It's an incompatible change, but it makes the two
2308        matching functions compatible, and I think it's the right thing to do.
2309    
2310    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
2311        so that the demonstration program is easily available in environments where
2312        PCRE has not been installed from source.
2313    
2314    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
2315        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
2316        library.
2317    
2318    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
2319        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
2320        is not the first non-POSIX option to be added. Clearly some people find
2321        these options useful.
2322    
2323    17. If a caller to the POSIX matching function regexec() passes a non-zero
2324        value for nmatch with a NULL value for pmatch, the value of
2325        nmatch is forced to zero.
2326    
2327    18. RunGrepTest did not have a test for the availability of the -u option of
2328        the diff command, as RunTest does. It now checks in the same way as
2329        RunTest, and also checks for the -b option.
2330    
2331    19. If an odd number of negated classes containing just a single character
2332        interposed, within parentheses, between a forward reference to a named
2333        subpattern and the definition of the subpattern, compilation crashed with
2334        an internal error, complaining that it could not find the referenced
2335        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
2336        [The bug was that it was starting one character too far in when skipping
2337        over the character class, thus treating the ] as data rather than
2338        terminating the class. This meant it could skip too much.]
2339    
2340    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
2341        /g option in pcretest when the pattern contains \K, which makes it possible
2342        to have an empty string match not at the start, even when the pattern is
2343        anchored. Updated pcretest and pcredemo to use this option.
2344    
2345    21. If the maximum number of capturing subpatterns in a recursion was greater
2346        than the maximum at the outer level, the higher number was returned, but
2347        with unset values at the outer level. The correct (outer level) value is
2348        now given.
2349    
2350    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
2351        PCRE did not set those parentheses (unlike Perl). I have now found a way to
2352        make it do so. The string so far is captured, making this feature
2353        compatible with Perl.
2354    
2355    23. The tests have been re-organized, adding tests 11 and 12, to make it
2356        possible to check the Perl 5.10 features against Perl 5.10.
2357    
2358    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
2359        pattern matches a fixed length string. PCRE did not allow this; now it
2360        does. Neither allows recursion.
2361    
2362    25. I finally figured out how to implement a request to provide the minimum
2363        length of subject string that was needed in order to match a given pattern.
2364        (It was back references and recursion that I had previously got hung up
2365        on.) This code has now been added to pcre_study(); it finds a lower bound
2366        to the length of subject needed. It is not necessarily the greatest lower
2367        bound, but using it to avoid searching strings that are too short does give
2368        some useful speed-ups. The value is available to calling programs via
2369        pcre_fullinfo().
2370    
2371    26. While implementing 25, I discovered to my embarrassment that pcretest had
2372        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
2373        study optimizations had never been tested with that matching function.
2374        Oops. What is worse, even when it was passed study data, there was a bug in
2375        pcre_dfa_exec() that meant it never actually used it. Double oops. There
2376        were also very few tests of studied patterns with pcre_dfa_exec().
2377    
2378    27. If (?| is used to create subpatterns with duplicate numbers, they are now
2379        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
2380        on the other side of the coin, they are no longer allowed to have different
2381        names, because these cannot be distinguished in PCRE, and this has caused
2382        confusion. (This is a difference from Perl.)
2383    
2384    28. When duplicate subpattern names are present (necessarily with different
2385        numbers, as required by 27 above), and a test is made by name in a
2386        conditional pattern, either for a subpattern having been matched, or for
2387        recursion in such a pattern, all the associated numbered subpatterns are
2388        tested, and the overall condition is true if the condition is true for any
2389        one of them. This is the way Perl works, and is also more like the way
2390        testing by number works.
2391    
2392    
2393    Version 7.9 11-Apr-09
2394  ---------------------  ---------------------
2395    
2396  1.  When building with support for bzlib/zlib (pcregrep) and/or readline  1.  When building with support for bzlib/zlib (pcregrep) and/or readline
2397      (pcretest), all targets were linked against these libraries. This included      (pcretest), all targets were linked against these libraries. This included
2398      libpcre, libpcreposix, and libpcrecpp, even though they do not use these      libpcre, libpcreposix, and libpcrecpp, even though they do not use these
2399      libraries. This caused unwanted dependencies to be created. This problem      libraries. This caused unwanted dependencies to be created. This problem
2400      has been fixed, and now only pcregrep is linked with bzlib/zlib and only      has been fixed, and now only pcregrep is linked with bzlib/zlib and only
2401      pcretest is linked with readline.      pcretest is linked with readline.
2402    
2403  2.  The "typedef int BOOL" in pcre_internal.h that was included inside the  2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
2404      "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been      "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
2405      moved outside it again, because FALSE and TRUE are already defined in AIX,      moved outside it again, because FALSE and TRUE are already defined in AIX,
2406      but BOOL is not.      but BOOL is not.
2407    
2408  3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and  3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
2409      PCRE_MATCH_LIMIT_RETURSION values as ints, when they should be long ints.      PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
2410    
2411  4.  The pcregrep documentation said spaces were inserted as well as colons (or  4.  The pcregrep documentation said spaces were inserted as well as colons (or
2412      hyphens) following file names and line numbers when outputting matching      hyphens) following file names and line numbers when outputting matching
2413      lines. This is not true; no spaces are inserted. I have also clarified the      lines. This is not true; no spaces are inserted. I have also clarified the
2414      wording for the --colour (or --color) option.      wording for the --colour (or --color) option.
2415    
2416  5.  In pcregrep, when --colour was used with -o, the list of matching strings  5.  In pcregrep, when --colour was used with -o, the list of matching strings
2417      was not coloured; this is different to GNU grep, so I have changed it to be      was not coloured; this is different to GNU grep, so I have changed it to be
2418      the same.      the same.
2419    
2420  6.  When --colo(u)r was used in pcregrep, only the first matching substring in  6.  When --colo(u)r was used in pcregrep, only the first matching substring in
2421      each matching line was coloured. Now it goes on to look for further matches      each matching line was coloured. Now it goes on to look for further matches
2422      of any of the test patterns, which is the same behaviour as GNU grep.      of any of the test patterns, which is the same behaviour as GNU grep.
2423    
2424  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
2425      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
2426      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this      locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2427      seems to be how GNU grep behaves.      seems to be how GNU grep behaves. [But see later change 40 for release
2428        8.33.]
2429    
2430  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
2431      start or after a newline", because the conditional assertion was not being      start or after a newline", because the conditional assertion was not being
2432      correctly handled. The rule now is that both the assertion and what follows      correctly handled. The rule now is that both the assertion and what follows
2433      in the first alternative must satisfy the test.      in the first alternative must satisfy the test.
2434    
2435  9.  If auto-callout was enabled in a pattern with a conditional group, PCRE  9.  If auto-callout was enabled in a pattern with a conditional group whose
2436      could crash during matching.      condition was an assertion, PCRE could crash during matching, both with
2437        pcre_exec() and pcre_dfa_exec().
2438  10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was  
2439      used for matching.  10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2440        used for matching.
2441  11. Unicode property support in character classes was not working for  
2442    11. Unicode property support in character classes was not working for
2443      characters (bytes) greater than 127 when not in UTF-8 mode.      characters (bytes) greater than 127 when not in UTF-8 mode.
2444    
2445  12. Added the -M command line option to pcretest.  12. Added the -M command line option to pcretest.
2446    
2447  14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.  14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2448    
2449  15. Added the PCRE_NO_START_OPTIMIZE match-time option.  15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2450    
2451    16. Added comments and documentation about mis-use of no_arg in the C++
2452        wrapper.
2453    
2454    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2455        from Martin Jerabek that uses macro names for all relevant character and
2456        string constants.
2457    
2458    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2459        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2460        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2461        these, but not everybody uses configure.
2462    
2463    19. A conditional group that had only one branch was not being correctly
2464        recognized as an item that could match an empty string. This meant that an
2465        enclosing group might also not be so recognized, causing infinite looping
2466        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2467        with the subject "ab", where knowledge that the repeated group can match
2468        nothing is needed in order to break the loop.
2469    
2470    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2471        exec(), but without supplying a callout function, matching went wrong.
2472    
2473    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2474        leak if the size of the offset vector was greater than 30. When the vector
2475        is smaller, the saved offsets during recursion go onto a local stack
2476        vector, but for larger vectors malloc() is used. It was failing to free
2477        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2478        error, in fact).
2479    
2480    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2481        heapframe that is used only when UTF-8 support is enabled. This caused no
2482        problem, but was untidy.
2483    
2484    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2485        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2486        included within another project.
2487    
2488    24. Steven Van Ingelgem's patches to add more options to the CMake support,
2489        slightly modified by me:
2490    
2491          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2492              not building pcregrep.
2493    
2494          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2495              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2496    
2497    25. Forward references, both numeric and by name, in patterns that made use of
2498        duplicate group numbers, could behave incorrectly or give incorrect errors,
2499        because when scanning forward to find the reference group, PCRE was not
2500        taking into account the duplicate group numbers. A pattern such as
2501        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2502    
2503    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2504        the feature of a custom pointer more persuasive (as requested by a user).
2505    
2506    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2507        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2508        pre-defined.
2509    
2510    28. Added support for (*UTF8) at the start of a pattern.
2511    
2512    29. Arrange for flags added by the "release type" setting in CMake to be shown
2513        in the configuration summary.
2514    
2515    
2516  Version 7.8 05-Sep-08  Version 7.8 05-Sep-08
2517  ---------------------  ---------------------
# Line 213  Version 7.7 07-May-08 Line 2668  Version 7.7 07-May-08
2668      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
2669      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
2670      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
2671      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
2672      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
2673      was subsequently set up correctly.)      was subsequently set up correctly.)
2674    

Legend:
Removed from v.389  
changed lines
  Added in v.1600

  ViewVC Help
Powered by ViewVC 1.1.5