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

Diff of /code/trunk/ChangeLog

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

revision 270 by ph10, Fri Nov 16 19:55:16 2007 UTC revision 1593 by zherczeg, Fri Aug 14 09:34:32 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.5 12-Nov-07  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 recursion. This issue was found by Karl Skomski
136        with a custom LLVM fuzzer.
137    
138    
139    Version 8.37 28-April-2015
140    --------------------------
141    
142    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
143        for those parentheses to be closed with whatever has been captured so far.
144        However, it was failing to mark any other groups between the hightest
145        capture so far and the currrent group as "unset". Thus, the ovector for
146        those groups contained whatever was previously there. An example is the
147        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
148    
149    2.  If an assertion condition was quantified with a minimum of zero (an odd
150        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
151    
152    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
153        unrecognized modifier, a crash could occur.
154    
155    4.  An attempt to do global matching in pcretest with a zero-length ovector
156        caused a crash.
157    
158    5.  Fixed a memory leak during matching that could occur for a subpattern
159        subroutine call (recursive or otherwise) if the number of captured groups
160        that had to be saved was greater than ten.
161    
162    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
163        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
164        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
165    
166    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
167        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
168    
169    8. When a pattern is compiled, it remembers the highest back reference so that
170       when matching, if the ovector is too small, extra memory can be obtained to
171       use instead. A conditional subpattern whose condition is a check on a
172       capture having happened, such as, for example in the pattern
173       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
174       setting the highest backreference number. This mattered only if pcre_exec()
175       was called with an ovector that was too small to hold the capture, and there
176       was no other kind of back reference (a situation which is probably quite
177       rare). The effect of the bug was that the condition was always treated as
178       FALSE when the capture could not be consulted, leading to a incorrect
179       behaviour by pcre_exec(). This bug has been fixed.
180    
181    9. A reference to a duplicated named group (either a back reference or a test
182       for being set in a conditional) that occurred in a part of the pattern where
183       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
184       to be incorrectly calculated, leading to overwriting.
185    
186    10. A mutually recursive set of back references such as (\2)(\1) caused a
187        segfault at study time (while trying to find the minimum matching length).
188        The infinite loop is now broken (with the minimum length unset, that is,
189        zero).
190    
191    11. If an assertion that was used as a condition was quantified with a minimum
192        of zero, matching went wrong. In particular, if the whole group had
193        unlimited repetition and could match an empty string, a segfault was
194        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
195        assertions to be quantified, but not if they are being used as conditions,
196        so the above pattern is faulted by Perl. PCRE has now been changed so that
197        it also rejects such patterns.
198    
199    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
200        failed to allow the zero-repeat case if pcre2_exec() was called with an
201        ovector too small to capture the group.
202    
203    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
204        Red Hat Product Security:
205    
206        (a) A crash if /K and /F were both set with the option to save the compiled
207        pattern.
208    
209        (b) Another crash if the option to print captured substrings in a callout
210        was combined with setting a null ovector, for example \O\C+ as a subject
211        string.
212    
213    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
214        forward reference repeated a large (but limited) number of times within a
215        repeated outer group that has a zero minimum quantifier, caused incorrect
216        code to be compiled, leading to the error "internal error:
217        previously-checked referenced subpattern not found" when an incorrect
218        memory address was read. This bug was reported as "heap overflow",
219        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
220        CVE-2015-2325.
221    
222    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
223        call within a group that also contained a recursive back reference caused
224        incorrect code to be compiled. This bug was reported as "heap overflow",
225        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
226        number CVE-2015-2326.
227    
228    24. Computing the size of the JIT read-only data in advance has been a source
229        of various issues, and new ones are still appear unfortunately. To fix
230        existing and future issues, size computation is eliminated from the code,
231        and replaced by on-demand memory allocation.
232    
233    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
234        adjacent to the end of the range, and the range contained characters with
235        more than one other case, caused incorrect behaviour when compiled in UTF
236        mode. In that example, the range a-j was left out of the class.
237    
238    26. Fix JIT compilation of conditional blocks, which assertion
239        is converted to (*FAIL). E.g: /(?(?!))/.
240    
241    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
242        discovered by the LLVM fuzzer.
243    
244    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
245        when this assertion was used as a condition, for example (?(?!)a|b). In
246        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
247        error about an unsupported item.
248    
249    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
250        possessification code could take exponential time to complete. A recursion
251        depth limit of 1000 has been imposed to limit the resources used by this
252        optimization.
253    
254    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
255        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
256        because \S ensures they are all in the class. The code for doing this was
257        interacting badly with the code for computing the amount of space needed to
258        compile the pattern, leading to a buffer overflow. This bug was discovered
259        by the LLVM fuzzer.
260    
261    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
262        other kinds of group caused stack overflow at compile time. This bug was
263        discovered by the LLVM fuzzer.
264    
265    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
266        between a subroutine call and its quantifier was incorrectly compiled,
267        leading to buffer overflow or other errors. This bug was discovered by the
268        LLVM fuzzer.
269    
270    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
271        assertion after (?(. The code was failing to check the character after
272        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
273        was discovered by the LLVM fuzzer.
274    
275    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
276        a fixed maximum following a group that contains a subroutine reference was
277        incorrectly compiled and could trigger buffer overflow. This bug was
278        discovered by the LLVM fuzzer.
279    
280    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
281        caused a stack overflow instead of the diagnosis of a non-fixed length
282        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
283    
284    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
285        (e.g. /(?<=\Ka)/) could make pcregrep loop.
286    
287    37. There was a similar problem to 36 in pcretest for global matches.
288    
289    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
290        and a subsequent item in the pattern caused a non-match, backtracking over
291        the repeated \X did not stop, but carried on past the start of the subject,
292        causing reference to random memory and/or a segfault. There were also some
293        other cases where backtracking after \C could crash. This set of bugs was
294        discovered by the LLVM fuzzer.
295    
296    39. The function for finding the minimum length of a matching string could take
297        a very long time if mutual recursion was present many times in a pattern,
298        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
299        method has been implemented. This infelicity was discovered by the LLVM
300        fuzzer.
301    
302    40. Static linking against the PCRE library using the pkg-config module was
303        failing on missing pthread symbols.
304    
305    
306    Version 8.36 26-September-2014
307    ------------------------------
308    
309    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
310        -Wmissing-field-initializers and -Wunused-parameter.
311    
312    2.  The tests for quantifiers being too big (greater than 65535) were being
313        applied after reading the number, and stupidly assuming that integer
314        overflow would give a negative number. The tests are now applied as the
315        numbers are read.
316    
317    3.  Tidy code in pcre_exec.c where two branches that used to be different are
318        now the same.
319    
320    4.  The JIT compiler did not generate match limit checks for certain
321        bracketed expressions with quantifiers. This may lead to exponential
322        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
323        issue should be resolved now.
324    
325    5.  Fixed an issue, which occures when nested alternatives are optimized
326        with table jumps.
327    
328    6.  Inserted two casts and changed some ints to size_t in the light of some
329        reported 64-bit compiler warnings (Bugzilla 1477).
330    
331    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
332        an empty string, which sometimes were behaving incorrectly in the
333        interpreter (though correctly in the JIT matcher). This pcretest input is
334        an example:
335    
336          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
337          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
338    
339        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
340        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
341        for an empty string was breaking the inner loop and carrying on at a lower
342        level, when possessive repeated groups should always return to a higher
343        level as they have no backtrack points in them. The empty string test now
344        occurs at the outer level.
345    
346    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
347        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
348    
349    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
350        doesn't).
351    
352    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
353        to hold a pattern to be incorrectly computed (too small) when there were
354        named back references to duplicated names. This could cause "internal
355        error: code overflow" or "double free or corruption" or other memory
356        handling errors.
357    
358    11. When named subpatterns had the same prefixes, back references could be
359        confused. For example, in this pattern:
360    
361          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
362    
363        the reference to 'Name' was incorrectly treated as a reference to a
364        duplicate name.
365    
366    12. A pattern such as /^s?c/mi8 where the optional character has more than
367        one "other case" was incorrectly compiled such that it would only try to
368        match starting at "c".
369    
370    13. When a pattern starting with \s was studied, VT was not included in the
371        list of possible starting characters; this should have been part of the
372        8.34/18 patch.
373    
374    14. If a character class started [\Qx]... where x is any character, the class
375        was incorrectly terminated at the ].
376    
377    15. If a pattern that started with a caseless match for a character with more
378        than one "other case" was studied, PCRE did not set up the starting code
379        unit bit map for the list of possible characters. Now it does. This is an
380        optimization improvement, not a bug fix.
381    
382    16. The Unicode data tables have been updated to Unicode 7.0.0.
383    
384    17. Fixed a number of memory leaks in pcregrep.
385    
386    18. Avoid a compiler warning (from some compilers) for a function call with
387        a cast that removes "const" from an lvalue by using an intermediate
388        variable (to which the compiler does not object).
389    
390    19. Incorrect code was compiled if a group that contained an internal recursive
391        back reference was optional (had quantifier with a minimum of zero). This
392        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
393        caused segmentation faults because of stack overflows at compile time.
394    
395    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
396        group that is quantified with an indefinite repeat, caused a compile-time
397        loop which used up all the system stack and provoked a segmentation fault.
398        This was not the same bug as 19 above.
399    
400    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
401        Patch by Mike Frysinger.
402    
403    
404    Version 8.35 04-April-2014
405    --------------------------
406    
407    1.  A new flag is set, when property checks are present in an XCLASS.
408        When this flag is not set, PCRE can perform certain optimizations
409        such as studying these XCLASS-es.
410    
411    2.  The auto-possessification of character sets were improved: a normal
412        and an extended character set can be compared now. Furthermore
413        the JIT compiler optimizes more character set checks.
414    
415    3.  Got rid of some compiler warnings for potentially uninitialized variables
416        that show up only when compiled with -O2.
417    
418    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
419        of a match later then the end of the match. The pcretest program was not
420        handling the case sensibly - it was outputting from the start to the next
421        binary zero. It now reports this situation in a message, and outputs the
422        text from the end to the start.
423    
424    5.  Fast forward search is improved in JIT. Instead of the first three
425        characters, any three characters with fixed position can be searched.
426        Search order: first, last, middle.
427    
428    6.  Improve character range checks in JIT. Characters are read by an inprecise
429        function now, which returns with an unknown value if the character code is
430        above a certain threshold (e.g: 256). The only limitation is that the value
431        must be bigger than the threshold as well. This function is useful when
432        the characters above the threshold are handled in the same way.
433    
434    7.  The macros whose names start with RAWUCHAR are placeholders for a future
435        mode in which only the bottom 21 bits of 32-bit data items are used. To
436        make this more memorable for those maintaining the code, the names have
437        been changed to start with UCHAR21, and an extensive comment has been added
438        to their definition.
439    
440    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
441        to the export list in Makefile.am (they were accidentally omitted from the
442        8.34 tarball).
443    
444    9.  The informational output from pcretest used the phrase "starting byte set"
445        which is inappropriate for the 16-bit and 32-bit libraries. As the output
446        for "first char" and "need char" really means "non-UTF-char", I've changed
447        "byte" to "char", and slightly reworded the output. The documentation about
448        these values has also been (I hope) clarified.
449    
450    10. Another JIT related optimization: use table jumps for selecting the correct
451        backtracking path, when more than four alternatives are present inside a
452        bracket.
453    
454    11. Empty match is not possible, when the minimum length is greater than zero,
455        and there is no \K in the pattern. JIT should avoid empty match checks in
456        such cases.
457    
458    12. In a caseless character class with UCP support, when a character with more
459        than one alternative case was not the first character of a range, not all
460        the alternative cases were added to the class. For example, s and \x{17f}
461        are both alternative cases for S: the class [RST] was handled correctly,
462        but [R-T] was not.
463    
464    13. The configure.ac file always checked for pthread support when JIT was
465        enabled. This is not used in Windows, so I have put this test inside a
466        check for the presence of windows.h (which was already tested for).
467    
468    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
469        The algorithm provides a way to skip certain starting offsets, and usually
470        faster than linear prefix searches.
471    
472    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
473        as for 'fr_FR' and 'french'. For some reason, however, it then used the
474        Windows-specific input and output files, which have 'french' screwed in.
475        So this could never have worked. One of the problems with locales is that
476        they aren't always the same. I have now updated RunTest so that it checks
477        the output of the locale test (test 3) against three different output
478        files, and it allows the test to pass if any one of them matches. With luck
479        this should make the test pass on some versions of Solaris where it was
480        failing. Because of the uncertainty, the script did not used to stop if
481        test 3 failed; it now does. If further versions of a French locale ever
482        come to light, they can now easily be added.
483    
484    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
485        there was a message during ./configure, but it did not stop. This now
486        provokes an error. The invalid example in README has been corrected.
487        If a value less than the minimum is given, the minimum value has always
488        been used, but now a warning is given.
489    
490    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
491        was a bug in the test system, which is now fixed. Also, the list of various
492        configurations that are tested for each release did not have one with both
493        16/32 bits and --enable-bar-anycrlf. It now does.
494    
495    18. pcretest was missing "-C bsr" for displaying the \R default setting.
496    
497    19. Little endian PowerPC systems are supported now by the JIT compiler.
498    
499    20. The fast forward newline mechanism could enter to an infinite loop on
500        certain invalid UTF-8 input. Although we don't support these cases
501        this issue can be fixed by a performance optimization.
502    
503    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
504        not take account if existing stack usage. There is now a new global
505        variable called pcre_stack_guard that can be set to point to an external
506        function to check stack availability. It is called at the start of
507        processing every parenthesized group.
508    
509    22. A typo in the code meant that in ungreedy mode the max/min qualifier
510        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
511        not match "ab".
512    
513    23. When UTF was disabled, the JIT program reported some incorrect compile
514        errors. These messages are silenced now.
515    
516    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
517        compiler.
518    
519    25. Change all the temporary files used in RunGrepTest to be different to those
520        used by RunTest so that the tests can be run simultaneously, for example by
521        "make -j check".
522    
523    
524    Version 8.34 15-December-2013
525    -----------------------------
526    
527    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
528        executable memory. Patch inspired by Carsten Klein.
529    
530    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
531        this macro is never tested and has no effect, because the work to support
532        coverage involves only compiling and linking options and special targets in
533        the Makefile. The comment in config.h implied that defining the macro would
534        enable coverage support, which is totally false. There was also support for
535        setting this macro in the CMake files (my fault, I just copied it from
536        configure). SUPPORT_GCOV has now been removed.
537    
538    3.  Make a small performance improvement in strlen16() and strlen32() in
539        pcretest.
540    
541    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
542        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
543        these cases). There was also one in pcretest.c.
544    
545    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
546    
547    6.  In UTF mode, the code for checking whether a group could match an empty
548        string (which is used for indefinitely repeated groups to allow for
549        breaking an infinite loop) was broken when the group contained a repeated
550        negated single-character class with a character that occupied more than one
551        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
552        UTF-8 mode). The effect was undefined: the group might or might not be
553        deemed as matching an empty string, or the program might have crashed.
554    
555    7.  The code for checking whether a group could match an empty string was not
556        recognizing that \h, \H, \v, \V, and \R must match a character.
557    
558    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
559        an empty string. If it can, pcretest shows this in its information output.
560    
561    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
562        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
563        matched by pcre_exec() without using JIT:
564    
565        (a) If the rest of the pattern did not match after a maximal run of
566            grapheme clusters, the code for backing up to try with fewer of them
567            did not always back up over a full grapheme when characters that do not
568            have the modifier quality were involved, e.g. Hangul syllables.
569    
570        (b) If the match point in a subject started with modifier character, and
571            there was no match, the code could incorrectly back up beyond the match
572            point, and potentially beyond the first character in the subject,
573            leading to a segfault or an incorrect match result.
574    
575    10. A conditional group with an assertion condition could lead to PCRE
576        recording an incorrect first data item for a match if no other first data
577        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
578        first data item, and therefore matched "ca" after "c" instead of at the
579        start.
580    
581    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
582        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
583    
584    12. The source of pcregrep now includes z/OS-specific code so that it can be
585        compiled for z/OS as part of the special z/OS distribution.
586    
587    13. Added the -T and -TM options to pcretest.
588    
589    14. The code in pcre_compile.c for creating the table of named capturing groups
590        has been refactored. Instead of creating the table dynamically during the
591        actual compiling pass, the information is remembered during the pre-compile
592        pass (on the stack unless there are more than 20 named groups, in which
593        case malloc() is used) and the whole table is created before the actual
594        compile happens. This has simplified the code (it is now nearly 150 lines
595        shorter) and prepared the way for better handling of references to groups
596        with duplicate names.
597    
598    15. A back reference to a named subpattern when there is more than one of the
599        same name now checks them in the order in which they appear in the pattern.
600        The first one that is set is used for the reference. Previously only the
601        first one was inspected. This change makes PCRE more compatible with Perl.
602    
603    16. Unicode character properties were updated from Unicode 6.3.0.
604    
605    17. The compile-time code for auto-possessification has been refactored, based
606        on a patch by Zoltan Herczeg. It now happens after instead of during
607        compilation. The code is cleaner, and more cases are handled. The option
608        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
609        options in pcretest are provided to set it. It can also be set by
610        (*NO_AUTO_POSSESS) at the start of a pattern.
611    
612    18. The character VT has been added to the default ("C" locale) set of
613        characters that match \s and are generally treated as white space,
614        following this same change in Perl 5.18. There is now no difference between
615        "Perl space" and "POSIX space". Whether VT is treated as white space in
616        other locales depends on the locale.
617    
618    19. The code for checking named groups as conditions, either for being set or
619        for being recursed, has been refactored (this is related to 14 and 15
620        above). Processing unduplicated named groups should now be as fast at
621        numerical groups, and processing duplicated groups should be faster than
622        before.
623    
624    20. Two patches to the CMake build system, by Alexander Barkov:
625    
626          (1) Replace the "source" command by "." in CMakeLists.txt because
627              "source" is a bash-ism.
628    
629          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
630              without these the CMake build does not work on Solaris.
631    
632    21. Perl has changed its handling of \8 and \9. If there is no previously
633        encountered capturing group of those numbers, they are treated as the
634        literal characters 8 and 9 instead of a binary zero followed by the
635        literals. PCRE now does the same.
636    
637    22. Following Perl, added \o{} to specify codepoints in octal, making it
638        possible to specify values greater than 0777 and also making them
639        unambiguous.
640    
641    23. Perl now gives an error for missing closing braces after \x{... instead of
642        treating the string as literal. PCRE now does the same.
643    
644    24. RunTest used to grumble if an inappropriate test was selected explicitly,
645        but just skip it when running all tests. This make it awkward to run ranges
646        of tests when one of them was inappropriate. Now it just skips any
647        inappropriate tests, as it always did when running all tests.
648    
649    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
650        character types such as \d or \w, too many callouts were inserted, and the
651        data that they returned was rubbish.
652    
653    26. In UCP mode, \s was not matching two of the characters that Perl matches,
654        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
655        were matched by \h. The code has now been refactored so that the lists of
656        the horizontal and vertical whitespace characters used for \h and \v (which
657        are defined only in one place) are now also used for \s.
658    
659    27. Add JIT support for the 64 bit TileGX architecture.
660        Patch by Jiong Wang (Tilera Corporation).
661    
662    28. Possessive quantifiers for classes (both explicit and automatically
663        generated) now use special opcodes instead of wrapping in ONCE brackets.
664    
665    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
666        (because it's meaningless), this was not happening when PCRE_CASELESS was
667        set. Not wrong, but inefficient.
668    
669    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
670        properties for \w, \d, etc) is present in a test regex. Otherwise if the
671        test contains no characters greater than 255, Perl doesn't realise it
672        should be using Unicode semantics.
673    
674    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
675        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
676        does in Unicode mode.
677    
678    32. Added the "forbid" facility to pcretest so that putting tests into the
679        wrong test files can sometimes be quickly detected.
680    
681    33. There is now a limit (default 250) on the depth of nesting of parentheses.
682        This limit is imposed to control the amount of system stack used at compile
683        time. It can be changed at build time by --with-parens-nest-limit=xxx or
684        the equivalent in CMake.
685    
686    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
687        errors. Perl warns for these when in warning mode, but PCRE has no facility
688        for giving warnings.
689    
690    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
691        However, this was not working for (?!) because it is optimized to (*FAIL),
692        for which PCRE does not allow quantifiers. The optimization is now disabled
693        when a quantifier follows (?!). I can't see any use for this, but it makes
694        things uniform.
695    
696    36. Perl no longer allows group names to start with digits, so I have made this
697        change also in PCRE. It simplifies the code a bit.
698    
699    37. In extended mode, Perl ignores spaces before a + that indicates a
700        possessive quantifier. PCRE allowed a space before the quantifier, but not
701        before the possessive +. It now does.
702    
703    38. The use of \K (reset reported match start) within a repeated possessive
704        group such as (a\Kb)*+ was not working.
705    
706    40. Document that the same character tables must be used at compile time and
707        run time, and that the facility to pass tables to pcre_exec() and
708        pcre_dfa_exec() is for use only with saved/restored patterns.
709    
710    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
711        features for Builds with MSVC:
712    
713        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
714           builds). The .rc files can be used to set FileDescription and many other
715           attributes.
716    
717        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
718           This allows higher-level build scripts which want .pdb files to avoid
719           hard-coding the exact files needed."
720    
721    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
722        mean "start of word" and "end of word", respectively, as a transition aid.
723    
724    43. A minimizing repeat of a class containing codepoints greater than 255 in
725        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
726        compiled to use the heap for recursion.
727    
728    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
729        is configured.
730    
731    
732    Version 8.33 28-May-2013
733    ------------------------
734    
735    1.  Added 'U' to some constants that are compared to unsigned integers, to
736        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
737        variables that are added to signed variables, to ensure the result is
738        signed and can be negated.
739    
740    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
741        CMake config files.
742    
743    3.  Revise the creation of config.h.generic so that all boolean macros are
744        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
745        overriding via -D on the command line possible.
746    
747    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
748        to unsigned int is reported to make a quite noticeable speed difference in
749        a specific Windows environment. Testing on Linux did also appear to show
750        some benefit (and it is clearly not harmful). Also fixed the definition of
751        Xop which should be unsigned.
752    
753    5.  Related to (4), changing the definition of the intermediate variable cc
754        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
755        improvements.
756    
757    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
758        unnecessary spaces.
759    
760    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
761        and later.
762    
763    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
764    
765    9.  Optimizing fast_forward_start_bits in JIT.
766    
767    10. Adding support for callouts in JIT, and fixing some issues revealed
768        during this work. Namely:
769    
770        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
771    
772        (b) Minimum length was not checked before the matching is started.
773    
774    11. The value of capture_last that is passed to callouts was incorrect in some
775        cases when there was a capture on one path that was subsequently abandoned
776        after a backtrack. Also, the capture_last value is now reset after a
777        recursion, since all captures are also reset in this case.
778    
779    12. The interpreter no longer returns the "too many substrings" error in the
780        case when an overflowing capture is in a branch that is subsequently
781        abandoned after a backtrack.
782    
783    13. In the pathological case when an offset vector of size 2 is used, pcretest
784        now prints out the matched string after a yield of 0 or 1.
785    
786    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
787        Only supported by the JIT compiler at the moment.
788    
789    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
790    
791    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
792        offset of the starting point of the matching process, provided the offsets
793        vector is large enough.
794    
795    17. The \A escape now records a lookbehind value of 1, though its execution
796        does not actually inspect the previous character. This is to ensure that,
797        in partial multi-segment matching, at least one character from the old
798        segment is retained when a new segment is processed. Otherwise, if there
799        are no lookbehinds in the pattern, \A might match incorrectly at the start
800        of a new segment.
801    
802    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
803    
804    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
805        gives some modest performance improvement in 8-bit mode.
806    
807    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
808        be expressed in certain programming languages using Universal Character
809        Names.
810    
811    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
812        which points out that "non characters" are not "characters that may not
813        appear in Unicode strings" but rather "characters that are reserved for
814        internal use and have only local meaning".
815    
816    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
817        there was a conditional group that depended on an assertion, if the
818        assertion was false, the callout that immediately followed the alternation
819        in the condition was skipped when pcre_exec() was used for matching.
820    
821    23. Allow an explicit callout to be inserted before an assertion that is the
822        condition for a conditional group, for compatibility with automatic
823        callouts, which always insert a callout at this point.
824    
825    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
826        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
827    
828    25. (*PRUNE) is now supported by the JIT compiler.
829    
830    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
831    
832    27. Fix the case where there are two or more SKIPs with arguments that may be
833        ignored.
834    
835    28. (*SKIP) is now supported by the JIT compiler.
836    
837    29. (*THEN) is now supported by the JIT compiler.
838    
839    30. Update RunTest with additional test selector options.
840    
841    31. The way PCRE handles backtracking verbs has been changed in two ways.
842    
843        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
844        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
845        backtracking. In some cases this makes it more Perl-compatible, but Perl's
846        rather obscure rules do not always do the same thing.
847    
848        (2) Previously, backtracking verbs were confined within assertions. This is
849        no longer the case for positive assertions, except for (*ACCEPT). Again,
850        this sometimes improves Perl compatibility, and sometimes does not.
851    
852    32. A number of tests that were in test 2 because Perl did things differently
853        have been moved to test 1, because either Perl or PCRE has changed, and
854        these tests are now compatible.
855    
856    32. Backtracking control verbs are now handled in the same way in JIT and
857        interpreter.
858    
859    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
860        contained a forward subroutine reference caused a compile error.
861    
862    34. Auto-detect and optimize limited repetitions in JIT.
863    
864    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
865        blocking (*UTF) etc.
866    
867    36. In the interpreter, maximizing pattern repetitions for characters and
868        character types now use tail recursion, which reduces stack usage.
869    
870    37. The value of the max lookbehind was not correctly preserved if a compiled
871        and saved regex was reloaded on a host of different endianness.
872    
873    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
874        of the compiled pattern block, expand the flags field from 16 to 32 bits
875        because it was almost full.
876    
877    39. Try madvise first before posix_madvise.
878    
879    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
880        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
881        original change locked out all matches of empty strings. This has been
882        changed so that one match of an empty string per line is recognized.
883        Subsequent searches on the same line (for colouring or for --only-matching,
884        for example) do not recognize empty strings.
885    
886    41. Applied a user patch to fix a number of spelling mistakes in comments.
887    
888    42. Data lines longer than 65536 caused pcretest to crash.
889    
890    43. Clarified the data type for length and startoffset arguments for pcre_exec
891        and pcre_dfa_exec in the function-specific man pages, where they were
892        explicitly stated to be in bytes, never having been updated. I also added
893        some clarification to the pcreapi man page.
894    
895    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
896        a segmentation fault.
897    
898    
899    Version 8.32 30-November-2012
900    -----------------------------
901    
902    1.  Improved JIT compiler optimizations for first character search and single
903        character iterators.
904    
905    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
906        Patch by Daniel Richard G.
907    
908    3.  Single character iterator optimizations in the JIT compiler.
909    
910    4.  Improved JIT compiler optimizations for character ranges.
911    
912    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
913        Reported by Giuseppe D'Angelo.
914    
915    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
916        of a line, was being set incorrectly in cases where .* appeared inside
917        atomic brackets at the start of a pattern, or where there was a subsequent
918        *PRUNE or *SKIP.
919    
920    7.  Improved instruction cache flush for POWER/PowerPC.
921        Patch by Daniel Richard G.
922    
923    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
924        grep:
925    
926        (a) There is now no limit to the number of patterns to be matched.
927    
928        (b) An error is given if a pattern is too long.
929    
930        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
931            are now supported.
932    
933        (d) --exclude-from and --include-from (multiple use) have been added.
934    
935        (e) Exclusions and inclusions now apply to all files and directories, not
936            just to those obtained from scanning a directory recursively.
937    
938        (f) Multiple uses of -f and --file-list are now supported.
939    
940        (g) In a Windows environment, the default for -d has been changed from
941            "read" (the GNU grep default) to "skip", because otherwise the presence
942            of a directory in the file list provokes an error.
943    
944        (h) The documentation has been revised and clarified in places.
945    
946    9.  Improve the matching speed of capturing brackets.
947    
948    10. Changed the meaning of \X so that it now matches a Unicode extended
949        grapheme cluster.
950    
951    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
952        out POSIX threads when JIT support is configured.
953    
954    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
955    
956    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
957        structure could go wrong in environments where size_t is not the same size
958        as int.
959    
960    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
961        set.
962    
963    15. The EBCDIC support had decayed; later updates to the code had included
964        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
965        general tidy up of EBCDIC-related issues, and the documentation was also
966        not quite right. There is now a test that can be run on ASCII systems to
967        check some of the EBCDIC-related things (but is it not a full test).
968    
969    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
970        in a small tidy to the code.
971    
972    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
973    
974    18. If the --only-matching (-o) option in pcregrep is specified multiple
975        times, each one causes appropriate output. For example, -o1 -o2 outputs the
976        substrings matched by the 1st and 2nd capturing parentheses. A separating
977        string can be specified by --om-separator (default empty).
978    
979    19. Improving the first n character searches.
980    
981    20. Turn case lists for horizontal and vertical white space into macros so that
982        they are defined only once.
983    
984    21. This set of changes together give more compatible Unicode case-folding
985        behaviour for characters that have more than one other case when UCP
986        support is available.
987    
988        (a) The Unicode property table now has offsets into a new table of sets of
989            three or more characters that are case-equivalent. The MultiStage2.py
990            script that generates these tables (the pcre_ucd.c file) now scans
991            CaseFolding.txt instead of UnicodeData.txt for character case
992            information.
993    
994        (b) The code for adding characters or ranges of characters to a character
995            class has been abstracted into a generalized function that also handles
996            case-independence. In UTF-mode with UCP support, this uses the new data
997            to handle characters with more than one other case.
998    
999        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
1000            whose other case is greater than 256 are now correctly matched
1001            caselessly. Previously, the high codepoint matched the low one, but not
1002            vice versa.
1003    
1004        (d) The processing of \h, \H, \v, and \ in character classes now makes use
1005            of the new class addition function, using character lists defined as
1006            macros alongside the case definitions of 20 above.
1007    
1008        (e) Caseless back references now work with characters that have more than
1009            one other case.
1010    
1011        (f) General caseless matching of characters with more than one other case
1012            is supported.
1013    
1014    22. Unicode character properties were updated from Unicode 6.2.0
1015    
1016    23. Improved CMake support under Windows. Patch by Daniel Richard G.
1017    
1018    24. Add support for 32-bit character strings, and UTF-32
1019    
1020    25. Major JIT compiler update (code refactoring and bugfixing).
1021        Experimental Sparc 32 support is added.
1022    
1023    26. Applied a modified version of Daniel Richard G's patch to create
1024        pcre.h.generic and config.h.generic by "make" instead of in the
1025        PrepareRelease script.
1026    
1027    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
1028        pcre_compile.c when checking for a zero character.
1029    
1030    28. Introducing a native interface for JIT. Through this interface, the compiled
1031        machine code can be directly executed. The purpose of this interface is to
1032        provide fast pattern matching, so several sanity checks are not performed.
1033        However, feature tests are still performed. The new interface provides
1034        1.4x speedup compared to the old one.
1035    
1036    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
1037        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
1038        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
1039    
1040    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
1041        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
1042        the "old" RFC 2279). Instead, it ended up passing a negative length to
1043        pcre_exec().
1044    
1045    31. Add support for GCC's visibility feature to hide internal functions.
1046    
1047    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
1048        "unknown -C option" after outputting 0 or 1.
1049    
1050    33. There is now support for generating a code coverage report for the test
1051        suite in environments where gcc is the compiler and lcov is installed. This
1052        is mainly for the benefit of the developers.
1053    
1054    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
1055        unaddressable using valgrind annotations, allowing valgrind to detect
1056        invalid memory accesses. This is mainly for the benefit of the developers.
1057    
1058    25. (*UTF) can now be used to start a pattern in any of the three libraries.
1059    
1060    26. Give configure error if --enable-cpp but no C++ compiler found.
1061    
1062    
1063    Version 8.31 06-July-2012
1064    -------------------------
1065    
1066    1.  Fixing a wrong JIT test case and some compiler warnings.
1067    
1068    2.  Removed a bashism from the RunTest script.
1069    
1070    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
1071        to unsigned type, result still unsigned" that was given by an MS compiler
1072        on encountering the code "-sizeof(xxx)".
1073    
1074    4.  Partial matching support is added to the JIT compiler.
1075    
1076    5.  Fixed several bugs concerned with partial matching of items that consist
1077        of more than one character:
1078    
1079        (a) /^(..)\1/ did not partially match "aba" because checking references was
1080            done on an "all or nothing" basis. This also applied to repeated
1081            references.
1082    
1083        (b) \R did not give a hard partial match if \r was found at the end of the
1084            subject.
1085    
1086        (c) \X did not give a hard partial match after matching one or more
1087            characters at the end of the subject.
1088    
1089        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1090            a partial match for the string "\r".
1091    
1092        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1093            a partial match for a CR character at the end of the subject string.
1094    
1095    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1096        running pcretest, the text "(JIT)" added to the output whenever JIT is
1097        actually used to run the match.
1098    
1099    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1100        or /S+[+] with a digit between 1 and 7.
1101    
1102    8.  OP_NOT now supports any UTF character not just single-byte ones.
1103    
1104    9.  (*MARK) control verb is now supported by the JIT compiler.
1105    
1106    10. The command "./RunTest list" lists the available tests without actually
1107        running any of them. (Because I keep forgetting what they all are.)
1108    
1109    11. Add PCRE_INFO_MAXLOOKBEHIND.
1110    
1111    12. Applied a (slightly modified) user-supplied patch that improves performance
1112        when the heap is used for recursion (compiled with --disable-stack-for-
1113        recursion). Instead of malloc and free for each heap frame each time a
1114        logical recursion happens, frames are retained on a chain and re-used where
1115        possible. This sometimes gives as much as 30% improvement.
1116    
1117    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1118        call.
1119    
1120    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1121    
1122    15. It is now possible to link pcretest with libedit as an alternative to
1123        libreadline.
1124    
1125    16. (*COMMIT) control verb is now supported by the JIT compiler.
1126    
1127    17. The Unicode data tables have been updated to Unicode 6.1.0.
1128    
1129    18. Added --file-list option to pcregrep.
1130    
1131    19. Added binary file support to pcregrep, including the -a, --binary-files,
1132        -I, and --text options.
1133    
1134    20. The madvise function is renamed for posix_madvise for QNX compatibility
1135        reasons. Fixed by Giuseppe D'Angelo.
1136    
1137    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1138    
1139    22. Changed the option for creating symbolic links for 16-bit man pages from
1140        -s to -sf so that re-installing does not cause issues.
1141    
1142    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1143    
1144    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1145        matches in certain environments (the workspace was not being correctly
1146        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1147        some of the workspace data at the beginning of a restart.
1148    
1149    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1150        was not doing so when it should - probably a typo introduced by SVN 528
1151        (change 8.10/14).
1152    
1153    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1154        \w+ when the character tables indicated that \x{c4} was a word character.
1155        There were several related cases, all because the tests for doing a table
1156        lookup were testing for characters less than 127 instead of 255.
1157    
1158    27. If a pattern contains capturing parentheses that are not used in a match,
1159        their slots in the ovector are set to -1. For those that are higher than
1160        any matched groups, this happens at the end of processing. In the case when
1161        there were back references that the ovector was too small to contain
1162        (causing temporary malloc'd memory to be used during matching), and the
1163        highest capturing number was not used, memory off the end of the ovector
1164        was incorrectly being set to -1. (It was using the size of the temporary
1165        memory instead of the true size.)
1166    
1167    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1168        ovector size, it uses memory at the end of the block that it has got.
1169    
1170    29. Check for an overlong MARK name and give an error at compile time. The
1171        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1172    
1173    30. JIT compiler update.
1174    
1175    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1176        Rill for the patch.
1177    
1178    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1179    
1180    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1181    
1182    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1183        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1184        was enabled.
1185    
1186    35. Improve JIT code generation for greedy plus quantifier.
1187    
1188    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1189        1 to "aa" instead of to an empty string. The bug affected repeated groups
1190        that could potentially match an empty string.
1191    
1192    37. Optimizing single character iterators in JIT.
1193    
1194    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1195        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1196        codepoints that are too big for the mode are faulted, and in a UTF mode,
1197        disallowed codepoints are also faulted.
1198    
1199    39. If PCRE was compiled with UTF support, in three places in the DFA
1200        matcher there was code that should only have been obeyed in UTF mode, but
1201        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1202        processing when bytes with values greater than 127 were present. In 16-bit
1203        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1204        both cases the values are those that cannot be the first data item in a UTF
1205        character. The three items that might have provoked this were recursions,
1206        possessively repeated groups, and atomic groups.
1207    
1208    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1209        and pcregrep, because some OS require shared objects to be explicitly
1210        passed to ld, causing the link step to fail if they are not.
1211    
1212    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1213        mode, patterns that started with \h* or \R* might be incorrectly matched.
1214    
1215    
1216    Version 8.30 04-February-2012
1217    -----------------------------
1218    
1219    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1220        name is defined in ctype.h.
1221    
1222    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1223        only in quite long subpatterns.
1224    
1225    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1226        since it was replaced by pcre_fullinfo() in February 2000.
1227    
1228    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1229        match a (*MARK), and the match failed at the start of the subject, a
1230        reference to memory before the start of the subject could occur. This bug
1231        was introduced by fix 17 of release 8.21.
1232    
1233    5.  A reference to an unset group with zero minimum repetition was giving
1234        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1235        /(another)?(\1?)test/ matched against "hello world test". This bug was
1236        introduced in release 8.13.
1237    
1238    6.  Add support for 16-bit character strings (a large amount of work involving
1239        many changes and refactorings).
1240    
1241    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1242        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1243        from a file.
1244    
1245    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1246        rounding is not applied in this particular case).
1247    
1248    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1249        if they appear, or are escaped, in patterns.
1250    
1251    10. Get rid of a number of -Wunused-but-set-variable warnings.
1252    
1253    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1254        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1255        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1256        also returns the mark "x". This bug applied to capturing parentheses,
1257        non-capturing parentheses, and atomic parentheses. It also applied to some
1258        assertions.
1259    
1260    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1261        information out of configure.ac instead of relying on pcre.h.generic, which
1262        is not stored in the repository.
1263    
1264    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1265        -lreadline.
1266    
1267    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1268    
1269    15. Applied Graycode's patch to put the top-level frame on the stack rather
1270        than the heap when not using the stack for recursion. This gives a
1271        performance improvement in many cases when recursion is not deep.
1272    
1273    16. Experimental code added to "pcretest -C" to output the stack frame size.
1274    
1275    
1276    Version 8.21 12-Dec-2011
1277    ------------------------
1278    
1279    1.  Updating the JIT compiler.
1280    
1281    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1282        are added as well.
1283    
1284    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1285        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1286        calling _pcre_jit_exec. Some extra comments are added.
1287    
1288    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1289        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1290        was introduced by change 18 for 8.20.
1291    
1292    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1293        ECMA-262 standard.
1294    
1295    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1296        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1297        This bug was probably introduced by change 9 of 8.13.
1298    
1299    7.  While fixing 6 above, I noticed that a number of other items were being
1300        incorrectly rejected as "not fixed length". This arose partly because newer
1301        opcodes had not been added to the fixed-length checking code. I have (a)
1302        corrected the bug and added tests for these items, and (b) arranged for an
1303        error to occur if an unknown opcode is encountered while checking for fixed
1304        length instead of just assuming "not fixed length". The items that were
1305        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1306        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1307        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1308    
1309    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1310        being incorrectly compiled and would have given unpredicatble results.
1311    
1312    9.  A possessively repeated subpattern with minimum repeat count greater than
1313        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1314        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1315        the first (A) could occur when it should not.
1316    
1317    10. Add a cast and remove a redundant test from the code.
1318    
1319    11. JIT should use pcre_malloc/pcre_free for allocation.
1320    
1321    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1322        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1323        is anything other than /usr, -L is still shown).
1324    
1325    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1326    
1327    14. Perl does not support \N without a following name in a [] class; PCRE now
1328        also gives an error.
1329    
1330    15. If a forward reference was repeated with an upper limit of around 2000,
1331        it caused the error "internal error: overran compiling workspace". The
1332        maximum number of forward references (including repeats) was limited by the
1333        internal workspace, and dependent on the LINK_SIZE. The code has been
1334        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1335        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1336        of around 200,000 forward references. While doing this, I also speeded up
1337        the filling in of repeated forward references.
1338    
1339    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1340        incorrectly expecting the subject to contain another "a" after the start.
1341    
1342    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1343        in the match, the SKIP should be ignored. This was not happening; instead
1344        the SKIP was being treated as NOMATCH. For patterns such as
1345        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1346        tested.
1347    
1348    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1349        now much more compatible with Perl, in particular in cases where the result
1350        is a non-match for a non-anchored pattern. For example, if
1351        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1352        "m", where previously it did not return a name. A side effect of this
1353        change is that for partial matches, the last encountered mark name is
1354        returned, as for non matches. A number of tests that were previously not
1355        Perl-compatible have been moved into the Perl-compatible test files. The
1356        refactoring has had the pleasing side effect of removing one argument from
1357        the match() function, thus reducing its stack requirements.
1358    
1359    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1360        subsequent uses of /S (without +) incorrectly behaved like /S+.
1361    
1362    21. Retrieve executable code size support for the JIT compiler and fixing
1363        some warnings.
1364    
1365    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1366        not work when the shorter character appeared right at the end of the
1367        subject string.
1368    
1369    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1370        systems.
1371    
1372    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1373        output it when the /M option is used in pcretest.
1374    
1375    25. The CheckMan script was not being included in the distribution. Also, added
1376        an explicit "perl" to run Perl scripts from the PrepareRelease script
1377        because this is reportedly needed in Windows.
1378    
1379    26. If study data was being save in a file and studying had not found a set of
1380        "starts with" bytes for the pattern, the data written to the file (though
1381        never used) was taken from uninitialized memory and so caused valgrind to
1382        complain.
1383    
1384    27. Updated RunTest.bat as provided by Sheri Pierce.
1385    
1386    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1387    
1388    29. Computation of memory usage for the table of capturing group names was
1389        giving an unnecessarily large value.
1390    
1391    
1392    Version 8.20 21-Oct-2011
1393    ------------------------
1394    
1395    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1396        a POSIX class. After further experiments with Perl, which convinced me that
1397        Perl has bugs and confusions, a closing square bracket is no longer allowed
1398        in a POSIX name. This bug also affected patterns with classes that started
1399        with full stops.
1400    
1401    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1402        captured substring, but while checking the failing first alternative,
1403        substring 1 is temporarily captured. If the output vector supplied to
1404        pcre_exec() was not big enough for this capture, the yield of the function
1405        was still zero ("insufficient space for captured substrings"). This cannot
1406        be totally fixed without adding another stack variable, which seems a lot
1407        of expense for a edge case. However, I have improved the situation in cases
1408        such as /(a)(b)x|abc/ matched against "abc", where the return code
1409        indicates that fewer than the maximum number of slots in the ovector have
1410        been set.
1411    
1412    3.  Related to (2) above: when there are more back references in a pattern than
1413        slots in the output vector, pcre_exec() uses temporary memory during
1414        matching, and copies in the captures as far as possible afterwards. It was
1415        using the entire output vector, but this conflicts with the specification
1416        that only 2/3 is used for passing back captured substrings. Now it uses
1417        only the first 2/3, for compatibility. This is, of course, another edge
1418        case.
1419    
1420    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1421        main code base, and can be used by building with --enable-jit. When this is
1422        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1423        runtime --no-jit option is given.
1424    
1425    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1426        ovector, the return from the function was zero, implying that there were
1427        other matches that did not fit. The correct "exactly full" value is now
1428        returned.
1429    
1430    6.  If a subpattern that was called recursively or as a subroutine contained
1431        (*PRUNE) or any other control that caused it to give a non-standard return,
1432        invalid errors such as "Error -26 (nested recursion at the same subject
1433        position)" or even infinite loops could occur.
1434    
1435    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1436        computing the minimum length on reaching *ACCEPT, and so ended up with the
1437        wrong value of 1 rather than 0. Further investigation indicates that
1438        computing a minimum subject length in the presence of *ACCEPT is difficult
1439        (think back references, subroutine calls), and so I have changed the code
1440        so that no minimum is registered for a pattern that contains *ACCEPT.
1441    
1442    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1443        it was not handled as intended. [But see 16 below.]
1444    
1445    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1446        Sheri Pierce.
1447    
1448    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1449        the first byte in a match must be "a".
1450    
1451    11. Change 17 for 8.13 increased the recursion depth for patterns like
1452        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1453        pattern contains any instances of (*THEN). If it does not, the old
1454        optimizations are restored. It would be nice to do this on a per-group
1455        basis, but at the moment that is not feasible.
1456    
1457    12. In some environments, the output of pcretest -C is CRLF terminated. This
1458        broke RunTest's code that checks for the link size. A single white space
1459        character after the value is now allowed for.
1460    
1461    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1462        For "fr", it uses the Windows-specific input and output files.
1463    
1464    14. If (*THEN) appeared in a group that was called recursively or as a
1465        subroutine, it did not work as intended. [But see next item.]
1466    
1467    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1468        pattern fragments (but not containing any | characters). If A and B are
1469        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1470        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1471        D. In other words, Perl considers parentheses that do not contain any |
1472        characters to be part of a surrounding alternative, whereas PCRE was
1473        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1474        differently. PCRE now behaves in the same way as Perl, except in the case
1475        of subroutine/recursion calls such as (?1) which have in any case always
1476        been different (but PCRE had them first :-).
1477    
1478    16. Related to 15 above: Perl does not treat the | in a conditional group as
1479        creating alternatives. Such a group is treated in the same way as an
1480        ordinary group without any | characters when processing (*THEN). PCRE has
1481        been changed to match Perl's behaviour.
1482    
1483    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1484        RunGrepTest script failed.
1485    
1486    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1487        inevitable for groups that contain captures, but it can lead to a lot of
1488        stack use in large patterns. The old behaviour has been restored for atomic
1489        groups that do not contain any capturing parentheses.
1490    
1491    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1492        suppress the check for a minimum subject length at run time. (If it was
1493        given to pcre_exec() or pcre_dfa_exec() it did work.)
1494    
1495    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1496        fail to work when decoding hex characters in data strings in EBCDIC
1497        environments.
1498    
1499    21. It appears that in at least one Mac OS environment, the isxdigit() function
1500        is implemented as a macro that evaluates to its argument more than once,
1501        contravening the C 90 Standard (I haven't checked a later standard). There
1502        was an instance in pcretest which caused it to go wrong when processing
1503        \x{...} escapes in subject strings. The has been rewritten to avoid using
1504        things like p++ in the argument of isxdigit().
1505    
1506    
1507    Version 8.13 16-Aug-2011
1508    ------------------------
1509    
1510    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1511    
1512    2.  Two minor typos in pcre_internal.h have been fixed.
1513    
1514    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1515        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1516        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1517    
1518    4.  There were a number of related bugs in the code for matching backrefences
1519        caselessly in UTF-8 mode when codes for the characters concerned were
1520        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1521        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1522        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1523        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1524        2-byte code at the end of the subject (it thought there wasn't enough data
1525        left).
1526    
1527    5.  Comprehensive information about what went wrong is now returned by
1528        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1529        as the output vector has at least 2 elements. The offset of the start of
1530        the failing character and a reason code are placed in the vector.
1531    
1532    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1533        now returned is for the first byte of the failing character, instead of the
1534        last byte inspected. This is an incompatible change, but I hope it is small
1535        enough not to be a problem. It makes the returned offset consistent with
1536        pcre_exec() and pcre_dfa_exec().
1537    
1538    7.  pcretest now gives a text phrase as well as the error number when
1539        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1540        failure, the offset and reason code are output.
1541    
1542    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1543        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1544        back over a single character (\n). This seems wrong (because it treated the
1545        two characters as a single entity when going forwards), conflicts with the
1546        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1547        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1548        has been changed.
1549    
1550    9.  Some internal refactoring has changed the processing so that the handling
1551        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1552        time (the PCRE_DOTALL option was changed this way some time ago: version
1553        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1554        which was always a bit of a fudge. It also means that there is one less
1555        argument for the match() function, which reduces its stack requirements
1556        slightly. This change also fixes an incompatibility with Perl: the pattern
1557        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1558    
1559    10. More internal refactoring has drastically reduced the number of recursive
1560        calls to match() for possessively repeated groups such as (abc)++ when
1561        using pcre_exec().
1562    
1563    11. While implementing 10, a number of bugs in the handling of groups were
1564        discovered and fixed:
1565    
1566        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1567        (a|)*(?1) gave a compile-time internal error.
1568        ((a|)+)+  did not notice that the outer group could match an empty string.
1569        (^a|^)+   was not marked as anchored.
1570        (.*a|.*)+ was not marked as matching at start or after a newline.
1571    
1572    12. Yet more internal refactoring has removed another argument from the match()
1573        function. Special calls to this function are now indicated by setting a
1574        value in a variable in the "match data" data block.
1575    
1576    13. Be more explicit in pcre_study() instead of relying on "default" for
1577        opcodes that mean there is no starting character; this means that when new
1578        ones are added and accidentally left out of pcre_study(), testing should
1579        pick them up.
1580    
1581    14. The -s option of pcretest has been documented for ages as being an old
1582        synonym of -m (show memory usage). I have changed it to mean "force study
1583        for every regex", that is, assume /S for every regex. This is similar to -i
1584        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1585        using it. It makes it easier to run collections of tests with and without
1586        study enabled, and thereby test pcre_study() more easily. All the standard
1587        tests are now run with and without -s (but some patterns can be marked as
1588        "never study" - see 20 below).
1589    
1590    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1591        restoration of the capturing data to the outer values was not happening
1592        correctly.
1593    
1594    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1595        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1596        pattern had matched an empty string, and so incorrectly returned a no
1597        match.
1598    
1599    17. There was optimizing code for the last branch of non-capturing parentheses,
1600        and also for the obeyed branch of a conditional subexpression, which used
1601        tail recursion to cut down on stack usage. Unfortunately, now that there is
1602        the possibility of (*THEN) occurring in these branches, tail recursion is
1603        no longer possible because the return has to be checked for (*THEN). These
1604        two optimizations have therefore been removed. [But see 8.20/11 above.]
1605    
1606    18. If a pattern containing \R was studied, it was assumed that \R always
1607        matched two bytes, thus causing the minimum subject length to be
1608        incorrectly computed because \R can also match just one byte.
1609    
1610    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1611        was incorrectly computed.
1612    
1613    20. If /S is present twice on a test pattern in pcretest input, it now
1614        *disables* studying, thereby overriding the use of -s on the command line
1615        (see 14 above). This is necessary for one or two tests to keep the output
1616        identical in both cases.
1617    
1618    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1619        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1620    
1621    22. When an atomic group that contained a capturing parenthesis was
1622        successfully matched, but the branch in which it appeared failed, the
1623        capturing was not being forgotten if a higher numbered group was later
1624        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1625        group 1 to "a", when in fact it should be unset. This applied to multi-
1626        branched capturing and non-capturing groups, repeated or not, and also to
1627        positive assertions (capturing in negative assertions does not happen
1628        in PCRE) and also to nested atomic groups.
1629    
1630    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1631        subject after a captured substring, to make it easier to tell which of a
1632        number of identical substrings has been captured.
1633    
1634    24. The way atomic groups are processed by pcre_exec() has been changed so that
1635        if they are repeated, backtracking one repetition now resets captured
1636        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1637        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1638        "aaa". Previously, it would have been "a". As part of this code
1639        refactoring, the way recursive calls are handled has also been changed.
1640    
1641    25. If an assertion condition captured any substrings, they were not passed
1642        back unless some other capturing happened later. For example, if
1643        (?(?=(a))a) was matched against "a", no capturing was returned.
1644    
1645    26. When studying a pattern that contained subroutine calls or assertions,
1646        the code for finding the minimum length of a possible match was handling
1647        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1648        group 1 called group 2 while simultaneously a separate group 2 called group
1649        1). A stack overflow occurred in this case. I have fixed this by limiting
1650        the recursion depth to 10.
1651    
1652    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1653        Fortmann. This supports explicit test numbers on the command line, and has
1654        argument validation and error reporting.
1655    
1656    28. An instance of \X with an unlimited repeat could fail if at any point the
1657        first character it looked at was a mark character.
1658    
1659    29. Some minor code refactoring concerning Unicode properties and scripts
1660        should reduce the stack requirement of match() slightly.
1661    
1662    30. Added the '=' option to pcretest to check the setting of unused capturing
1663        slots at the end of the pattern, which are documented as being -1, but are
1664        not included in the return count.
1665    
1666    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1667        compiled something random. Now it gives a compile-time error (as does
1668        Perl).
1669    
1670    32. A *MARK encountered during the processing of a positive assertion is now
1671        recorded and passed back (compatible with Perl).
1672    
1673    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1674        had alternative anchored branches, the search for a second match in a line
1675        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1676        matched the line "0102" twice. The same bug affected patterns that started
1677        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1678        twice.
1679    
1680    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1681        does, and because of capturing effects, quantifying parenthesized
1682        assertions may at times be useful. Quantifiers are now allowed for
1683        parenthesized assertions.
1684    
1685    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1686    
1687    36. \g was being checked for fancy things in a character class, when it should
1688        just be a literal "g".
1689    
1690    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1691        appearance of a nested POSIX class supersedes an apparent external class.
1692        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1693        unescaped square brackets may also appear as part of class names. For
1694        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1695        more like Perl. (But see 8.20/1 above.)
1696    
1697    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1698        was because it thought it was \N{name}, which is not supported).
1699    
1700    39. Add minix to OS list not supporting the -S option in pcretest.
1701    
1702    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1703        cannot analyze patterns in sufficient detail to catch mutual recursions
1704        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1705        subgroup is called recursively as a subpattern for a second time at the
1706        same position in the subject string. In previous releases this might have
1707        been caught by the recursion limit, or it might have run out of stack.
1708    
1709    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1710        happen only once. PCRE was, however incorrectly giving a compile time error
1711        "recursive call could loop indefinitely" because it cannot analyze the
1712        pattern in sufficient detail. The compile time test no longer happens when
1713        PCRE is compiling a conditional subpattern, but actual runaway loops are
1714        now caught at runtime (see 40 above).
1715    
1716    42. It seems that Perl allows any characters other than a closing parenthesis
1717        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1718        has been changed to be the same.
1719    
1720    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1721        as not to get warnings when autogen.sh is called. Also changed
1722        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1723    
1724    44. To help people who use pcregrep to scan files containing exceedingly long
1725        lines, the following changes have been made:
1726    
1727        (a) The default value of the buffer size parameter has been increased from
1728            8K to 20K. (The actual buffer used is three times this size.)
1729    
1730        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1731            PCRE is built.
1732    
1733        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1734            to be set at run time.
1735    
1736        (d) Numerical values in pcregrep options can be followed by K or M, for
1737            example --buffer-size=50K.
1738    
1739        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1740            given and the return code is set to 2.
1741    
1742    45. Add a pointer to the latest mark to the callout data block.
1743    
1744    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1745        partial match of an empty string instead of no match. This was specific to
1746        the use of ".".
1747    
1748    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1749        complete match instead of a partial match. This bug was dependent on both
1750        the PCRE_UTF8 and PCRE_DOTALL options being set.
1751    
1752    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1753        starting byte set, because \b was not being ignored.
1754    
1755    
1756    Version 8.12 15-Jan-2011
1757    ------------------------
1758    
1759    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1760        checks for such things as part of the documentation building process.
1761    
1762    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1763        --match-limit and --recursion-limit options (added for 8.11). In
1764        particular, this made one of the standard tests fail. (The integer value
1765        went into the wrong half of a long int.)
1766    
1767    3.  If the --colour option was given to pcregrep with -v (invert match), it
1768        did strange things, either producing crazy output, or crashing. It should,
1769        of course, ignore a request for colour when reporting lines that do not
1770        match.
1771    
1772    4.  Another pcregrep bug caused similar problems if --colour was specified with
1773        -M (multiline) and the pattern match finished with a line ending.
1774    
1775    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1776        matched in multiline mode, the following line was shown as part of the
1777        match. This seems wrong, so I have changed it.
1778    
1779    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1780        the check for further matches in the same line (so they could be coloured)
1781        to overrun the end of the current line. If another match was found, it was
1782        incorrectly shown (and then shown again when found in the next line).
1783    
1784    7.  If pcregrep was compiled under Windows, there was a reference to the
1785        function pcregrep_exit() before it was defined. I am assuming this was
1786        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1787        reported by a user. I've moved the definition above the reference.
1788    
1789    
1790    Version 8.11 10-Dec-2010
1791    ------------------------
1792    
1793    1.  (*THEN) was not working properly if there were untried alternatives prior
1794        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1795        backtracked to try for "b" instead of moving to the next alternative branch
1796        at the same level (in this case, to look for "c"). The Perl documentation
1797        is clear that when (*THEN) is backtracked onto, it goes to the "next
1798        alternative in the innermost enclosing group".
1799    
1800    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1801        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1802        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1803        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1804        (*THEN).
1805    
1806    3.  If \s appeared in a character class, it removed the VT character from
1807        the class, even if it had been included by some previous item, for example
1808        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1809        of \s, but is part of the POSIX "space" class.)
1810    
1811    4.  A partial match never returns an empty string (because you can always
1812        match an empty string at the end of the subject); however the checking for
1813        an empty string was starting at the "start of match" point. This has been
1814        changed to the "earliest inspected character" point, because the returned
1815        data for a partial match starts at this character. This means that, for
1816        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1817        (previously it gave "no match").
1818    
1819    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1820        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1821        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1822        has an implication that the given string is incomplete (because a partial
1823        match is preferred over a full match). For this reason, these items now
1824        give a partial match in this situation. [Aside: previously, the one case
1825        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1826        match rather than a full match, which was wrong by the old rules, but is
1827        now correct.]
1828    
1829    6.  There was a bug in the handling of #-introduced comments, recognized when
1830        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1831        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1832        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1833        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1834        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1835        places in pcre_compile().
1836    
1837    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1838        comments when looking ahead for named forward references to subpatterns,
1839        the only newline sequence it recognized was NL. It now handles newlines
1840        according to the set newline convention.
1841    
1842    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1843        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1844        cater for a lack of strerror(). These oversights have been fixed.
1845    
1846    9.  Added --match-limit and --recursion-limit to pcregrep.
1847    
1848    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1849    
1850    11. When the -o option was used, pcregrep was setting a return code of 1, even
1851        when matches were found, and --line-buffered was not being honoured.
1852    
1853    12. Added an optional parentheses number to the -o and --only-matching options
1854        of pcregrep.
1855    
1856    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1857        can match an empty string. The code to do it in pcretest and pcredemo
1858        needed fixing:
1859    
1860        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1861            only one byte after an empty string match just before CRLF (this case
1862            just got forgotten; "any" and "anycrlf" were OK).
1863    
1864        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1865            mode when an empty string match preceded an ASCII character followed by
1866            a non-ASCII character. (The code for advancing by one character rather
1867            than one byte was nonsense.)
1868    
1869        (c) The pcredemo.c sample program did not have any code at all to handle
1870            the cases when CRLF is a valid newline sequence.
1871    
1872    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1873        as a starting offset was within the subject string. There is now a new
1874        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1875        negative or greater than the length of the string. In order to test this,
1876        pcretest is extended to allow the setting of negative starting offsets.
1877    
1878    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1879        starting offset points to the beginning of a UTF-8 character was
1880        unnecessarily clumsy. I tidied it up.
1881    
1882    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1883        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1884    
1885    17. Nobody had reported that the --include_dir option, which was added in
1886        release 7.7 should have been called --include-dir (hyphen, not underscore)
1887        for compatibility with GNU grep. I have changed it to --include-dir, but
1888        left --include_dir as an undocumented synonym, and the same for
1889        --exclude-dir, though that is not available in GNU grep, at least as of
1890        release 2.5.4.
1891    
1892    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1893        characters from a string of bytes) have been redefined so as not to use
1894        loops, in order to improve performance in some environments. At the same
1895        time, I abstracted some of the common code into auxiliary macros to save
1896        repetition (this should not affect the compiled code).
1897    
1898    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1899        compile-time error is now given if \c is not followed by an ASCII
1900        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1901        different, and any byte value is allowed.)
1902    
1903    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1904        START_OPTIMIZE option, which is now allowed at compile time - but just
1905        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1906        to pcregrep and other applications that have no direct access to PCRE
1907        options. The new /Y option in pcretest sets this option when calling
1908        pcre_compile().
1909    
1910    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1911        back references. Groups containing recursive back references were forced to
1912        be atomic by that change, but in the case of named groups, the amount of
1913        memory required was incorrectly computed, leading to "Failed: internal
1914        error: code overflow". This has been fixed.
1915    
1916    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1917        pcretest.c, to avoid build problems in some Borland environments.
1918    
1919    
1920    Version 8.10 25-Jun-2010
1921    ------------------------
1922    
1923    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1924        THEN.
1925    
1926    2.  (*ACCEPT) was not working when inside an atomic group.
1927    
1928    3.  Inside a character class, \B is treated as a literal by default, but
1929        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1930        causes the error). The code is unchanged, but I tidied the documentation.
1931    
1932    4.  Inside a character class, PCRE always treated \R and \X as literals,
1933        whereas Perl faults them if its -w option is set. I have changed PCRE so
1934        that it faults them when PCRE_EXTRA is set.
1935    
1936    5.  Added support for \N, which always matches any character other than
1937        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1938    
1939    6.  When compiling pcregrep with newer versions of gcc which may have
1940        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1941        declared with attribute warn_unused_result" were given. Just casting the
1942        result to (void) does not stop the warnings; a more elaborate fudge is
1943        needed. I've used a macro to implement this.
1944    
1945    7.  Minor change to pcretest.c to avoid a compiler warning.
1946    
1947    8.  Added four artifical Unicode properties to help with an option to make
1948        \s etc use properties (see next item). The new properties are: Xan
1949        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1950    
1951    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1952        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1953        this option. Modified pcretest to add /W to test this facility. Added
1954        REG_UCP to make it available via the POSIX interface.
1955    
1956    10. Added --line-buffered to pcregrep.
1957    
1958    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1959        studied, and the match started with a letter with a code point greater than
1960        127 whose first byte was different to the first byte of the other case of
1961        the letter, the other case of this starting letter was not recognized
1962        (#976).
1963    
1964    12. If a pattern that was studied started with a repeated Unicode property
1965        test, for example, \p{Nd}+, there was the theoretical possibility of
1966        setting up an incorrect bitmap of starting bytes, but fortunately it could
1967        not have actually happened in practice until change 8 above was made (it
1968        added property types that matched character-matching opcodes).
1969    
1970    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1971        possible starting bytes for non-anchored patterns.
1972    
1973    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1974        \R, and also a number of cases that involve Unicode properties, both
1975        explicit and implicit when PCRE_UCP is set.
1976    
1977    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1978        input, it could crash or give wrong results if characters with values
1979        greater than 0xc0 were present in the subject string. (Detail: it assumed
1980        UTF-8 input when processing these items.)
1981    
1982    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1983        size_t is 64-bit (#991).
1984    
1985    17. Added a check for running out of memory when PCRE is compiled with
1986        --disable-stack-for-recursion (#990).
1987    
1988    18. If the last data line in a file for pcretest does not have a newline on
1989        the end, a newline was missing in the output.
1990    
1991    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1992        less than 128) in its various bitmaps. However, there is a facility for
1993        generating tables according to the current locale when PCRE is compiled. It
1994        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1995        characters, are recognized by isspace() and therefore were getting set in
1996        these tables, and indeed these tables seem to approximate to ISO 8859. This
1997        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1998        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1999        which of course cannot start UTF-8 characters. I have changed the code so
2000        that only real ASCII characters (less than 128) and the correct starting
2001        bytes for UTF-8 encodings are set for characters greater than 127 when in
2002        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
2003        altogether.)
2004    
2005    20. Added the /T option to pcretest so as to be able to run tests with non-
2006        standard character tables, thus making it possible to include the tests
2007        used for 19 above in the standard set of tests.
2008    
2009    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
2010        reference to a subpattern the other side of a comment that contains an
2011        opening parenthesis caused either an internal compiling error, or a
2012        reference to the wrong subpattern.
2013    
2014    
2015    Version 8.02 19-Mar-2010
2016    ------------------------
2017    
2018    1.  The Unicode data tables have been updated to Unicode 5.2.0.
2019    
2020    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
2021        configured.
2022    
2023    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
2024        original author of that file, following a query about its status.
2025    
2026    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
2027        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
2028    
2029    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
2030        quantifier applied to a forward-referencing subroutine call, could compile
2031        incorrect code or give the error "internal error: previously-checked
2032        referenced subpattern not found".
2033    
2034    6.  Both MS Visual Studio and Symbian OS have problems with initializing
2035        variables to point to external functions. For these systems, therefore,
2036        pcre_malloc etc. are now initialized to local functions that call the
2037        relevant global functions.
2038    
2039    7.  There were two entries missing in the vectors called coptable and poptable
2040        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
2041        I've fixed the data, and added a kludgy way of testing at compile time that
2042        the lengths are correct (equal to the number of opcodes).
2043    
2044    8.  Following on from 7, I added a similar kludge to check the length of the
2045        eint vector in pcreposix.c.
2046    
2047    9.  Error texts for pcre_compile() are held as one long string to avoid too
2048        much relocation at load time. To find a text, the string is searched,
2049        counting zeros. There was no check for running off the end of the string,
2050        which could happen if a new error number was added without updating the
2051        string.
2052    
2053    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
2054    
2055    11. \K was not working if it appeared in an atomic group or in a group that
2056        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
2057        \K is "not well defined" if used in an assertion. PCRE now accepts it if
2058        the assertion is positive, but not if it is negative.
2059    
2060    12. Change 11 fortuitously reduced the size of the stack frame used in the
2061        "match()" function of pcre_exec.c by one pointer. Forthcoming
2062        implementation of support for (*MARK) will need an extra pointer on the
2063        stack; I have reserved it now, so that the stack frame size does not
2064        decrease.
2065    
2066    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
2067        item in branch that calls a recursion is a subroutine call - as in the
2068        second branch in the above example - was incorrectly given the compile-
2069        time error "recursive call could loop indefinitely" because pcre_compile()
2070        was not correctly checking the subroutine for matching a non-empty string.
2071    
2072    14. The checks for overrunning compiling workspace could trigger after an
2073        overrun had occurred. This is a "should never occur" error, but it can be
2074        triggered by pathological patterns such as hundreds of nested parentheses.
2075        The checks now trigger 100 bytes before the end of the workspace.
2076    
2077    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
2078    
2079    
2080    Version 8.01 19-Jan-2010
2081    ------------------------
2082    
2083    1.  If a pattern contained a conditional subpattern with only one branch (in
2084        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
2085        computed the wrong minimum data length (which is of course zero for such
2086        subpatterns). This could cause incorrect "no match" results.
2087    
2088    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
2089        the pattern is reset in the first branch, pcre_compile() failed with
2090        "internal error: code overflow at offset...". This happened only when
2091        the reset was to the original external option setting. (An optimization
2092        abstracts leading options settings into an external setting, which was the
2093        cause of this.)
2094    
2095    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
2096        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
2097        assertion pattern did not match (meaning that the assertion was true), it
2098        was incorrectly treated as false if the SKIP had been reached during the
2099        matching. This also applied to assertions used as conditions.
2100    
2101    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
2102        assertion subpattern, including such a pattern used as a condition,
2103        unpredictable results occurred, instead of the error return
2104        PCRE_ERROR_DFA_UITEM.
2105    
2106    5.  The C++ GlobalReplace function was not working like Perl for the special
2107        situation when an empty string is matched. It now does the fancy magic
2108        stuff that is necessary.
2109    
2110    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
2111        removed. (These were left over from very, very early versions of PCRE.)
2112    
2113    7.  Some cosmetic changes to the code to make life easier when compiling it
2114        as part of something else:
2115    
2116        (a) Change DEBUG to PCRE_DEBUG.
2117    
2118        (b) In pcre_compile(), rename the member of the "branch_chain" structure
2119            called "current" as "current_branch", to prevent a collision with the
2120            Linux macro when compiled as a kernel module.
2121    
2122        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
2123            prevent a collision with the Linux macro when compiled as a kernel
2124            module.
2125    
2126    8.  In pcre_compile() there are some checks for integer overflows that used to
2127        cast potentially large values to (double). This has been changed to that
2128        when building, a check for int64_t is made, and if it is found, it is used
2129        instead, thus avoiding the use of floating point arithmetic. (There is no
2130        other use of FP in PCRE.) If int64_t is not found, the fallback is to
2131        double.
2132    
2133    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
2134        2005 (difference between two addresses compared to an unsigned value).
2135    
2136    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
2137        custom one, because of the following reported problem in Windows:
2138    
2139          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
2140              under Win32.
2141          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
2142              therefore missing the function definition.
2143          - The compiler thus generates a "C" signature for the test function.
2144          - The linker fails to find the "C" function.
2145          - PCRE fails to configure if asked to do so against libbz2.
2146    
2147    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
2148        messages were output:
2149    
2150          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
2151          rerunning libtoolize, to keep the correct libtool macros in-tree.
2152          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
2153    
2154        I have done both of these things.
2155    
2156    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
2157        most of the time, it *can* run out if it is given a pattern that contains a
2158        runaway infinite recursion. I updated the discussion in the pcrestack man
2159        page.
2160    
2161    13. Now that we have gone to the x.xx style of version numbers, the minor
2162        version may start with zero. Using 08 or 09 is a bad idea because users
2163        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
2164        interpreted as invalid octal numbers. I've updated the previous comment in
2165        configure.ac, and also added a check that gives an error if 08 or 09 are
2166        used.
2167    
2168    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
2169        causing partial matching to fail when the end of the subject matched \W
2170        in a UTF-8 pattern where \W was quantified with a minimum of 3.
2171    
2172    15. There were some discrepancies between the declarations in pcre_internal.h
2173        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
2174        their definitions. The declarations used "const uschar *" and the
2175        definitions used USPTR. Even though USPTR is normally defined as "const
2176        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
2177        reported that: "This difference in casting confuses some C++ compilers, for
2178        example, SunCC recognizes above declarations as different functions and
2179        generates broken code for hbpcre." I have changed the declarations to use
2180        USPTR.
2181    
2182    16. GNU libtool is named differently on some systems. The autogen.sh script now
2183        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
2184        (FreeBSD).
2185    
2186    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
2187        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
2188        comment: "Figure out how to create a longlong from a string: strtoll and
2189        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
2190        instance, but it only takes 2 args instead of 3!"
2191    
2192    18. A subtle bug concerned with back references has been fixed by a change of
2193        specification, with a corresponding code fix. A pattern such as
2194        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
2195        refers, was giving matches when it shouldn't. For example, xa=xaaa would
2196        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
2197        same bug. Such groups have to be quantified to be useful, or contained
2198        inside another quantified group. (If there's no repetition, the reference
2199        can never match.) The problem arises because, having left the group and
2200        moved on to the rest of the pattern, a later failure that backtracks into
2201        the group uses the captured value from the final iteration of the group
2202        rather than the correct earlier one. I have fixed this in PCRE by forcing
2203        any group that contains a reference to itself to be an atomic group; that
2204        is, there cannot be any backtracking into it once it has completed. This is
2205        similar to recursive and subroutine calls.
2206    
2207    
2208    Version 8.00 19-Oct-09
2209    ----------------------
2210    
2211    1.  The table for translating pcre_compile() error codes into POSIX error codes
2212        was out-of-date, and there was no check on the pcre_compile() error code
2213        being within the table. This could lead to an OK return being given in
2214        error.
2215    
2216    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
2217        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
2218        in a Windows environment.
2219    
2220    3.  The pcregrep --count option prints the count for each file even when it is
2221        zero, as does GNU grep. However, pcregrep was also printing all files when
2222        --files-with-matches was added. Now, when both options are given, it prints
2223        counts only for those files that have at least one match. (GNU grep just
2224        prints the file name in this circumstance, but including the count seems
2225        more useful - otherwise, why use --count?) Also ensured that the
2226        combination -clh just lists non-zero counts, with no names.
2227    
2228    4.  The long form of the pcregrep -F option was incorrectly implemented as
2229        --fixed_strings instead of --fixed-strings. This is an incompatible change,
2230        but it seems right to fix it, and I didn't think it was worth preserving
2231        the old behaviour.
2232    
2233    5.  The command line items --regex=pattern and --regexp=pattern were not
2234        recognized by pcregrep, which required --regex pattern or --regexp pattern
2235        (with a space rather than an '='). The man page documented the '=' forms,
2236        which are compatible with GNU grep; these now work.
2237    
2238    6.  No libpcreposix.pc file was created for pkg-config; there was just
2239        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
2240    
2241    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
2242        when UCP support is not needed, by modifying the Python script that
2243        generates it from Unicode data files. This should not matter if the module
2244        is correctly used as a library, but I received one complaint about 50K of
2245        unwanted data. My guess is that the person linked everything into his
2246        program rather than using a library. Anyway, it does no harm.
2247    
2248    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
2249        was a minimum greater than 1 for a wide character in a possessive
2250        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
2251        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
2252        character. Chaos in the form of incorrect output or a compiling loop could
2253        result.
2254    
2255    9.  The restrictions on what a pattern can contain when partial matching is
2256        requested for pcre_exec() have been removed. All patterns can now be
2257        partially matched by this function. In addition, if there are at least two
2258        slots in the offset vector, the offset of the earliest inspected character
2259        for the match and the offset of the end of the subject are set in them when
2260        PCRE_ERROR_PARTIAL is returned.
2261    
2262    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
2263        synonymous with PCRE_PARTIAL, for backwards compatibility, and
2264        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
2265        and may be more useful for multi-segment matching.
2266    
2267    11. Partial matching with pcre_exec() is now more intuitive. A partial match
2268        used to be given if ever the end of the subject was reached; now it is
2269        given only if matching could not proceed because another character was
2270        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
2271        string "Z", which now yields "no match" instead of "partial match". In the
2272        case of pcre_dfa_exec(), "no match" is given if every matching path for the
2273        final character ended with (*FAIL).
2274    
2275    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
2276        if the pattern had a "must contain" character that was already found in the
2277        earlier partial match, unless partial matching was again requested. For
2278        example, with the pattern /dog.(body)?/, the "must contain" character is
2279        "g". If the first part-match was for the string "dog", restarting with
2280        "sbody" failed. This bug has been fixed.
2281    
2282    13. The string returned by pcre_dfa_exec() after a partial match has been
2283        changed so that it starts at the first inspected character rather than the
2284        first character of the match. This makes a difference only if the pattern
2285        starts with a lookbehind assertion or \b or \B (\K is not supported by
2286        pcre_dfa_exec()). It's an incompatible change, but it makes the two
2287        matching functions compatible, and I think it's the right thing to do.
2288    
2289    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
2290        so that the demonstration program is easily available in environments where
2291        PCRE has not been installed from source.
2292    
2293    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
2294        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
2295        library.
2296    
2297    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
2298        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
2299        is not the first non-POSIX option to be added. Clearly some people find
2300        these options useful.
2301    
2302    17. If a caller to the POSIX matching function regexec() passes a non-zero
2303        value for nmatch with a NULL value for pmatch, the value of
2304        nmatch is forced to zero.
2305    
2306    18. RunGrepTest did not have a test for the availability of the -u option of
2307        the diff command, as RunTest does. It now checks in the same way as
2308        RunTest, and also checks for the -b option.
2309    
2310    19. If an odd number of negated classes containing just a single character
2311        interposed, within parentheses, between a forward reference to a named
2312        subpattern and the definition of the subpattern, compilation crashed with
2313        an internal error, complaining that it could not find the referenced
2314        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
2315        [The bug was that it was starting one character too far in when skipping
2316        over the character class, thus treating the ] as data rather than
2317        terminating the class. This meant it could skip too much.]
2318    
2319    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
2320        /g option in pcretest when the pattern contains \K, which makes it possible
2321        to have an empty string match not at the start, even when the pattern is
2322        anchored. Updated pcretest and pcredemo to use this option.
2323    
2324    21. If the maximum number of capturing subpatterns in a recursion was greater
2325        than the maximum at the outer level, the higher number was returned, but
2326        with unset values at the outer level. The correct (outer level) value is
2327        now given.
2328    
2329    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
2330        PCRE did not set those parentheses (unlike Perl). I have now found a way to
2331        make it do so. The string so far is captured, making this feature
2332        compatible with Perl.
2333    
2334    23. The tests have been re-organized, adding tests 11 and 12, to make it
2335        possible to check the Perl 5.10 features against Perl 5.10.
2336    
2337    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
2338        pattern matches a fixed length string. PCRE did not allow this; now it
2339        does. Neither allows recursion.
2340    
2341    25. I finally figured out how to implement a request to provide the minimum
2342        length of subject string that was needed in order to match a given pattern.
2343        (It was back references and recursion that I had previously got hung up
2344        on.) This code has now been added to pcre_study(); it finds a lower bound
2345        to the length of subject needed. It is not necessarily the greatest lower
2346        bound, but using it to avoid searching strings that are too short does give
2347        some useful speed-ups. The value is available to calling programs via
2348        pcre_fullinfo().
2349    
2350    26. While implementing 25, I discovered to my embarrassment that pcretest had
2351        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
2352        study optimizations had never been tested with that matching function.
2353        Oops. What is worse, even when it was passed study data, there was a bug in
2354        pcre_dfa_exec() that meant it never actually used it. Double oops. There
2355        were also very few tests of studied patterns with pcre_dfa_exec().
2356    
2357    27. If (?| is used to create subpatterns with duplicate numbers, they are now
2358        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
2359        on the other side of the coin, they are no longer allowed to have different
2360        names, because these cannot be distinguished in PCRE, and this has caused
2361        confusion. (This is a difference from Perl.)
2362    
2363    28. When duplicate subpattern names are present (necessarily with different
2364        numbers, as required by 27 above), and a test is made by name in a
2365        conditional pattern, either for a subpattern having been matched, or for
2366        recursion in such a pattern, all the associated numbered subpatterns are
2367        tested, and the overall condition is true if the condition is true for any
2368        one of them. This is the way Perl works, and is also more like the way
2369        testing by number works.
2370    
2371    
2372    Version 7.9 11-Apr-09
2373    ---------------------
2374    
2375    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
2376        (pcretest), all targets were linked against these libraries. This included
2377        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
2378        libraries. This caused unwanted dependencies to be created. This problem
2379        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
2380        pcretest is linked with readline.
2381    
2382    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
2383        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
2384        moved outside it again, because FALSE and TRUE are already defined in AIX,
2385        but BOOL is not.
2386    
2387    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
2388        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
2389    
2390    4.  The pcregrep documentation said spaces were inserted as well as colons (or
2391        hyphens) following file names and line numbers when outputting matching
2392        lines. This is not true; no spaces are inserted. I have also clarified the
2393        wording for the --colour (or --color) option.
2394    
2395    5.  In pcregrep, when --colour was used with -o, the list of matching strings
2396        was not coloured; this is different to GNU grep, so I have changed it to be
2397        the same.
2398    
2399    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
2400        each matching line was coloured. Now it goes on to look for further matches
2401        of any of the test patterns, which is the same behaviour as GNU grep.
2402    
2403    7.  A pattern that could match an empty string could cause pcregrep to loop; it
2404        doesn't make sense to accept an empty string match in pcregrep, so I have
2405        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2406        seems to be how GNU grep behaves. [But see later change 40 for release
2407        8.33.]
2408    
2409    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2410        start or after a newline", because the conditional assertion was not being
2411        correctly handled. The rule now is that both the assertion and what follows
2412        in the first alternative must satisfy the test.
2413    
2414    9.  If auto-callout was enabled in a pattern with a conditional group whose
2415        condition was an assertion, PCRE could crash during matching, both with
2416        pcre_exec() and pcre_dfa_exec().
2417    
2418    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2419        used for matching.
2420    
2421    11. Unicode property support in character classes was not working for
2422        characters (bytes) greater than 127 when not in UTF-8 mode.
2423    
2424    12. Added the -M command line option to pcretest.
2425    
2426    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2427    
2428    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2429    
2430    16. Added comments and documentation about mis-use of no_arg in the C++
2431        wrapper.
2432    
2433    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2434        from Martin Jerabek that uses macro names for all relevant character and
2435        string constants.
2436    
2437    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2438        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2439        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2440        these, but not everybody uses configure.
2441    
2442    19. A conditional group that had only one branch was not being correctly
2443        recognized as an item that could match an empty string. This meant that an
2444        enclosing group might also not be so recognized, causing infinite looping
2445        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2446        with the subject "ab", where knowledge that the repeated group can match
2447        nothing is needed in order to break the loop.
2448    
2449    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2450        exec(), but without supplying a callout function, matching went wrong.
2451    
2452    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2453        leak if the size of the offset vector was greater than 30. When the vector
2454        is smaller, the saved offsets during recursion go onto a local stack
2455        vector, but for larger vectors malloc() is used. It was failing to free
2456        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2457        error, in fact).
2458    
2459    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2460        heapframe that is used only when UTF-8 support is enabled. This caused no
2461        problem, but was untidy.
2462    
2463    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2464        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2465        included within another project.
2466    
2467    24. Steven Van Ingelgem's patches to add more options to the CMake support,
2468        slightly modified by me:
2469    
2470          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2471              not building pcregrep.
2472    
2473          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2474              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2475    
2476    25. Forward references, both numeric and by name, in patterns that made use of
2477        duplicate group numbers, could behave incorrectly or give incorrect errors,
2478        because when scanning forward to find the reference group, PCRE was not
2479        taking into account the duplicate group numbers. A pattern such as
2480        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2481    
2482    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2483        the feature of a custom pointer more persuasive (as requested by a user).
2484    
2485    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2486        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2487        pre-defined.
2488    
2489    28. Added support for (*UTF8) at the start of a pattern.
2490    
2491    29. Arrange for flags added by the "release type" setting in CMake to be shown
2492        in the configuration summary.
2493    
2494    
2495    Version 7.8 05-Sep-08
2496    ---------------------
2497    
2498    1.  Replaced UCP searching code with optimized version as implemented for Ad
2499        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2500        stage table and inline lookup instead of a function, giving speed ups of 2
2501        to 5 times on some simple patterns that I tested. Permission was given to
2502        distribute the MultiStage2.py script that generates the tables (it's not in
2503        the tarball, but is in the Subversion repository).
2504    
2505    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2506        scripts.
2507    
2508    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2509        a group with a zero qualifier. The result of the study could be incorrect,
2510        or the function might crash, depending on the pattern.
2511    
2512    4.  Caseless matching was not working for non-ASCII characters in back
2513        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2514        It now works when Unicode Property Support is available.
2515    
2516    5.  In pcretest, an escape such as \x{de} in the data was always generating
2517        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2518        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2519        truncation.
2520    
2521    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2522    
2523    7.  Added two (int) casts to pcregrep when printing the difference of two
2524        pointers, in case they are 64-bit values.
2525    
2526    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2527        test 2 if it fails.
2528    
2529    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2530        and a #define of that name to empty if it is not externally set. This is to
2531        allow users of MSVC to set it if necessary.
2532    
2533    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2534        the convenience functions in the pcre_get.c source file.
2535    
2536    11. An option change at the start of a pattern that had top-level alternatives
2537        could cause overwriting and/or a crash. This command provoked a crash in
2538        some environments:
2539    
2540          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2541    
2542        This potential security problem was recorded as CVE-2008-2371.
2543    
2544    12. For a pattern where the match had to start at the beginning or immediately
2545        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2546        pcre_dfa_exec() could read past the end of the passed subject if there was
2547        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2548        pcretest so that it places the subject at the end of its malloc-ed buffer.
2549    
2550    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2551        exec() might read past the end of the data buffer in UTF-8 mode.
2552    
2553    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2554        the data contained the byte 0x85 as part of a UTF-8 character within its
2555        first line. This applied both to normal and DFA matching.
2556    
2557    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2558        /^[^d]*?$/8 failed to match "abc".
2559    
2560    16. Added a missing copyright notice to pcrecpp_internal.h.
2561    
2562    17. Make it more clear in the documentation that values returned from
2563        pcre_exec() in ovector are byte offsets, not character counts.
2564    
2565    18. Tidied a few places to stop certain compilers from issuing warnings.
2566    
2567    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2568        supplied by Stefan Weber. I made a further small update for 7.8 because
2569        there is a change of source arrangements: the pcre_searchfuncs.c module is
2570        replaced by pcre_ucd.c.
2571    
2572    
2573    Version 7.7 07-May-08
2574    ---------------------
2575    
2576    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2577        a string to a long long, pretend we don't even have a long long." This is
2578        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2579    
2580    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2581        pre-7.6 versions, which defined a global no_arg variable instead of putting
2582        it in the RE class. (See also #8 below.)
2583    
2584    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2585        Lopes.
2586    
2587    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2588    
2589        (1) The include/exclude patterns were being applied to the whole pathnames
2590            of files, instead of just to the final components.
2591    
2592        (2) If there was more than one level of directory, the subdirectories were
2593            skipped unless they satisfied the include/exclude conditions. This is
2594            inconsistent with GNU grep (and could even be seen as contrary to the
2595            pcregrep specification - which I improved to make it absolutely clear).
2596            The action now is always to scan all levels of directory, and just
2597            apply the include/exclude patterns to regular files.
2598    
2599    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2600        --exclude_dir in the tests to avoid scanning .svn directories.
2601    
2602    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2603        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2604        doesn't support NULs in patterns.
2605    
2606    7.  Added some missing "const"s to declarations of static tables in
2607        pcre_compile.c and pcre_dfa_exec.c.
2608    
2609    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2610        caused by fix #2  above. (Subsequently also a second patch to fix the
2611        first patch. And a third patch - this was a messy problem.)
2612    
2613    9.  Applied Craig's patch to remove the use of push_back().
2614    
2615    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2616        matching function regexec().
2617    
2618    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2619        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2620        references. PCRE supports relative numbers with this syntax (I don't think
2621        Oniguruma does).
2622    
2623    12. Previously, a group with a zero repeat such as (...){0} was completely
2624        omitted from the compiled regex. However, this means that if the group
2625        was called as a subroutine from elsewhere in the pattern, things went wrong
2626        (an internal error was given). Such groups are now left in the compiled
2627        pattern, with a new opcode that causes them to be skipped at execution
2628        time.
2629    
2630    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2631        to the way PCRE behaves:
2632    
2633        (a) A lone ] character is dis-allowed (Perl treats it as data).
2634    
2635        (b) A back reference to an unmatched subpattern matches an empty string
2636            (Perl fails the current match path).
2637    
2638        (c) A data ] in a character class must be notated as \] because if the
2639            first data character in a class is ], it defines an empty class. (In
2640            Perl it is not possible to have an empty class.) The empty class []
2641            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2642            The negative empty class [^] matches any one character, independently
2643            of the DOTALL setting.
2644    
2645    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2646        non-existent subpattern following a character class starting with ']' and
2647        containing () gave an internal compiling error instead of "reference to
2648        non-existent subpattern". Fortunately, when the pattern did exist, the
2649        compiled code was correct. (When scanning forwards to check for the
2650        existence of the subpattern, it was treating the data ']' as terminating
2651        the class, so got the count wrong. When actually compiling, the reference
2652        was subsequently set up correctly.)
2653    
2654    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2655        it was being rejected as not supported by pcre_dfa_exec(), even though
2656        other assertions are supported. I have made pcre_dfa_exec() support
2657        (*FAIL).
2658    
2659    16. The implementation of 13c above involved the invention of a new opcode,
2660        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2661        cannot be changed at match time, I realized I could make a small
2662        improvement to matching performance by compiling OP_ALLANY instead of
2663        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2664        on the OP_ANY path.
2665    
2666    17. Compiling pcretest on Windows with readline support failed without the
2667        following two fixes: (1) Make the unistd.h include conditional on
2668        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2669    
2670    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2671        ncurses library to be included for pcretest when ReadLine support is
2672        requested, but also to allow for it to be overridden. This patch came from
2673        Daniel Bergström.
2674    
2675    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2676        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2677        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2678        spotting this.
2679    
2680    
2681    Version 7.6 28-Jan-08
2682    ---------------------
2683    
2684    1.  A character class containing a very large number of characters with
2685        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2686        overflow.
2687    
2688    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2689        HAVE_LONG_LONG is not defined.
2690    
2691    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2692        bring them up to date and include new features. This patch includes:
2693    
2694        - Fixed PH's badly added libz and libbz2 support.
2695        - Fixed a problem with static linking.
2696        - Added pcredemo. [But later removed - see 7 below.]
2697        - Fixed dftables problem and added an option.
2698        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2699            HAVE_LONG_LONG.
2700        - Added readline support for pcretest.
2701        - Added an listing of the option settings after cmake has run.
2702    
2703    4.  A user submitted a patch to Makefile that makes it easy to create
2704        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2705        Makefile.am that cause it to include this special target, without
2706        affecting anything else. Note that the same mingw target plus all
2707        the other distribution libraries and programs are now supported
2708        when configuring with CMake (see 6 below) instead of with
2709        Configure/Make.
2710    
2711    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2712        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2713        exported in the Windows port". It has not yet been confirmed that the patch
2714        solves the problem, but it does no harm.
2715    
2716    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2717        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2718        with CMake, and also correct the comment about stack recursion.
2719    
2720    7.  Remove the automatic building of pcredemo from the ./configure system and
2721        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2722        of a program that users should build themselves after PCRE is installed, so
2723        building it automatically is not really right. What is more, it gave
2724        trouble in some build environments.
2725    
2726    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2727    
2728    
2729    Version 7.5 10-Jan-08
2730  ---------------------  ---------------------
2731    
2732  1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'  1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2733      values in parens when parsing an RE using the C++ wrapper."      values in parens when parsing an RE using the C++ wrapper."
2734    
2735  2.  Negative specials like \S did not work in character classes in UTF-8 mode.  2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2736      Characters greater than 255 were excluded from the class instead of being      Characters greater than 255 were excluded from the class instead of being
2737      included.      included.
2738    
2739  3.  The same bug as (2) above applied to negated POSIX classes such as  3.  The same bug as (2) above applied to negated POSIX classes such as
2740      [:^space:].      [:^space:].
2741    
2742  4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it  4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2743      defined or documented. It seems to have been a typo for PCRE_STATIC, so      defined or documented. It seems to have been a typo for PCRE_STATIC, so
2744      I have changed it.      I have changed it.
2745    
2746  5.  The construct (?&) was not diagnosed as a syntax error (it referenced the  5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2747      first named subpattern) and a construct such as (?&a) would reference the      first named subpattern) and a construct such as (?&a) would reference the
2748      first named subpattern whose name started with "a" (in other words, the      first named subpattern whose name started with "a" (in other words, the
2749      length check was missing).      length check was missing). Both these problems are fixed. "Subpattern name
2750        expected" is now given for (?&) (a zero-length name), and this patch also
2751        makes it give the same error for \k'' (previously it complained that that
2752        was a reference to a non-existent subpattern).
2753    
2754  6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;  6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2755      this is right because (?- can be followed by option settings as well as by      this is right because (?- can be followed by option settings as well as by
2756      digits. I have, however, made the messages clearer.      digits. I have, however, made the messages clearer.
2757    
2758  7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns  7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2759      than the number used in the conditional) now cause a compile-time error.      than the number used in the conditional) now cause a compile-time error.
2760      This is actually not compatible with Perl, which accepts such patterns, but      This is actually not compatible with Perl, which accepts such patterns, but
2761      treats the conditional as always being FALSE (as PCRE used to), but it      treats the conditional as always being FALSE (as PCRE used to), but it
2762      seems to me that giving a diagnostic is better.      seems to me that giving a diagnostic is better.
2763    
2764    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2765        and messages.
2766    
2767    9.  Fix two occurrences of "backslash" in comments that should have been
2768        "backspace".
2769    
2770    10. Remove two redundant lines of code that can never be obeyed (their function
2771        was moved elsewhere).
2772    
2773    11. The program that makes PCRE's Unicode character property table had a bug
2774        which caused it to generate incorrect table entries for sequences of
2775        characters that have the same character type, but are in different scripts.
2776        It amalgamated them into a single range, with the script of the first of
2777        them. In other words, some characters were in the wrong script. There were
2778        thirteen such cases, affecting characters in the following ranges:
2779    
2780          U+002b0 - U+002c1
2781          U+0060c - U+0060d
2782          U+0061e - U+00612
2783          U+0064b - U+0065e
2784          U+0074d - U+0076d
2785          U+01800 - U+01805
2786          U+01d00 - U+01d77
2787          U+01d9b - U+01dbf
2788          U+0200b - U+0200f
2789          U+030fc - U+030fe
2790          U+03260 - U+0327f
2791          U+0fb46 - U+0fbb1
2792          U+10450 - U+1049d
2793    
2794    12. The -o option (show only the matching part of a line) for pcregrep was not
2795        compatible with GNU grep in that, if there was more than one match in a
2796        line, it showed only the first of them. It now behaves in the same way as
2797        GNU grep.
2798    
2799    13. If the -o and -v options were combined for pcregrep, it printed a blank
2800        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2801        does the same. The return code can be used to tell if there were any
2802        non-matching lines.
2803    
2804    14. Added --file-offsets and --line-offsets to pcregrep.
2805    
2806    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2807        infinitely looping recursion. The bug was that positive lookaheads were not
2808        being skipped when checking for a possible empty match (negative lookaheads
2809        and both kinds of lookbehind were skipped).
2810    
2811    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2812        inclusion of <windows.h> to before rather than after the definition of
2813        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2814    
2815    17. Specifying a possessive quantifier with a specific limit for a Unicode
2816        character property caused pcre_compile() to compile bad code, which led at
2817        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2818        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2819        caused the error; without that there was no problem.
2820    
2821    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2822    
2823    19. Added --enable-pcretest-libreadline.
2824    
2825    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2826        RE::GlobalReplace(). As a result, the number of replacements returned was
2827        double what it should be. I removed one of the increments, but Craig sent a
2828        later patch that removed the other one (the right fix) and added unit tests
2829        that check the return values (which was not done before).
2830    
2831    21. Several CMake things:
2832    
2833        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2834            the names libpcre and libpcreposix, not just pcre and pcreposix.
2835    
2836        (2) The above change means that pcretest and pcregrep are now correctly
2837            linked with the newly-built libraries, not previously installed ones.
2838    
2839        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2840    
2841    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2842        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2843        UTF-8 newline character). The key issue is that the pattern starts .*;
2844        this means that the match must be either at the beginning, or after a
2845        newline. The bug was in the code for advancing after a failed match and
2846        checking that the new position followed a newline. It was not taking
2847        account of UTF-8 characters correctly.
2848    
2849    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2850        character classes. PCRE was not treating the sequence [:...:] as a
2851        character class unless the ... were all letters. Perl, however, seems to
2852        allow any characters between [: and :], though of course it rejects as
2853        unknown any "names" that contain non-letters, because all the known class
2854        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2855        for example, whereas PCRE did not - it did not recognize a POSIX character
2856        class. This seemed a bit dangerous, so the code has been changed to be
2857        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2858        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2859        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2860        Perl does, and where it didn't before.
2861    
2862    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2863        Windows environments %n is disabled by default.
2864    
2865    
2866  Version 7.4 21-Sep-07  Version 7.4 21-Sep-07
2867  ---------------------  ---------------------

Legend:
Removed from v.270  
changed lines
  Added in v.1593

  ViewVC Help
Powered by ViewVC 1.1.5