ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 167 by ph10, Wed May 9 15:53:54 2007 UTC revision 1609 by ph10, Mon Nov 23 12:39:23 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4  Version 7.2 01-May-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.
7    Version 8.38 23-November-2015
8    -----------------------------
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.
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.
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.
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.
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.
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.
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.)
43    8.  There was no check for integer overflow in subroutine calls such as (?123).
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.
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.
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|)+)+$()/.
55    12. In an EBCDIC environment, pcretest was mishandling the escape sequences
56        \a and \e in test subject lines.
58    13. In an EBCDIC environment, \a in a pattern was converted to the ASCII
59        instead of the EBCDIC value.
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.
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.
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.
72    17. \h within a class was incorrectly compiled in EBCDIC environments.
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))|(?=(.))))/.
78    19. JIT should return with error when the compiled pattern requires more stack
79        space than the maximum.
81    20. A possessively repeated conditional group that could match an empty string,
82        for example, /(?(R))*+/, was incorrectly compiled.
84    21. Fix infinite recursion in the JIT compiler when certain patterns such as
85        /(?:|a|){100}x/ are analysed.
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.
91    23. Pathological patterns containing many nested occurrences of [: caused
92        pcre_compile() to run for a very long time.
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.
98    25. If (?R was followed by - or + incorrect behaviour happened instead of a
99        diagnostic.
101    26. Arrange to give up on finding the minimum matching length for overly
102        complex patterns.
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.
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.
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.
117    30. Error messages for syntax errors following \g and \k were giving inaccurate
118        offsets in the pattern.
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.
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.
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.
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.
135    35. Match limit check added to JIT recursion. This issue was found by Karl
136        Skomski with a custom LLVM fuzzer.
138    36. Yet another case similar to 27 above has been circumvented by an
139        unconditional allocation of extra memory. This issue is fixed "properly" in
140        PCRE2 by refactoring the way references are handled. Wen Guanxing
141        from Venustech ADLAB discovered this bug.
143    37. Fix two assertion fails in JIT. These issues were found by Karl Skomski
144        with a custom LLVM fuzzer.
146    38. Fixed a corner case of range optimization in JIT.
148    39. An incorrect error "overran compiling workspace" was given if there were
149        exactly enough group forward references such that the last one extended
150        into the workspace safety margin. The next one would have expanded the
151        workspace. The test for overflow was not including the safety margin.
153    40. A match limit issue is fixed in JIT which was found by Karl Skomski
154        with a custom LLVM fuzzer.
156    41. Remove the use of /dev/null in testdata/testinput2, because it doesn't
157        work under Windows. (Why has it taken so long for anyone to notice?)
159    42. In a character class such as [\W\p{Any}] where both a negative-type escape
160        ("not a word character") and a property escape were present, the property
161        escape was being ignored.
163    43. Fix crash caused by very long (*MARK) or (*THEN) names.
165    44. A sequence such as [[:punct:]b] that is, a POSIX character class followed
166        by a single ASCII character in a class item, was incorrectly compiled in
167        UCP mode. The POSIX class got lost, but only if the single character
168        followed it.
170    45. [:punct:] in UCP mode was matching some characters in the range 128-255
171        that should not have been matched.
173    46. If [:^ascii:] or [:^xdigit:] or [:^cntrl:] are present in a non-negated
174        class, all characters with code points greater than 255 are in the class.
175        When a Unicode property was also in the class (if PCRE_UCP is set, escapes
176        such as \w are turned into Unicode properties), wide characters were not
177        correctly handled, and could fail to match.
180    Version 8.37 28-April-2015
181    --------------------------
183    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
184        for those parentheses to be closed with whatever has been captured so far.
185        However, it was failing to mark any other groups between the hightest
186        capture so far and the currrent group as "unset". Thus, the ovector for
187        those groups contained whatever was previously there. An example is the
188        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
190    2.  If an assertion condition was quantified with a minimum of zero (an odd
191        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
193    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
194        unrecognized modifier, a crash could occur.
196    4.  An attempt to do global matching in pcretest with a zero-length ovector
197        caused a crash.
199    5.  Fixed a memory leak during matching that could occur for a subpattern
200        subroutine call (recursive or otherwise) if the number of captured groups
201        that had to be saved was greater than ten.
203    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
204        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
205        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
207    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
208        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
210    8. When a pattern is compiled, it remembers the highest back reference so that
211       when matching, if the ovector is too small, extra memory can be obtained to
212       use instead. A conditional subpattern whose condition is a check on a
213       capture having happened, such as, for example in the pattern
214       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
215       setting the highest backreference number. This mattered only if pcre_exec()
216       was called with an ovector that was too small to hold the capture, and there
217       was no other kind of back reference (a situation which is probably quite
218       rare). The effect of the bug was that the condition was always treated as
219       FALSE when the capture could not be consulted, leading to a incorrect
220       behaviour by pcre_exec(). This bug has been fixed.
222    9. A reference to a duplicated named group (either a back reference or a test
223       for being set in a conditional) that occurred in a part of the pattern where
224       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
225       to be incorrectly calculated, leading to overwriting.
227    10. A mutually recursive set of back references such as (\2)(\1) caused a
228        segfault at study time (while trying to find the minimum matching length).
229        The infinite loop is now broken (with the minimum length unset, that is,
230        zero).
232    11. If an assertion that was used as a condition was quantified with a minimum
233        of zero, matching went wrong. In particular, if the whole group had
234        unlimited repetition and could match an empty string, a segfault was
235        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
236        assertions to be quantified, but not if they are being used as conditions,
237        so the above pattern is faulted by Perl. PCRE has now been changed so that
238        it also rejects such patterns.
240    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
241        failed to allow the zero-repeat case if pcre2_exec() was called with an
242        ovector too small to capture the group.
244    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
245        Red Hat Product Security:
247        (a) A crash if /K and /F were both set with the option to save the compiled
248        pattern.
250        (b) Another crash if the option to print captured substrings in a callout
251        was combined with setting a null ovector, for example \O\C+ as a subject
252        string.
254    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
255        forward reference repeated a large (but limited) number of times within a
256        repeated outer group that has a zero minimum quantifier, caused incorrect
257        code to be compiled, leading to the error "internal error:
258        previously-checked referenced subpattern not found" when an incorrect
259        memory address was read. This bug was reported as "heap overflow",
260        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
261        CVE-2015-2325.
263    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
264        call within a group that also contained a recursive back reference caused
265        incorrect code to be compiled. This bug was reported as "heap overflow",
266        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
267        number CVE-2015-2326.
269    24. Computing the size of the JIT read-only data in advance has been a source
270        of various issues, and new ones are still appear unfortunately. To fix
271        existing and future issues, size computation is eliminated from the code,
272        and replaced by on-demand memory allocation.
274    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
275        adjacent to the end of the range, and the range contained characters with
276        more than one other case, caused incorrect behaviour when compiled in UTF
277        mode. In that example, the range a-j was left out of the class.
279    26. Fix JIT compilation of conditional blocks, which assertion
280        is converted to (*FAIL). E.g: /(?(?!))/.
282    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
283        discovered by the LLVM fuzzer.
285    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
286        when this assertion was used as a condition, for example (?(?!)a|b). In
287        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
288        error about an unsupported item.
290    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
291        possessification code could take exponential time to complete. A recursion
292        depth limit of 1000 has been imposed to limit the resources used by this
293        optimization.
295    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
296        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
297        because \S ensures they are all in the class. The code for doing this was
298        interacting badly with the code for computing the amount of space needed to
299        compile the pattern, leading to a buffer overflow. This bug was discovered
300        by the LLVM fuzzer.
302    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
303        other kinds of group caused stack overflow at compile time. This bug was
304        discovered by the LLVM fuzzer.
306    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
307        between a subroutine call and its quantifier was incorrectly compiled,
308        leading to buffer overflow or other errors. This bug was discovered by the
309        LLVM fuzzer.
311    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
312        assertion after (?(. The code was failing to check the character after
313        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
314        was discovered by the LLVM fuzzer.
316    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
317        a fixed maximum following a group that contains a subroutine reference was
318        incorrectly compiled and could trigger buffer overflow. This bug was
319        discovered by the LLVM fuzzer.
321    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
322        caused a stack overflow instead of the diagnosis of a non-fixed length
323        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
325    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
326        (e.g. /(?<=\Ka)/) could make pcregrep loop.
328    37. There was a similar problem to 36 in pcretest for global matches.
330    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
331        and a subsequent item in the pattern caused a non-match, backtracking over
332        the repeated \X did not stop, but carried on past the start of the subject,
333        causing reference to random memory and/or a segfault. There were also some
334        other cases where backtracking after \C could crash. This set of bugs was
335        discovered by the LLVM fuzzer.
337    39. The function for finding the minimum length of a matching string could take
338        a very long time if mutual recursion was present many times in a pattern,
339        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
340        method has been implemented. This infelicity was discovered by the LLVM
341        fuzzer.
343    40. Static linking against the PCRE library using the pkg-config module was
344        failing on missing pthread symbols.
347    Version 8.36 26-September-2014
348    ------------------------------
350    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
351        -Wmissing-field-initializers and -Wunused-parameter.
353    2.  The tests for quantifiers being too big (greater than 65535) were being
354        applied after reading the number, and stupidly assuming that integer
355        overflow would give a negative number. The tests are now applied as the
356        numbers are read.
358    3.  Tidy code in pcre_exec.c where two branches that used to be different are
359        now the same.
361    4.  The JIT compiler did not generate match limit checks for certain
362        bracketed expressions with quantifiers. This may lead to exponential
363        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
364        issue should be resolved now.
366    5.  Fixed an issue, which occures when nested alternatives are optimized
367        with table jumps.
369    6.  Inserted two casts and changed some ints to size_t in the light of some
370        reported 64-bit compiler warnings (Bugzilla 1477).
372    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
373        an empty string, which sometimes were behaving incorrectly in the
374        interpreter (though correctly in the JIT matcher). This pcretest input is
375        an example:
377          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
380        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
381        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
382        for an empty string was breaking the inner loop and carrying on at a lower
383        level, when possessive repeated groups should always return to a higher
384        level as they have no backtrack points in them. The empty string test now
385        occurs at the outer level.
387    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
388        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
390    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
391        doesn't).
393    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
394        to hold a pattern to be incorrectly computed (too small) when there were
395        named back references to duplicated names. This could cause "internal
396        error: code overflow" or "double free or corruption" or other memory
397        handling errors.
399    11. When named subpatterns had the same prefixes, back references could be
400        confused. For example, in this pattern:
402          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
404        the reference to 'Name' was incorrectly treated as a reference to a
405        duplicate name.
407    12. A pattern such as /^s?c/mi8 where the optional character has more than
408        one "other case" was incorrectly compiled such that it would only try to
409        match starting at "c".
411    13. When a pattern starting with \s was studied, VT was not included in the
412        list of possible starting characters; this should have been part of the
413        8.34/18 patch.
415    14. If a character class started [\Qx]... where x is any character, the class
416        was incorrectly terminated at the ].
418    15. If a pattern that started with a caseless match for a character with more
419        than one "other case" was studied, PCRE did not set up the starting code
420        unit bit map for the list of possible characters. Now it does. This is an
421        optimization improvement, not a bug fix.
423    16. The Unicode data tables have been updated to Unicode 7.0.0.
425    17. Fixed a number of memory leaks in pcregrep.
427    18. Avoid a compiler warning (from some compilers) for a function call with
428        a cast that removes "const" from an lvalue by using an intermediate
429        variable (to which the compiler does not object).
431    19. Incorrect code was compiled if a group that contained an internal recursive
432        back reference was optional (had quantifier with a minimum of zero). This
433        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
434        caused segmentation faults because of stack overflows at compile time.
436    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
437        group that is quantified with an indefinite repeat, caused a compile-time
438        loop which used up all the system stack and provoked a segmentation fault.
439        This was not the same bug as 19 above.
441    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
442        Patch by Mike Frysinger.
445    Version 8.35 04-April-2014
446    --------------------------
448    1.  A new flag is set, when property checks are present in an XCLASS.
449        When this flag is not set, PCRE can perform certain optimizations
450        such as studying these XCLASS-es.
452    2.  The auto-possessification of character sets were improved: a normal
453        and an extended character set can be compared now. Furthermore
454        the JIT compiler optimizes more character set checks.
456    3.  Got rid of some compiler warnings for potentially uninitialized variables
457        that show up only when compiled with -O2.
459    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
460        of a match later then the end of the match. The pcretest program was not
461        handling the case sensibly - it was outputting from the start to the next
462        binary zero. It now reports this situation in a message, and outputs the
463        text from the end to the start.
465    5.  Fast forward search is improved in JIT. Instead of the first three
466        characters, any three characters with fixed position can be searched.
467        Search order: first, last, middle.
469    6.  Improve character range checks in JIT. Characters are read by an inprecise
470        function now, which returns with an unknown value if the character code is
471        above a certain threshold (e.g: 256). The only limitation is that the value
472        must be bigger than the threshold as well. This function is useful when
473        the characters above the threshold are handled in the same way.
475    7.  The macros whose names start with RAWUCHAR are placeholders for a future
476        mode in which only the bottom 21 bits of 32-bit data items are used. To
477        make this more memorable for those maintaining the code, the names have
478        been changed to start with UCHAR21, and an extensive comment has been added
479        to their definition.
481    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
482        to the export list in Makefile.am (they were accidentally omitted from the
483        8.34 tarball).
485    9.  The informational output from pcretest used the phrase "starting byte set"
486        which is inappropriate for the 16-bit and 32-bit libraries. As the output
487        for "first char" and "need char" really means "non-UTF-char", I've changed
488        "byte" to "char", and slightly reworded the output. The documentation about
489        these values has also been (I hope) clarified.
491    10. Another JIT related optimization: use table jumps for selecting the correct
492        backtracking path, when more than four alternatives are present inside a
493        bracket.
495    11. Empty match is not possible, when the minimum length is greater than zero,
496        and there is no \K in the pattern. JIT should avoid empty match checks in
497        such cases.
499    12. In a caseless character class with UCP support, when a character with more
500        than one alternative case was not the first character of a range, not all
501        the alternative cases were added to the class. For example, s and \x{17f}
502        are both alternative cases for S: the class [RST] was handled correctly,
503        but [R-T] was not.
505    13. The configure.ac file always checked for pthread support when JIT was
506        enabled. This is not used in Windows, so I have put this test inside a
507        check for the presence of windows.h (which was already tested for).
509    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
510        The algorithm provides a way to skip certain starting offsets, and usually
511        faster than linear prefix searches.
513    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
514        as for 'fr_FR' and 'french'. For some reason, however, it then used the
515        Windows-specific input and output files, which have 'french' screwed in.
516        So this could never have worked. One of the problems with locales is that
517        they aren't always the same. I have now updated RunTest so that it checks
518        the output of the locale test (test 3) against three different output
519        files, and it allows the test to pass if any one of them matches. With luck
520        this should make the test pass on some versions of Solaris where it was
521        failing. Because of the uncertainty, the script did not used to stop if
522        test 3 failed; it now does. If further versions of a French locale ever
523        come to light, they can now easily be added.
525    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
526        there was a message during ./configure, but it did not stop. This now
527        provokes an error. The invalid example in README has been corrected.
528        If a value less than the minimum is given, the minimum value has always
529        been used, but now a warning is given.
531    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
532        was a bug in the test system, which is now fixed. Also, the list of various
533        configurations that are tested for each release did not have one with both
534        16/32 bits and --enable-bar-anycrlf. It now does.
536    18. pcretest was missing "-C bsr" for displaying the \R default setting.
538    19. Little endian PowerPC systems are supported now by the JIT compiler.
540    20. The fast forward newline mechanism could enter to an infinite loop on
541        certain invalid UTF-8 input. Although we don't support these cases
542        this issue can be fixed by a performance optimization.
544    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
545        not take account if existing stack usage. There is now a new global
546        variable called pcre_stack_guard that can be set to point to an external
547        function to check stack availability. It is called at the start of
548        processing every parenthesized group.
550    22. A typo in the code meant that in ungreedy mode the max/min qualifier
551        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
552        not match "ab".
554    23. When UTF was disabled, the JIT program reported some incorrect compile
555        errors. These messages are silenced now.
557    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
558        compiler.
560    25. Change all the temporary files used in RunGrepTest to be different to those
561        used by RunTest so that the tests can be run simultaneously, for example by
562        "make -j check".
565    Version 8.34 15-December-2013
566    -----------------------------
568    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
569        executable memory. Patch inspired by Carsten Klein.
571    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
572        this macro is never tested and has no effect, because the work to support
573        coverage involves only compiling and linking options and special targets in
574        the Makefile. The comment in config.h implied that defining the macro would
575        enable coverage support, which is totally false. There was also support for
576        setting this macro in the CMake files (my fault, I just copied it from
577        configure). SUPPORT_GCOV has now been removed.
579    3.  Make a small performance improvement in strlen16() and strlen32() in
580        pcretest.
582    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
583        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
584        these cases). There was also one in pcretest.c.
586    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
588    6.  In UTF mode, the code for checking whether a group could match an empty
589        string (which is used for indefinitely repeated groups to allow for
590        breaking an infinite loop) was broken when the group contained a repeated
591        negated single-character class with a character that occupied more than one
592        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
593        UTF-8 mode). The effect was undefined: the group might or might not be
594        deemed as matching an empty string, or the program might have crashed.
596    7.  The code for checking whether a group could match an empty string was not
597        recognizing that \h, \H, \v, \V, and \R must match a character.
599    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
600        an empty string. If it can, pcretest shows this in its information output.
602    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
603        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
604        matched by pcre_exec() without using JIT:
606        (a) If the rest of the pattern did not match after a maximal run of
607            grapheme clusters, the code for backing up to try with fewer of them
608            did not always back up over a full grapheme when characters that do not
609            have the modifier quality were involved, e.g. Hangul syllables.
611        (b) If the match point in a subject started with modifier character, and
612            there was no match, the code could incorrectly back up beyond the match
613            point, and potentially beyond the first character in the subject,
614            leading to a segfault or an incorrect match result.
616    10. A conditional group with an assertion condition could lead to PCRE
617        recording an incorrect first data item for a match if no other first data
618        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
619        first data item, and therefore matched "ca" after "c" instead of at the
620        start.
622    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
623        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
625    12. The source of pcregrep now includes z/OS-specific code so that it can be
626        compiled for z/OS as part of the special z/OS distribution.
628    13. Added the -T and -TM options to pcretest.
630    14. The code in pcre_compile.c for creating the table of named capturing groups
631        has been refactored. Instead of creating the table dynamically during the
632        actual compiling pass, the information is remembered during the pre-compile
633        pass (on the stack unless there are more than 20 named groups, in which
634        case malloc() is used) and the whole table is created before the actual
635        compile happens. This has simplified the code (it is now nearly 150 lines
636        shorter) and prepared the way for better handling of references to groups
637        with duplicate names.
639    15. A back reference to a named subpattern when there is more than one of the
640        same name now checks them in the order in which they appear in the pattern.
641        The first one that is set is used for the reference. Previously only the
642        first one was inspected. This change makes PCRE more compatible with Perl.
644    16. Unicode character properties were updated from Unicode 6.3.0.
646    17. The compile-time code for auto-possessification has been refactored, based
647        on a patch by Zoltan Herczeg. It now happens after instead of during
648        compilation. The code is cleaner, and more cases are handled. The option
649        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
650        options in pcretest are provided to set it. It can also be set by
651        (*NO_AUTO_POSSESS) at the start of a pattern.
653    18. The character VT has been added to the default ("C" locale) set of
654        characters that match \s and are generally treated as white space,
655        following this same change in Perl 5.18. There is now no difference between
656        "Perl space" and "POSIX space". Whether VT is treated as white space in
657        other locales depends on the locale.
659    19. The code for checking named groups as conditions, either for being set or
660        for being recursed, has been refactored (this is related to 14 and 15
661        above). Processing unduplicated named groups should now be as fast at
662        numerical groups, and processing duplicated groups should be faster than
663        before.
665    20. Two patches to the CMake build system, by Alexander Barkov:
667          (1) Replace the "source" command by "." in CMakeLists.txt because
668              "source" is a bash-ism.
670          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
671              without these the CMake build does not work on Solaris.
673    21. Perl has changed its handling of \8 and \9. If there is no previously
674        encountered capturing group of those numbers, they are treated as the
675        literal characters 8 and 9 instead of a binary zero followed by the
676        literals. PCRE now does the same.
678    22. Following Perl, added \o{} to specify codepoints in octal, making it
679        possible to specify values greater than 0777 and also making them
680        unambiguous.
682    23. Perl now gives an error for missing closing braces after \x{... instead of
683        treating the string as literal. PCRE now does the same.
685    24. RunTest used to grumble if an inappropriate test was selected explicitly,
686        but just skip it when running all tests. This make it awkward to run ranges
687        of tests when one of them was inappropriate. Now it just skips any
688        inappropriate tests, as it always did when running all tests.
690    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
691        character types such as \d or \w, too many callouts were inserted, and the
692        data that they returned was rubbish.
694    26. In UCP mode, \s was not matching two of the characters that Perl matches,
695        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
696        were matched by \h. The code has now been refactored so that the lists of
697        the horizontal and vertical whitespace characters used for \h and \v (which
698        are defined only in one place) are now also used for \s.
700    27. Add JIT support for the 64 bit TileGX architecture.
701        Patch by Jiong Wang (Tilera Corporation).
703    28. Possessive quantifiers for classes (both explicit and automatically
704        generated) now use special opcodes instead of wrapping in ONCE brackets.
706    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
707        (because it's meaningless), this was not happening when PCRE_CASELESS was
708        set. Not wrong, but inefficient.
710    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
711        properties for \w, \d, etc) is present in a test regex. Otherwise if the
712        test contains no characters greater than 255, Perl doesn't realise it
713        should be using Unicode semantics.
715    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
716        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
717        does in Unicode mode.
719    32. Added the "forbid" facility to pcretest so that putting tests into the
720        wrong test files can sometimes be quickly detected.
722    33. There is now a limit (default 250) on the depth of nesting of parentheses.
723        This limit is imposed to control the amount of system stack used at compile
724        time. It can be changed at build time by --with-parens-nest-limit=xxx or
725        the equivalent in CMake.
727    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
728        errors. Perl warns for these when in warning mode, but PCRE has no facility
729        for giving warnings.
731    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
732        However, this was not working for (?!) because it is optimized to (*FAIL),
733        for which PCRE does not allow quantifiers. The optimization is now disabled
734        when a quantifier follows (?!). I can't see any use for this, but it makes
735        things uniform.
737    36. Perl no longer allows group names to start with digits, so I have made this
738        change also in PCRE. It simplifies the code a bit.
740    37. In extended mode, Perl ignores spaces before a + that indicates a
741        possessive quantifier. PCRE allowed a space before the quantifier, but not
742        before the possessive +. It now does.
744    38. The use of \K (reset reported match start) within a repeated possessive
745        group such as (a\Kb)*+ was not working.
747    40. Document that the same character tables must be used at compile time and
748        run time, and that the facility to pass tables to pcre_exec() and
749        pcre_dfa_exec() is for use only with saved/restored patterns.
751    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
752        features for Builds with MSVC:
754        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
755           builds). The .rc files can be used to set FileDescription and many other
756           attributes.
758        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
759           This allows higher-level build scripts which want .pdb files to avoid
760           hard-coding the exact files needed."
762    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
763        mean "start of word" and "end of word", respectively, as a transition aid.
765    43. A minimizing repeat of a class containing codepoints greater than 255 in
766        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
767        compiled to use the heap for recursion.
769    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
770        is configured.
773    Version 8.33 28-May-2013
774    ------------------------
776    1.  Added 'U' to some constants that are compared to unsigned integers, to
777        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
778        variables that are added to signed variables, to ensure the result is
779        signed and can be negated.
781    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
782        CMake config files.
784    3.  Revise the creation of config.h.generic so that all boolean macros are
785        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
786        overriding via -D on the command line possible.
788    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
789        to unsigned int is reported to make a quite noticeable speed difference in
790        a specific Windows environment. Testing on Linux did also appear to show
791        some benefit (and it is clearly not harmful). Also fixed the definition of
792        Xop which should be unsigned.
794    5.  Related to (4), changing the definition of the intermediate variable cc
795        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
796        improvements.
798    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
799        unnecessary spaces.
801    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
802        and later.
804    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
806    9.  Optimizing fast_forward_start_bits in JIT.
808    10. Adding support for callouts in JIT, and fixing some issues revealed
809        during this work. Namely:
811        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
813        (b) Minimum length was not checked before the matching is started.
815    11. The value of capture_last that is passed to callouts was incorrect in some
816        cases when there was a capture on one path that was subsequently abandoned
817        after a backtrack. Also, the capture_last value is now reset after a
818        recursion, since all captures are also reset in this case.
820    12. The interpreter no longer returns the "too many substrings" error in the
821        case when an overflowing capture is in a branch that is subsequently
822        abandoned after a backtrack.
824    13. In the pathological case when an offset vector of size 2 is used, pcretest
825        now prints out the matched string after a yield of 0 or 1.
827    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
828        Only supported by the JIT compiler at the moment.
830    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
832    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
833        offset of the starting point of the matching process, provided the offsets
834        vector is large enough.
836    17. The \A escape now records a lookbehind value of 1, though its execution
837        does not actually inspect the previous character. This is to ensure that,
838        in partial multi-segment matching, at least one character from the old
839        segment is retained when a new segment is processed. Otherwise, if there
840        are no lookbehinds in the pattern, \A might match incorrectly at the start
841        of a new segment.
843    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
845    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
846        gives some modest performance improvement in 8-bit mode.
848    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
849        be expressed in certain programming languages using Universal Character
850        Names.
852    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
853        which points out that "non characters" are not "characters that may not
854        appear in Unicode strings" but rather "characters that are reserved for
855        internal use and have only local meaning".
857    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
858        there was a conditional group that depended on an assertion, if the
859        assertion was false, the callout that immediately followed the alternation
860        in the condition was skipped when pcre_exec() was used for matching.
862    23. Allow an explicit callout to be inserted before an assertion that is the
863        condition for a conditional group, for compatibility with automatic
864        callouts, which always insert a callout at this point.
866    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
867        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
869    25. (*PRUNE) is now supported by the JIT compiler.
871    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
873    27. Fix the case where there are two or more SKIPs with arguments that may be
874        ignored.
876    28. (*SKIP) is now supported by the JIT compiler.
878    29. (*THEN) is now supported by the JIT compiler.
880    30. Update RunTest with additional test selector options.
882    31. The way PCRE handles backtracking verbs has been changed in two ways.
884        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
885        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
886        backtracking. In some cases this makes it more Perl-compatible, but Perl's
887        rather obscure rules do not always do the same thing.
889        (2) Previously, backtracking verbs were confined within assertions. This is
890        no longer the case for positive assertions, except for (*ACCEPT). Again,
891        this sometimes improves Perl compatibility, and sometimes does not.
893    32. A number of tests that were in test 2 because Perl did things differently
894        have been moved to test 1, because either Perl or PCRE has changed, and
895        these tests are now compatible.
897    32. Backtracking control verbs are now handled in the same way in JIT and
898        interpreter.
900    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
901        contained a forward subroutine reference caused a compile error.
903    34. Auto-detect and optimize limited repetitions in JIT.
905    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
906        blocking (*UTF) etc.
908    36. In the interpreter, maximizing pattern repetitions for characters and
909        character types now use tail recursion, which reduces stack usage.
911    37. The value of the max lookbehind was not correctly preserved if a compiled
912        and saved regex was reloaded on a host of different endianness.
914    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
915        of the compiled pattern block, expand the flags field from 16 to 32 bits
916        because it was almost full.
918    39. Try madvise first before posix_madvise.
920    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
921        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
922        original change locked out all matches of empty strings. This has been
923        changed so that one match of an empty string per line is recognized.
924        Subsequent searches on the same line (for colouring or for --only-matching,
925        for example) do not recognize empty strings.
927    41. Applied a user patch to fix a number of spelling mistakes in comments.
929    42. Data lines longer than 65536 caused pcretest to crash.
931    43. Clarified the data type for length and startoffset arguments for pcre_exec
932        and pcre_dfa_exec in the function-specific man pages, where they were
933        explicitly stated to be in bytes, never having been updated. I also added
934        some clarification to the pcreapi man page.
936    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
937        a segmentation fault.
940    Version 8.32 30-November-2012
941    -----------------------------
943    1.  Improved JIT compiler optimizations for first character search and single
944        character iterators.
946    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
947        Patch by Daniel Richard G.
949    3.  Single character iterator optimizations in the JIT compiler.
951    4.  Improved JIT compiler optimizations for character ranges.
953    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
954        Reported by Giuseppe D'Angelo.
956    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
957        of a line, was being set incorrectly in cases where .* appeared inside
958        atomic brackets at the start of a pattern, or where there was a subsequent
959        *PRUNE or *SKIP.
961    7.  Improved instruction cache flush for POWER/PowerPC.
962        Patch by Daniel Richard G.
964    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
965        grep:
967        (a) There is now no limit to the number of patterns to be matched.
969        (b) An error is given if a pattern is too long.
971        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
972            are now supported.
974        (d) --exclude-from and --include-from (multiple use) have been added.
976        (e) Exclusions and inclusions now apply to all files and directories, not
977            just to those obtained from scanning a directory recursively.
979        (f) Multiple uses of -f and --file-list are now supported.
981        (g) In a Windows environment, the default for -d has been changed from
982            "read" (the GNU grep default) to "skip", because otherwise the presence
983            of a directory in the file list provokes an error.
985        (h) The documentation has been revised and clarified in places.
987    9.  Improve the matching speed of capturing brackets.
989    10. Changed the meaning of \X so that it now matches a Unicode extended
990        grapheme cluster.
992    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
993        out POSIX threads when JIT support is configured.
995    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
997    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
998        structure could go wrong in environments where size_t is not the same size
999        as int.
1001    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
1002        set.
1004    15. The EBCDIC support had decayed; later updates to the code had included
1005        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
1006        general tidy up of EBCDIC-related issues, and the documentation was also
1007        not quite right. There is now a test that can be run on ASCII systems to
1008        check some of the EBCDIC-related things (but is it not a full test).
1010    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
1011        in a small tidy to the code.
1013    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
1015    18. If the --only-matching (-o) option in pcregrep is specified multiple
1016        times, each one causes appropriate output. For example, -o1 -o2 outputs the
1017        substrings matched by the 1st and 2nd capturing parentheses. A separating
1018        string can be specified by --om-separator (default empty).
1020    19. Improving the first n character searches.
1022    20. Turn case lists for horizontal and vertical white space into macros so that
1023        they are defined only once.
1025    21. This set of changes together give more compatible Unicode case-folding
1026        behaviour for characters that have more than one other case when UCP
1027        support is available.
1029        (a) The Unicode property table now has offsets into a new table of sets of
1030            three or more characters that are case-equivalent. The MultiStage2.py
1031            script that generates these tables (the pcre_ucd.c file) now scans
1032            CaseFolding.txt instead of UnicodeData.txt for character case
1033            information.
1035        (b) The code for adding characters or ranges of characters to a character
1036            class has been abstracted into a generalized function that also handles
1037            case-independence. In UTF-mode with UCP support, this uses the new data
1038            to handle characters with more than one other case.
1040        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
1041            whose other case is greater than 256 are now correctly matched
1042            caselessly. Previously, the high codepoint matched the low one, but not
1043            vice versa.
1045        (d) The processing of \h, \H, \v, and \ in character classes now makes use
1046            of the new class addition function, using character lists defined as
1047            macros alongside the case definitions of 20 above.
1049        (e) Caseless back references now work with characters that have more than
1050            one other case.
1052        (f) General caseless matching of characters with more than one other case
1053            is supported.
1055    22. Unicode character properties were updated from Unicode 6.2.0
1057    23. Improved CMake support under Windows. Patch by Daniel Richard G.
1059    24. Add support for 32-bit character strings, and UTF-32
1061    25. Major JIT compiler update (code refactoring and bugfixing).
1062        Experimental Sparc 32 support is added.
1064    26. Applied a modified version of Daniel Richard G's patch to create
1065        pcre.h.generic and config.h.generic by "make" instead of in the
1066        PrepareRelease script.
1068    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
1069        pcre_compile.c when checking for a zero character.
1071    28. Introducing a native interface for JIT. Through this interface, the compiled
1072        machine code can be directly executed. The purpose of this interface is to
1073        provide fast pattern matching, so several sanity checks are not performed.
1074        However, feature tests are still performed. The new interface provides
1075        1.4x speedup compared to the old one.
1077    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
1078        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
1079        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
1081    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
1082        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
1083        the "old" RFC 2279). Instead, it ended up passing a negative length to
1084        pcre_exec().
1086    31. Add support for GCC's visibility feature to hide internal functions.
1088    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
1089        "unknown -C option" after outputting 0 or 1.
1091    33. There is now support for generating a code coverage report for the test
1092        suite in environments where gcc is the compiler and lcov is installed. This
1093        is mainly for the benefit of the developers.
1095    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
1096        unaddressable using valgrind annotations, allowing valgrind to detect
1097        invalid memory accesses. This is mainly for the benefit of the developers.
1099    25. (*UTF) can now be used to start a pattern in any of the three libraries.
1101    26. Give configure error if --enable-cpp but no C++ compiler found.
1104    Version 8.31 06-July-2012
1105    -------------------------
1107    1.  Fixing a wrong JIT test case and some compiler warnings.
1109    2.  Removed a bashism from the RunTest script.
1111    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
1112        to unsigned type, result still unsigned" that was given by an MS compiler
1113        on encountering the code "-sizeof(xxx)".
1115    4.  Partial matching support is added to the JIT compiler.
1117    5.  Fixed several bugs concerned with partial matching of items that consist
1118        of more than one character:
1120        (a) /^(..)\1/ did not partially match "aba" because checking references was
1121            done on an "all or nothing" basis. This also applied to repeated
1122            references.
1124        (b) \R did not give a hard partial match if \r was found at the end of the
1125            subject.
1127        (c) \X did not give a hard partial match after matching one or more
1128            characters at the end of the subject.
1130        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1131            a partial match for the string "\r".
1133        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1134            a partial match for a CR character at the end of the subject string.
1136    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1137        running pcretest, the text "(JIT)" added to the output whenever JIT is
1138        actually used to run the match.
1140    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1141        or /S+[+] with a digit between 1 and 7.
1143    8.  OP_NOT now supports any UTF character not just single-byte ones.
1145    9.  (*MARK) control verb is now supported by the JIT compiler.
1147    10. The command "./RunTest list" lists the available tests without actually
1148        running any of them. (Because I keep forgetting what they all are.)
1152    12. Applied a (slightly modified) user-supplied patch that improves performance
1153        when the heap is used for recursion (compiled with --disable-stack-for-
1154        recursion). Instead of malloc and free for each heap frame each time a
1155        logical recursion happens, frames are retained on a chain and re-used where
1156        possible. This sometimes gives as much as 30% improvement.
1158    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1159        call.
1161    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1163    15. It is now possible to link pcretest with libedit as an alternative to
1164        libreadline.
1166    16. (*COMMIT) control verb is now supported by the JIT compiler.
1168    17. The Unicode data tables have been updated to Unicode 6.1.0.
1170    18. Added --file-list option to pcregrep.
1172    19. Added binary file support to pcregrep, including the -a, --binary-files,
1173        -I, and --text options.
1175    20. The madvise function is renamed for posix_madvise for QNX compatibility
1176        reasons. Fixed by Giuseppe D'Angelo.
1178    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1180    22. Changed the option for creating symbolic links for 16-bit man pages from
1181        -s to -sf so that re-installing does not cause issues.
1183    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1185    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1186        matches in certain environments (the workspace was not being correctly
1187        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1188        some of the workspace data at the beginning of a restart.
1190    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1191        was not doing so when it should - probably a typo introduced by SVN 528
1192        (change 8.10/14).
1194    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1195        \w+ when the character tables indicated that \x{c4} was a word character.
1196        There were several related cases, all because the tests for doing a table
1197        lookup were testing for characters less than 127 instead of 255.
1199    27. If a pattern contains capturing parentheses that are not used in a match,
1200        their slots in the ovector are set to -1. For those that are higher than
1201        any matched groups, this happens at the end of processing. In the case when
1202        there were back references that the ovector was too small to contain
1203        (causing temporary malloc'd memory to be used during matching), and the
1204        highest capturing number was not used, memory off the end of the ovector
1205        was incorrectly being set to -1. (It was using the size of the temporary
1206        memory instead of the true size.)
1208    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1209        ovector size, it uses memory at the end of the block that it has got.
1211    29. Check for an overlong MARK name and give an error at compile time. The
1212        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1214    30. JIT compiler update.
1216    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1217        Rill for the patch.
1219    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1221    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1223    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1224        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1225        was enabled.
1227    35. Improve JIT code generation for greedy plus quantifier.
1229    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1230        1 to "aa" instead of to an empty string. The bug affected repeated groups
1231        that could potentially match an empty string.
1233    37. Optimizing single character iterators in JIT.
1235    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1236        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1237        codepoints that are too big for the mode are faulted, and in a UTF mode,
1238        disallowed codepoints are also faulted.
1240    39. If PCRE was compiled with UTF support, in three places in the DFA
1241        matcher there was code that should only have been obeyed in UTF mode, but
1242        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1243        processing when bytes with values greater than 127 were present. In 16-bit
1244        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1245        both cases the values are those that cannot be the first data item in a UTF
1246        character. The three items that might have provoked this were recursions,
1247        possessively repeated groups, and atomic groups.
1249    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1250        and pcregrep, because some OS require shared objects to be explicitly
1251        passed to ld, causing the link step to fail if they are not.
1253    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1254        mode, patterns that started with \h* or \R* might be incorrectly matched.
1257    Version 8.30 04-February-2012
1258    -----------------------------
1260    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1261        name is defined in ctype.h.
1263    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1264        only in quite long subpatterns.
1266    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1267        since it was replaced by pcre_fullinfo() in February 2000.
1269    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1270        match a (*MARK), and the match failed at the start of the subject, a
1271        reference to memory before the start of the subject could occur. This bug
1272        was introduced by fix 17 of release 8.21.
1274    5.  A reference to an unset group with zero minimum repetition was giving
1275        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1276        /(another)?(\1?)test/ matched against "hello world test". This bug was
1277        introduced in release 8.13.
1279    6.  Add support for 16-bit character strings (a large amount of work involving
1280        many changes and refactorings).
1282    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1283        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1284        from a file.
1286    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1287        rounding is not applied in this particular case).
1289    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1290        if they appear, or are escaped, in patterns.
1292    10. Get rid of a number of -Wunused-but-set-variable warnings.
1294    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1295        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1296        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1297        also returns the mark "x". This bug applied to capturing parentheses,
1298        non-capturing parentheses, and atomic parentheses. It also applied to some
1299        assertions.
1301    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1302        information out of configure.ac instead of relying on pcre.h.generic, which
1303        is not stored in the repository.
1305    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1306        -lreadline.
1308    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1310    15. Applied Graycode's patch to put the top-level frame on the stack rather
1311        than the heap when not using the stack for recursion. This gives a
1312        performance improvement in many cases when recursion is not deep.
1314    16. Experimental code added to "pcretest -C" to output the stack frame size.
1317    Version 8.21 12-Dec-2011
1318    ------------------------
1320    1.  Updating the JIT compiler.
1322    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1323        are added as well.
1325    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1326        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1327        calling _pcre_jit_exec. Some extra comments are added.
1329    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1330        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1331        was introduced by change 18 for 8.20.
1333    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1334        ECMA-262 standard.
1336    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1337        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1338        This bug was probably introduced by change 9 of 8.13.
1340    7.  While fixing 6 above, I noticed that a number of other items were being
1341        incorrectly rejected as "not fixed length". This arose partly because newer
1342        opcodes had not been added to the fixed-length checking code. I have (a)
1343        corrected the bug and added tests for these items, and (b) arranged for an
1344        error to occur if an unknown opcode is encountered while checking for fixed
1345        length instead of just assuming "not fixed length". The items that were
1346        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1347        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1348        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1350    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1351        being incorrectly compiled and would have given unpredicatble results.
1353    9.  A possessively repeated subpattern with minimum repeat count greater than
1354        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1355        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1356        the first (A) could occur when it should not.
1358    10. Add a cast and remove a redundant test from the code.
1360    11. JIT should use pcre_malloc/pcre_free for allocation.
1362    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1363        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1364        is anything other than /usr, -L is still shown).
1366    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1368    14. Perl does not support \N without a following name in a [] class; PCRE now
1369        also gives an error.
1371    15. If a forward reference was repeated with an upper limit of around 2000,
1372        it caused the error "internal error: overran compiling workspace". The
1373        maximum number of forward references (including repeats) was limited by the
1374        internal workspace, and dependent on the LINK_SIZE. The code has been
1375        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1376        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1377        of around 200,000 forward references. While doing this, I also speeded up
1378        the filling in of repeated forward references.
1380    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1381        incorrectly expecting the subject to contain another "a" after the start.
1383    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1384        in the match, the SKIP should be ignored. This was not happening; instead
1385        the SKIP was being treated as NOMATCH. For patterns such as
1386        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1387        tested.
1389    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1390        now much more compatible with Perl, in particular in cases where the result
1391        is a non-match for a non-anchored pattern. For example, if
1392        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1393        "m", where previously it did not return a name. A side effect of this
1394        change is that for partial matches, the last encountered mark name is
1395        returned, as for non matches. A number of tests that were previously not
1396        Perl-compatible have been moved into the Perl-compatible test files. The
1397        refactoring has had the pleasing side effect of removing one argument from
1398        the match() function, thus reducing its stack requirements.
1400    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1401        subsequent uses of /S (without +) incorrectly behaved like /S+.
1403    21. Retrieve executable code size support for the JIT compiler and fixing
1404        some warnings.
1406    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1407        not work when the shorter character appeared right at the end of the
1408        subject string.
1410    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1411        systems.
1413    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1414        output it when the /M option is used in pcretest.
1416    25. The CheckMan script was not being included in the distribution. Also, added
1417        an explicit "perl" to run Perl scripts from the PrepareRelease script
1418        because this is reportedly needed in Windows.
1420    26. If study data was being save in a file and studying had not found a set of
1421        "starts with" bytes for the pattern, the data written to the file (though
1422        never used) was taken from uninitialized memory and so caused valgrind to
1423        complain.
1425    27. Updated RunTest.bat as provided by Sheri Pierce.
1427    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1429    29. Computation of memory usage for the table of capturing group names was
1430        giving an unnecessarily large value.
1433    Version 8.20 21-Oct-2011
1434    ------------------------
1436    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1437        a POSIX class. After further experiments with Perl, which convinced me that
1438        Perl has bugs and confusions, a closing square bracket is no longer allowed
1439        in a POSIX name. This bug also affected patterns with classes that started
1440        with full stops.
1442    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1443        captured substring, but while checking the failing first alternative,
1444        substring 1 is temporarily captured. If the output vector supplied to
1445        pcre_exec() was not big enough for this capture, the yield of the function
1446        was still zero ("insufficient space for captured substrings"). This cannot
1447        be totally fixed without adding another stack variable, which seems a lot
1448        of expense for a edge case. However, I have improved the situation in cases
1449        such as /(a)(b)x|abc/ matched against "abc", where the return code
1450        indicates that fewer than the maximum number of slots in the ovector have
1451        been set.
1453    3.  Related to (2) above: when there are more back references in a pattern than
1454        slots in the output vector, pcre_exec() uses temporary memory during
1455        matching, and copies in the captures as far as possible afterwards. It was
1456        using the entire output vector, but this conflicts with the specification
1457        that only 2/3 is used for passing back captured substrings. Now it uses
1458        only the first 2/3, for compatibility. This is, of course, another edge
1459        case.
1461    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1462        main code base, and can be used by building with --enable-jit. When this is
1463        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1464        runtime --no-jit option is given.
1466    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1467        ovector, the return from the function was zero, implying that there were
1468        other matches that did not fit. The correct "exactly full" value is now
1469        returned.
1471    6.  If a subpattern that was called recursively or as a subroutine contained
1472        (*PRUNE) or any other control that caused it to give a non-standard return,
1473        invalid errors such as "Error -26 (nested recursion at the same subject
1474        position)" or even infinite loops could occur.
1476    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1477        computing the minimum length on reaching *ACCEPT, and so ended up with the
1478        wrong value of 1 rather than 0. Further investigation indicates that
1479        computing a minimum subject length in the presence of *ACCEPT is difficult
1480        (think back references, subroutine calls), and so I have changed the code
1481        so that no minimum is registered for a pattern that contains *ACCEPT.
1483    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1484        it was not handled as intended. [But see 16 below.]
1486    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1487        Sheri Pierce.
1489    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1490        the first byte in a match must be "a".
1492    11. Change 17 for 8.13 increased the recursion depth for patterns like
1493        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1494        pattern contains any instances of (*THEN). If it does not, the old
1495        optimizations are restored. It would be nice to do this on a per-group
1496        basis, but at the moment that is not feasible.
1498    12. In some environments, the output of pcretest -C is CRLF terminated. This
1499        broke RunTest's code that checks for the link size. A single white space
1500        character after the value is now allowed for.
1502    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1503        For "fr", it uses the Windows-specific input and output files.
1505    14. If (*THEN) appeared in a group that was called recursively or as a
1506        subroutine, it did not work as intended. [But see next item.]
1508    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1509        pattern fragments (but not containing any | characters). If A and B are
1510        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1511        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1512        D. In other words, Perl considers parentheses that do not contain any |
1513        characters to be part of a surrounding alternative, whereas PCRE was
1514        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1515        differently. PCRE now behaves in the same way as Perl, except in the case
1516        of subroutine/recursion calls such as (?1) which have in any case always
1517        been different (but PCRE had them first :-).
1519    16. Related to 15 above: Perl does not treat the | in a conditional group as
1520        creating alternatives. Such a group is treated in the same way as an
1521        ordinary group without any | characters when processing (*THEN). PCRE has
1522        been changed to match Perl's behaviour.
1524    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1525        RunGrepTest script failed.
1527    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1528        inevitable for groups that contain captures, but it can lead to a lot of
1529        stack use in large patterns. The old behaviour has been restored for atomic
1530        groups that do not contain any capturing parentheses.
1532    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1533        suppress the check for a minimum subject length at run time. (If it was
1534        given to pcre_exec() or pcre_dfa_exec() it did work.)
1536    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1537        fail to work when decoding hex characters in data strings in EBCDIC
1538        environments.
1540    21. It appears that in at least one Mac OS environment, the isxdigit() function
1541        is implemented as a macro that evaluates to its argument more than once,
1542        contravening the C 90 Standard (I haven't checked a later standard). There
1543        was an instance in pcretest which caused it to go wrong when processing
1544        \x{...} escapes in subject strings. The has been rewritten to avoid using
1545        things like p++ in the argument of isxdigit().
1548    Version 8.13 16-Aug-2011
1549    ------------------------
1551    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1553    2.  Two minor typos in pcre_internal.h have been fixed.
1555    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1556        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1557        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1559    4.  There were a number of related bugs in the code for matching backrefences
1560        caselessly in UTF-8 mode when codes for the characters concerned were
1561        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1562        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1563        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1564        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1565        2-byte code at the end of the subject (it thought there wasn't enough data
1566        left).
1568    5.  Comprehensive information about what went wrong is now returned by
1569        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1570        as the output vector has at least 2 elements. The offset of the start of
1571        the failing character and a reason code are placed in the vector.
1573    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1574        now returned is for the first byte of the failing character, instead of the
1575        last byte inspected. This is an incompatible change, but I hope it is small
1576        enough not to be a problem. It makes the returned offset consistent with
1577        pcre_exec() and pcre_dfa_exec().
1579    7.  pcretest now gives a text phrase as well as the error number when
1580        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1581        failure, the offset and reason code are output.
1583    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1584        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1585        back over a single character (\n). This seems wrong (because it treated the
1586        two characters as a single entity when going forwards), conflicts with the
1587        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1588        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1589        has been changed.
1591    9.  Some internal refactoring has changed the processing so that the handling
1592        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1593        time (the PCRE_DOTALL option was changed this way some time ago: version
1594        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1595        which was always a bit of a fudge. It also means that there is one less
1596        argument for the match() function, which reduces its stack requirements
1597        slightly. This change also fixes an incompatibility with Perl: the pattern
1598        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1600    10. More internal refactoring has drastically reduced the number of recursive
1601        calls to match() for possessively repeated groups such as (abc)++ when
1602        using pcre_exec().
1604    11. While implementing 10, a number of bugs in the handling of groups were
1605        discovered and fixed:
1607        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1608        (a|)*(?1) gave a compile-time internal error.
1609        ((a|)+)+  did not notice that the outer group could match an empty string.
1610        (^a|^)+   was not marked as anchored.
1611        (.*a|.*)+ was not marked as matching at start or after a newline.
1613    12. Yet more internal refactoring has removed another argument from the match()
1614        function. Special calls to this function are now indicated by setting a
1615        value in a variable in the "match data" data block.
1617    13. Be more explicit in pcre_study() instead of relying on "default" for
1618        opcodes that mean there is no starting character; this means that when new
1619        ones are added and accidentally left out of pcre_study(), testing should
1620        pick them up.
1622    14. The -s option of pcretest has been documented for ages as being an old
1623        synonym of -m (show memory usage). I have changed it to mean "force study
1624        for every regex", that is, assume /S for every regex. This is similar to -i
1625        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1626        using it. It makes it easier to run collections of tests with and without
1627        study enabled, and thereby test pcre_study() more easily. All the standard
1628        tests are now run with and without -s (but some patterns can be marked as
1629        "never study" - see 20 below).
1631    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1632        restoration of the capturing data to the outer values was not happening
1633        correctly.
1635    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1636        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1637        pattern had matched an empty string, and so incorrectly returned a no
1638        match.
1640    17. There was optimizing code for the last branch of non-capturing parentheses,
1641        and also for the obeyed branch of a conditional subexpression, which used
1642        tail recursion to cut down on stack usage. Unfortunately, now that there is
1643        the possibility of (*THEN) occurring in these branches, tail recursion is
1644        no longer possible because the return has to be checked for (*THEN). These
1645        two optimizations have therefore been removed. [But see 8.20/11 above.]
1647    18. If a pattern containing \R was studied, it was assumed that \R always
1648        matched two bytes, thus causing the minimum subject length to be
1649        incorrectly computed because \R can also match just one byte.
1651    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1652        was incorrectly computed.
1654    20. If /S is present twice on a test pattern in pcretest input, it now
1655        *disables* studying, thereby overriding the use of -s on the command line
1656        (see 14 above). This is necessary for one or two tests to keep the output
1657        identical in both cases.
1659    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1660        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1662    22. When an atomic group that contained a capturing parenthesis was
1663        successfully matched, but the branch in which it appeared failed, the
1664        capturing was not being forgotten if a higher numbered group was later
1665        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1666        group 1 to "a", when in fact it should be unset. This applied to multi-
1667        branched capturing and non-capturing groups, repeated or not, and also to
1668        positive assertions (capturing in negative assertions does not happen
1669        in PCRE) and also to nested atomic groups.
1671    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1672        subject after a captured substring, to make it easier to tell which of a
1673        number of identical substrings has been captured.
1675    24. The way atomic groups are processed by pcre_exec() has been changed so that
1676        if they are repeated, backtracking one repetition now resets captured
1677        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1678        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1679        "aaa". Previously, it would have been "a". As part of this code
1680        refactoring, the way recursive calls are handled has also been changed.
1682    25. If an assertion condition captured any substrings, they were not passed
1683        back unless some other capturing happened later. For example, if
1684        (?(?=(a))a) was matched against "a", no capturing was returned.
1686    26. When studying a pattern that contained subroutine calls or assertions,
1687        the code for finding the minimum length of a possible match was handling
1688        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1689        group 1 called group 2 while simultaneously a separate group 2 called group
1690        1). A stack overflow occurred in this case. I have fixed this by limiting
1691        the recursion depth to 10.
1693    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1694        Fortmann. This supports explicit test numbers on the command line, and has
1695        argument validation and error reporting.
1697    28. An instance of \X with an unlimited repeat could fail if at any point the
1698        first character it looked at was a mark character.
1700    29. Some minor code refactoring concerning Unicode properties and scripts
1701        should reduce the stack requirement of match() slightly.
1703    30. Added the '=' option to pcretest to check the setting of unused capturing
1704        slots at the end of the pattern, which are documented as being -1, but are
1705        not included in the return count.
1707    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1708        compiled something random. Now it gives a compile-time error (as does
1709        Perl).
1711    32. A *MARK encountered during the processing of a positive assertion is now
1712        recorded and passed back (compatible with Perl).
1714    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1715        had alternative anchored branches, the search for a second match in a line
1716        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1717        matched the line "0102" twice. The same bug affected patterns that started
1718        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1719        twice.
1721    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1722        does, and because of capturing effects, quantifying parenthesized
1723        assertions may at times be useful. Quantifiers are now allowed for
1724        parenthesized assertions.
1726    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1728    36. \g was being checked for fancy things in a character class, when it should
1729        just be a literal "g".
1731    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1732        appearance of a nested POSIX class supersedes an apparent external class.
1733        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1734        unescaped square brackets may also appear as part of class names. For
1735        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1736        more like Perl. (But see 8.20/1 above.)
1738    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1739        was because it thought it was \N{name}, which is not supported).
1741    39. Add minix to OS list not supporting the -S option in pcretest.
1743    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1744        cannot analyze patterns in sufficient detail to catch mutual recursions
1745        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1746        subgroup is called recursively as a subpattern for a second time at the
1747        same position in the subject string. In previous releases this might have
1748        been caught by the recursion limit, or it might have run out of stack.
1750    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1751        happen only once. PCRE was, however incorrectly giving a compile time error
1752        "recursive call could loop indefinitely" because it cannot analyze the
1753        pattern in sufficient detail. The compile time test no longer happens when
1754        PCRE is compiling a conditional subpattern, but actual runaway loops are
1755        now caught at runtime (see 40 above).
1757    42. It seems that Perl allows any characters other than a closing parenthesis
1758        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1759        has been changed to be the same.
1761    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1762        as not to get warnings when autogen.sh is called. Also changed
1763        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1765    44. To help people who use pcregrep to scan files containing exceedingly long
1766        lines, the following changes have been made:
1768        (a) The default value of the buffer size parameter has been increased from
1769            8K to 20K. (The actual buffer used is three times this size.)
1771        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1772            PCRE is built.
1774        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1775            to be set at run time.
1777        (d) Numerical values in pcregrep options can be followed by K or M, for
1778            example --buffer-size=50K.
1780        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1781            given and the return code is set to 2.
1783    45. Add a pointer to the latest mark to the callout data block.
1785    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1786        partial match of an empty string instead of no match. This was specific to
1787        the use of ".".
1789    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1790        complete match instead of a partial match. This bug was dependent on both
1791        the PCRE_UTF8 and PCRE_DOTALL options being set.
1793    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1794        starting byte set, because \b was not being ignored.
1797    Version 8.12 15-Jan-2011
1798    ------------------------
1800    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1801        checks for such things as part of the documentation building process.
1803    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1804        --match-limit and --recursion-limit options (added for 8.11). In
1805        particular, this made one of the standard tests fail. (The integer value
1806        went into the wrong half of a long int.)
1808    3.  If the --colour option was given to pcregrep with -v (invert match), it
1809        did strange things, either producing crazy output, or crashing. It should,
1810        of course, ignore a request for colour when reporting lines that do not
1811        match.
1813    4.  Another pcregrep bug caused similar problems if --colour was specified with
1814        -M (multiline) and the pattern match finished with a line ending.
1816    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1817        matched in multiline mode, the following line was shown as part of the
1818        match. This seems wrong, so I have changed it.
1820    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1821        the check for further matches in the same line (so they could be coloured)
1822        to overrun the end of the current line. If another match was found, it was
1823        incorrectly shown (and then shown again when found in the next line).
1825    7.  If pcregrep was compiled under Windows, there was a reference to the
1826        function pcregrep_exit() before it was defined. I am assuming this was
1827        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1828        reported by a user. I've moved the definition above the reference.
1831    Version 8.11 10-Dec-2010
1832    ------------------------
1834    1.  (*THEN) was not working properly if there were untried alternatives prior
1835        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1836        backtracked to try for "b" instead of moving to the next alternative branch
1837        at the same level (in this case, to look for "c"). The Perl documentation
1838        is clear that when (*THEN) is backtracked onto, it goes to the "next
1839        alternative in the innermost enclosing group".
1841    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1842        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1843        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1844        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1845        (*THEN).
1847    3.  If \s appeared in a character class, it removed the VT character from
1848        the class, even if it had been included by some previous item, for example
1849        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1850        of \s, but is part of the POSIX "space" class.)
1852    4.  A partial match never returns an empty string (because you can always
1853        match an empty string at the end of the subject); however the checking for
1854        an empty string was starting at the "start of match" point. This has been
1855        changed to the "earliest inspected character" point, because the returned
1856        data for a partial match starts at this character. This means that, for
1857        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1858        (previously it gave "no match").
1860    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1861        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1862        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1863        has an implication that the given string is incomplete (because a partial
1864        match is preferred over a full match). For this reason, these items now
1865        give a partial match in this situation. [Aside: previously, the one case
1866        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1867        match rather than a full match, which was wrong by the old rules, but is
1868        now correct.]
1870    6.  There was a bug in the handling of #-introduced comments, recognized when
1871        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1872        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1873        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1874        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1875        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1876        places in pcre_compile().
1878    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1879        comments when looking ahead for named forward references to subpatterns,
1880        the only newline sequence it recognized was NL. It now handles newlines
1881        according to the set newline convention.
1883    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1884        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1885        cater for a lack of strerror(). These oversights have been fixed.
1887    9.  Added --match-limit and --recursion-limit to pcregrep.
1889    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1891    11. When the -o option was used, pcregrep was setting a return code of 1, even
1892        when matches were found, and --line-buffered was not being honoured.
1894    12. Added an optional parentheses number to the -o and --only-matching options
1895        of pcregrep.
1897    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1898        can match an empty string. The code to do it in pcretest and pcredemo
1899        needed fixing:
1901        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1902            only one byte after an empty string match just before CRLF (this case
1903            just got forgotten; "any" and "anycrlf" were OK).
1905        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1906            mode when an empty string match preceded an ASCII character followed by
1907            a non-ASCII character. (The code for advancing by one character rather
1908            than one byte was nonsense.)
1910        (c) The pcredemo.c sample program did not have any code at all to handle
1911            the cases when CRLF is a valid newline sequence.
1913    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1914        as a starting offset was within the subject string. There is now a new
1915        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1916        negative or greater than the length of the string. In order to test this,
1917        pcretest is extended to allow the setting of negative starting offsets.
1919    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1920        starting offset points to the beginning of a UTF-8 character was
1921        unnecessarily clumsy. I tidied it up.
1923    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1924        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1926    17. Nobody had reported that the --include_dir option, which was added in
1927        release 7.7 should have been called --include-dir (hyphen, not underscore)
1928        for compatibility with GNU grep. I have changed it to --include-dir, but
1929        left --include_dir as an undocumented synonym, and the same for
1930        --exclude-dir, though that is not available in GNU grep, at least as of
1931        release 2.5.4.
1933    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1934        characters from a string of bytes) have been redefined so as not to use
1935        loops, in order to improve performance in some environments. At the same
1936        time, I abstracted some of the common code into auxiliary macros to save
1937        repetition (this should not affect the compiled code).
1939    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1940        compile-time error is now given if \c is not followed by an ASCII
1941        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1942        different, and any byte value is allowed.)
1944    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1945        START_OPTIMIZE option, which is now allowed at compile time - but just
1946        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1947        to pcregrep and other applications that have no direct access to PCRE
1948        options. The new /Y option in pcretest sets this option when calling
1949        pcre_compile().
1951    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1952        back references. Groups containing recursive back references were forced to
1953        be atomic by that change, but in the case of named groups, the amount of
1954        memory required was incorrectly computed, leading to "Failed: internal
1955        error: code overflow". This has been fixed.
1957    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1958        pcretest.c, to avoid build problems in some Borland environments.
1961    Version 8.10 25-Jun-2010
1962    ------------------------
1964    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1965        THEN.
1967    2.  (*ACCEPT) was not working when inside an atomic group.
1969    3.  Inside a character class, \B is treated as a literal by default, but
1970        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1971        causes the error). The code is unchanged, but I tidied the documentation.
1973    4.  Inside a character class, PCRE always treated \R and \X as literals,
1974        whereas Perl faults them if its -w option is set. I have changed PCRE so
1975        that it faults them when PCRE_EXTRA is set.
1977    5.  Added support for \N, which always matches any character other than
1978        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1980    6.  When compiling pcregrep with newer versions of gcc which may have
1981        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1982        declared with attribute warn_unused_result" were given. Just casting the
1983        result to (void) does not stop the warnings; a more elaborate fudge is
1984        needed. I've used a macro to implement this.
1986    7.  Minor change to pcretest.c to avoid a compiler warning.
1988    8.  Added four artifical Unicode properties to help with an option to make
1989        \s etc use properties (see next item). The new properties are: Xan
1990        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1992    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1993        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1994        this option. Modified pcretest to add /W to test this facility. Added
1995        REG_UCP to make it available via the POSIX interface.
1997    10. Added --line-buffered to pcregrep.
1999    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
2000        studied, and the match started with a letter with a code point greater than
2001        127 whose first byte was different to the first byte of the other case of
2002        the letter, the other case of this starting letter was not recognized
2003        (#976).
2005    12. If a pattern that was studied started with a repeated Unicode property
2006        test, for example, \p{Nd}+, there was the theoretical possibility of
2007        setting up an incorrect bitmap of starting bytes, but fortunately it could
2008        not have actually happened in practice until change 8 above was made (it
2009        added property types that matched character-matching opcodes).
2011    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
2012        possible starting bytes for non-anchored patterns.
2014    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
2015        \R, and also a number of cases that involve Unicode properties, both
2016        explicit and implicit when PCRE_UCP is set.
2018    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
2019        input, it could crash or give wrong results if characters with values
2020        greater than 0xc0 were present in the subject string. (Detail: it assumed
2021        UTF-8 input when processing these items.)
2023    16. Added a lot of (int) casts to avoid compiler warnings in systems where
2024        size_t is 64-bit (#991).
2026    17. Added a check for running out of memory when PCRE is compiled with
2027        --disable-stack-for-recursion (#990).
2029    18. If the last data line in a file for pcretest does not have a newline on
2030        the end, a newline was missing in the output.
2032    19. The default pcre_chartables.c file recognizes only ASCII characters (values
2033        less than 128) in its various bitmaps. However, there is a facility for
2034        generating tables according to the current locale when PCRE is compiled. It
2035        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
2036        characters, are recognized by isspace() and therefore were getting set in
2037        these tables, and indeed these tables seem to approximate to ISO 8859. This
2038        caused a problem in UTF-8 mode when pcre_study() was used to create a list
2039        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
2040        which of course cannot start UTF-8 characters. I have changed the code so
2041        that only real ASCII characters (less than 128) and the correct starting
2042        bytes for UTF-8 encodings are set for characters greater than 127 when in
2043        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
2044        altogether.)
2046    20. Added the /T option to pcretest so as to be able to run tests with non-
2047        standard character tables, thus making it possible to include the tests
2048        used for 19 above in the standard set of tests.
2050    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
2051        reference to a subpattern the other side of a comment that contains an
2052        opening parenthesis caused either an internal compiling error, or a
2053        reference to the wrong subpattern.
2056    Version 8.02 19-Mar-2010
2057    ------------------------
2059    1.  The Unicode data tables have been updated to Unicode 5.2.0.
2061    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
2062        configured.
2064    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
2065        original author of that file, following a query about its status.
2067    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
2068        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
2070    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
2071        quantifier applied to a forward-referencing subroutine call, could compile
2072        incorrect code or give the error "internal error: previously-checked
2073        referenced subpattern not found".
2075    6.  Both MS Visual Studio and Symbian OS have problems with initializing
2076        variables to point to external functions. For these systems, therefore,
2077        pcre_malloc etc. are now initialized to local functions that call the
2078        relevant global functions.
2080    7.  There were two entries missing in the vectors called coptable and poptable
2081        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
2082        I've fixed the data, and added a kludgy way of testing at compile time that
2083        the lengths are correct (equal to the number of opcodes).
2085    8.  Following on from 7, I added a similar kludge to check the length of the
2086        eint vector in pcreposix.c.
2088    9.  Error texts for pcre_compile() are held as one long string to avoid too
2089        much relocation at load time. To find a text, the string is searched,
2090        counting zeros. There was no check for running off the end of the string,
2091        which could happen if a new error number was added without updating the
2092        string.
2094    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
2096    11. \K was not working if it appeared in an atomic group or in a group that
2097        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
2098        \K is "not well defined" if used in an assertion. PCRE now accepts it if
2099        the assertion is positive, but not if it is negative.
2101    12. Change 11 fortuitously reduced the size of the stack frame used in the
2102        "match()" function of pcre_exec.c by one pointer. Forthcoming
2103        implementation of support for (*MARK) will need an extra pointer on the
2104        stack; I have reserved it now, so that the stack frame size does not
2105        decrease.
2107    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
2108        item in branch that calls a recursion is a subroutine call - as in the
2109        second branch in the above example - was incorrectly given the compile-
2110        time error "recursive call could loop indefinitely" because pcre_compile()
2111        was not correctly checking the subroutine for matching a non-empty string.
2113    14. The checks for overrunning compiling workspace could trigger after an
2114        overrun had occurred. This is a "should never occur" error, but it can be
2115        triggered by pathological patterns such as hundreds of nested parentheses.
2116        The checks now trigger 100 bytes before the end of the workspace.
2118    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
2121    Version 8.01 19-Jan-2010
2122    ------------------------
2124    1.  If a pattern contained a conditional subpattern with only one branch (in
2125        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
2126        computed the wrong minimum data length (which is of course zero for such
2127        subpatterns). This could cause incorrect "no match" results.
2129    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
2130        the pattern is reset in the first branch, pcre_compile() failed with
2131        "internal error: code overflow at offset...". This happened only when
2132        the reset was to the original external option setting. (An optimization
2133        abstracts leading options settings into an external setting, which was the
2134        cause of this.)
2136    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
2137        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
2138        assertion pattern did not match (meaning that the assertion was true), it
2139        was incorrectly treated as false if the SKIP had been reached during the
2140        matching. This also applied to assertions used as conditions.
2142    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
2143        assertion subpattern, including such a pattern used as a condition,
2144        unpredictable results occurred, instead of the error return
2147    5.  The C++ GlobalReplace function was not working like Perl for the special
2148        situation when an empty string is matched. It now does the fancy magic
2149        stuff that is necessary.
2151    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
2152        removed. (These were left over from very, very early versions of PCRE.)
2154    7.  Some cosmetic changes to the code to make life easier when compiling it
2155        as part of something else:
2157        (a) Change DEBUG to PCRE_DEBUG.
2159        (b) In pcre_compile(), rename the member of the "branch_chain" structure
2160            called "current" as "current_branch", to prevent a collision with the
2161            Linux macro when compiled as a kernel module.
2163        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
2164            prevent a collision with the Linux macro when compiled as a kernel
2165            module.
2167    8.  In pcre_compile() there are some checks for integer overflows that used to
2168        cast potentially large values to (double). This has been changed to that
2169        when building, a check for int64_t is made, and if it is found, it is used
2170        instead, thus avoiding the use of floating point arithmetic. (There is no
2171        other use of FP in PCRE.) If int64_t is not found, the fallback is to
2172        double.
2174    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
2175        2005 (difference between two addresses compared to an unsigned value).
2177    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
2178        custom one, because of the following reported problem in Windows:
2180          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
2181              under Win32.
2182          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
2183              therefore missing the function definition.
2184          - The compiler thus generates a "C" signature for the test function.
2185          - The linker fails to find the "C" function.
2186          - PCRE fails to configure if asked to do so against libbz2.
2188    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
2189        messages were output:
2191          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
2192          rerunning libtoolize, to keep the correct libtool macros in-tree.
2193          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
2195        I have done both of these things.
2197    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
2198        most of the time, it *can* run out if it is given a pattern that contains a
2199        runaway infinite recursion. I updated the discussion in the pcrestack man
2200        page.
2202    13. Now that we have gone to the x.xx style of version numbers, the minor
2203        version may start with zero. Using 08 or 09 is a bad idea because users
2204        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
2205        interpreted as invalid octal numbers. I've updated the previous comment in
2206        configure.ac, and also added a check that gives an error if 08 or 09 are
2207        used.
2209    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
2210        causing partial matching to fail when the end of the subject matched \W
2211        in a UTF-8 pattern where \W was quantified with a minimum of 3.
2213    15. There were some discrepancies between the declarations in pcre_internal.h
2214        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
2215        their definitions. The declarations used "const uschar *" and the
2216        definitions used USPTR. Even though USPTR is normally defined as "const
2217        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
2218        reported that: "This difference in casting confuses some C++ compilers, for
2219        example, SunCC recognizes above declarations as different functions and
2220        generates broken code for hbpcre." I have changed the declarations to use
2221        USPTR.
2223    16. GNU libtool is named differently on some systems. The autogen.sh script now
2224        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
2225        (FreeBSD).
2227    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
2228        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
2229        comment: "Figure out how to create a longlong from a string: strtoll and
2230        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
2231        instance, but it only takes 2 args instead of 3!"
2233    18. A subtle bug concerned with back references has been fixed by a change of
2234        specification, with a corresponding code fix. A pattern such as
2235        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
2236        refers, was giving matches when it shouldn't. For example, xa=xaaa would
2237        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
2238        same bug. Such groups have to be quantified to be useful, or contained
2239        inside another quantified group. (If there's no repetition, the reference
2240        can never match.) The problem arises because, having left the group and
2241        moved on to the rest of the pattern, a later failure that backtracks into
2242        the group uses the captured value from the final iteration of the group
2243        rather than the correct earlier one. I have fixed this in PCRE by forcing
2244        any group that contains a reference to itself to be an atomic group; that
2245        is, there cannot be any backtracking into it once it has completed. This is
2246        similar to recursive and subroutine calls.
2249    Version 8.00 19-Oct-09
2250    ----------------------
2252    1.  The table for translating pcre_compile() error codes into POSIX error codes
2253        was out-of-date, and there was no check on the pcre_compile() error code
2254        being within the table. This could lead to an OK return being given in
2255        error.
2257    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
2258        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
2259        in a Windows environment.
2261    3.  The pcregrep --count option prints the count for each file even when it is
2262        zero, as does GNU grep. However, pcregrep was also printing all files when
2263        --files-with-matches was added. Now, when both options are given, it prints
2264        counts only for those files that have at least one match. (GNU grep just
2265        prints the file name in this circumstance, but including the count seems
2266        more useful - otherwise, why use --count?) Also ensured that the
2267        combination -clh just lists non-zero counts, with no names.
2269    4.  The long form of the pcregrep -F option was incorrectly implemented as
2270        --fixed_strings instead of --fixed-strings. This is an incompatible change,
2271        but it seems right to fix it, and I didn't think it was worth preserving
2272        the old behaviour.
2274    5.  The command line items --regex=pattern and --regexp=pattern were not
2275        recognized by pcregrep, which required --regex pattern or --regexp pattern
2276        (with a space rather than an '='). The man page documented the '=' forms,
2277        which are compatible with GNU grep; these now work.
2279    6.  No libpcreposix.pc file was created for pkg-config; there was just
2280        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
2282    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
2283        when UCP support is not needed, by modifying the Python script that
2284        generates it from Unicode data files. This should not matter if the module
2285        is correctly used as a library, but I received one complaint about 50K of
2286        unwanted data. My guess is that the person linked everything into his
2287        program rather than using a library. Anyway, it does no harm.
2289    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
2290        was a minimum greater than 1 for a wide character in a possessive
2291        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
2292        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
2293        character. Chaos in the form of incorrect output or a compiling loop could
2294        result.
2296    9.  The restrictions on what a pattern can contain when partial matching is
2297        requested for pcre_exec() have been removed. All patterns can now be
2298        partially matched by this function. In addition, if there are at least two
2299        slots in the offset vector, the offset of the earliest inspected character
2300        for the match and the offset of the end of the subject are set in them when
2301        PCRE_ERROR_PARTIAL is returned.
2303    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
2304        synonymous with PCRE_PARTIAL, for backwards compatibility, and
2305        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
2306        and may be more useful for multi-segment matching.
2308    11. Partial matching with pcre_exec() is now more intuitive. A partial match
2309        used to be given if ever the end of the subject was reached; now it is
2310        given only if matching could not proceed because another character was
2311        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
2312        string "Z", which now yields "no match" instead of "partial match". In the
2313        case of pcre_dfa_exec(), "no match" is given if every matching path for the
2314        final character ended with (*FAIL).
2316    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
2317        if the pattern had a "must contain" character that was already found in the
2318        earlier partial match, unless partial matching was again requested. For
2319        example, with the pattern /dog.(body)?/, the "must contain" character is
2320        "g". If the first part-match was for the string "dog", restarting with
2321        "sbody" failed. This bug has been fixed.
2323    13. The string returned by pcre_dfa_exec() after a partial match has been
2324        changed so that it starts at the first inspected character rather than the
2325        first character of the match. This makes a difference only if the pattern
2326        starts with a lookbehind assertion or \b or \B (\K is not supported by
2327        pcre_dfa_exec()). It's an incompatible change, but it makes the two
2328        matching functions compatible, and I think it's the right thing to do.
2330    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
2331        so that the demonstration program is easily available in environments where
2332        PCRE has not been installed from source.
2334    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
2335        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
2336        library.
2338    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
2339        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
2340        is not the first non-POSIX option to be added. Clearly some people find
2341        these options useful.
2343    17. If a caller to the POSIX matching function regexec() passes a non-zero
2344        value for nmatch with a NULL value for pmatch, the value of
2345        nmatch is forced to zero.
2347    18. RunGrepTest did not have a test for the availability of the -u option of
2348        the diff command, as RunTest does. It now checks in the same way as
2349        RunTest, and also checks for the -b option.
2351    19. If an odd number of negated classes containing just a single character
2352        interposed, within parentheses, between a forward reference to a named
2353        subpattern and the definition of the subpattern, compilation crashed with
2354        an internal error, complaining that it could not find the referenced
2355        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
2356        [The bug was that it was starting one character too far in when skipping
2357        over the character class, thus treating the ] as data rather than
2358        terminating the class. This meant it could skip too much.]
2360    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
2361        /g option in pcretest when the pattern contains \K, which makes it possible
2362        to have an empty string match not at the start, even when the pattern is
2363        anchored. Updated pcretest and pcredemo to use this option.
2365    21. If the maximum number of capturing subpatterns in a recursion was greater
2366        than the maximum at the outer level, the higher number was returned, but
2367        with unset values at the outer level. The correct (outer level) value is
2368        now given.
2370    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
2371        PCRE did not set those parentheses (unlike Perl). I have now found a way to
2372        make it do so. The string so far is captured, making this feature
2373        compatible with Perl.
2375    23. The tests have been re-organized, adding tests 11 and 12, to make it
2376        possible to check the Perl 5.10 features against Perl 5.10.
2378    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
2379        pattern matches a fixed length string. PCRE did not allow this; now it
2380        does. Neither allows recursion.
2382    25. I finally figured out how to implement a request to provide the minimum
2383        length of subject string that was needed in order to match a given pattern.
2384        (It was back references and recursion that I had previously got hung up
2385        on.) This code has now been added to pcre_study(); it finds a lower bound
2386        to the length of subject needed. It is not necessarily the greatest lower
2387        bound, but using it to avoid searching strings that are too short does give
2388        some useful speed-ups. The value is available to calling programs via
2389        pcre_fullinfo().
2391    26. While implementing 25, I discovered to my embarrassment that pcretest had
2392        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
2393        study optimizations had never been tested with that matching function.
2394        Oops. What is worse, even when it was passed study data, there was a bug in
2395        pcre_dfa_exec() that meant it never actually used it. Double oops. There
2396        were also very few tests of studied patterns with pcre_dfa_exec().
2398    27. If (?| is used to create subpatterns with duplicate numbers, they are now
2399        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
2400        on the other side of the coin, they are no longer allowed to have different
2401        names, because these cannot be distinguished in PCRE, and this has caused
2402        confusion. (This is a difference from Perl.)
2404    28. When duplicate subpattern names are present (necessarily with different
2405        numbers, as required by 27 above), and a test is made by name in a
2406        conditional pattern, either for a subpattern having been matched, or for
2407        recursion in such a pattern, all the associated numbered subpatterns are
2408        tested, and the overall condition is true if the condition is true for any
2409        one of them. This is the way Perl works, and is also more like the way
2410        testing by number works.
2413    Version 7.9 11-Apr-09
2414    ---------------------
2416    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
2417        (pcretest), all targets were linked against these libraries. This included
2418        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
2419        libraries. This caused unwanted dependencies to be created. This problem
2420        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
2421        pcretest is linked with readline.
2423    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
2424        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
2425        moved outside it again, because FALSE and TRUE are already defined in AIX,
2426        but BOOL is not.
2428    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
2429        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
2431    4.  The pcregrep documentation said spaces were inserted as well as colons (or
2432        hyphens) following file names and line numbers when outputting matching
2433        lines. This is not true; no spaces are inserted. I have also clarified the
2434        wording for the --colour (or --color) option.
2436    5.  In pcregrep, when --colour was used with -o, the list of matching strings
2437        was not coloured; this is different to GNU grep, so I have changed it to be
2438        the same.
2440    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
2441        each matching line was coloured. Now it goes on to look for further matches
2442        of any of the test patterns, which is the same behaviour as GNU grep.
2444    7.  A pattern that could match an empty string could cause pcregrep to loop; it
2445        doesn't make sense to accept an empty string match in pcregrep, so I have
2446        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2447        seems to be how GNU grep behaves. [But see later change 40 for release
2448        8.33.]
2450    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2451        start or after a newline", because the conditional assertion was not being
2452        correctly handled. The rule now is that both the assertion and what follows
2453        in the first alternative must satisfy the test.
2455    9.  If auto-callout was enabled in a pattern with a conditional group whose
2456        condition was an assertion, PCRE could crash during matching, both with
2457        pcre_exec() and pcre_dfa_exec().
2459    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2460        used for matching.
2462    11. Unicode property support in character classes was not working for
2463        characters (bytes) greater than 127 when not in UTF-8 mode.
2465    12. Added the -M command line option to pcretest.
2467    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2469    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2471    16. Added comments and documentation about mis-use of no_arg in the C++
2472        wrapper.
2474    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2475        from Martin Jerabek that uses macro names for all relevant character and
2476        string constants.
2478    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2479        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2480        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2481        these, but not everybody uses configure.
2483    19. A conditional group that had only one branch was not being correctly
2484        recognized as an item that could match an empty string. This meant that an
2485        enclosing group might also not be so recognized, causing infinite looping
2486        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2487        with the subject "ab", where knowledge that the repeated group can match
2488        nothing is needed in order to break the loop.
2490    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2491        exec(), but without supplying a callout function, matching went wrong.
2493    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2494        leak if the size of the offset vector was greater than 30. When the vector
2495        is smaller, the saved offsets during recursion go onto a local stack
2496        vector, but for larger vectors malloc() is used. It was failing to free
2497        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2498        error, in fact).
2500    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2501        heapframe that is used only when UTF-8 support is enabled. This caused no
2502        problem, but was untidy.
2504    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2505        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2506        included within another project.
2508    24. Steven Van Ingelgem's patches to add more options to the CMake support,
2509        slightly modified by me:
2511          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2512              not building pcregrep.
2514          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2515              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2517    25. Forward references, both numeric and by name, in patterns that made use of
2518        duplicate group numbers, could behave incorrectly or give incorrect errors,
2519        because when scanning forward to find the reference group, PCRE was not
2520        taking into account the duplicate group numbers. A pattern such as
2521        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2523    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2524        the feature of a custom pointer more persuasive (as requested by a user).
2526    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2527        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2528        pre-defined.
2530    28. Added support for (*UTF8) at the start of a pattern.
2532    29. Arrange for flags added by the "release type" setting in CMake to be shown
2533        in the configuration summary.
2536    Version 7.8 05-Sep-08
2537    ---------------------
2539    1.  Replaced UCP searching code with optimized version as implemented for Ad
2540        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2541        stage table and inline lookup instead of a function, giving speed ups of 2
2542        to 5 times on some simple patterns that I tested. Permission was given to
2543        distribute the MultiStage2.py script that generates the tables (it's not in
2544        the tarball, but is in the Subversion repository).
2546    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2547        scripts.
2549    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2550        a group with a zero qualifier. The result of the study could be incorrect,
2551        or the function might crash, depending on the pattern.
2553    4.  Caseless matching was not working for non-ASCII characters in back
2554        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2555        It now works when Unicode Property Support is available.
2557    5.  In pcretest, an escape such as \x{de} in the data was always generating
2558        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2559        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2560        truncation.
2562    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2564    7.  Added two (int) casts to pcregrep when printing the difference of two
2565        pointers, in case they are 64-bit values.
2567    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2568        test 2 if it fails.
2570    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2571        and a #define of that name to empty if it is not externally set. This is to
2572        allow users of MSVC to set it if necessary.
2574    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2575        the convenience functions in the pcre_get.c source file.
2577    11. An option change at the start of a pattern that had top-level alternatives
2578        could cause overwriting and/or a crash. This command provoked a crash in
2579        some environments:
2581          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2583        This potential security problem was recorded as CVE-2008-2371.
2585    12. For a pattern where the match had to start at the beginning or immediately
2586        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2587        pcre_dfa_exec() could read past the end of the passed subject if there was
2588        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2589        pcretest so that it places the subject at the end of its malloc-ed buffer.
2591    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2592        exec() might read past the end of the data buffer in UTF-8 mode.
2594    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2595        the data contained the byte 0x85 as part of a UTF-8 character within its
2596        first line. This applied both to normal and DFA matching.
2598    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2599        /^[^d]*?$/8 failed to match "abc".
2601    16. Added a missing copyright notice to pcrecpp_internal.h.
2603    17. Make it more clear in the documentation that values returned from
2604        pcre_exec() in ovector are byte offsets, not character counts.
2606    18. Tidied a few places to stop certain compilers from issuing warnings.
2608    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2609        supplied by Stefan Weber. I made a further small update for 7.8 because
2610        there is a change of source arrangements: the pcre_searchfuncs.c module is
2611        replaced by pcre_ucd.c.
2614    Version 7.7 07-May-08
2615    ---------------------
2617    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2618        a string to a long long, pretend we don't even have a long long." This is
2619        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2621    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2622        pre-7.6 versions, which defined a global no_arg variable instead of putting
2623        it in the RE class. (See also #8 below.)
2625    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2626        Lopes.
2628    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2630        (1) The include/exclude patterns were being applied to the whole pathnames
2631            of files, instead of just to the final components.
2633        (2) If there was more than one level of directory, the subdirectories were
2634            skipped unless they satisfied the include/exclude conditions. This is
2635            inconsistent with GNU grep (and could even be seen as contrary to the
2636            pcregrep specification - which I improved to make it absolutely clear).
2637            The action now is always to scan all levels of directory, and just
2638            apply the include/exclude patterns to regular files.
2640    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2641        --exclude_dir in the tests to avoid scanning .svn directories.
2643    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2644        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2645        doesn't support NULs in patterns.
2647    7.  Added some missing "const"s to declarations of static tables in
2648        pcre_compile.c and pcre_dfa_exec.c.
2650    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2651        caused by fix #2  above. (Subsequently also a second patch to fix the
2652        first patch. And a third patch - this was a messy problem.)
2654    9.  Applied Craig's patch to remove the use of push_back().
2656    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2657        matching function regexec().
2659    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2660        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2661        references. PCRE supports relative numbers with this syntax (I don't think
2662        Oniguruma does).
2664    12. Previously, a group with a zero repeat such as (...){0} was completely
2665        omitted from the compiled regex. However, this means that if the group
2666        was called as a subroutine from elsewhere in the pattern, things went wrong
2667        (an internal error was given). Such groups are now left in the compiled
2668        pattern, with a new opcode that causes them to be skipped at execution
2669        time.
2671    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2672        to the way PCRE behaves:
2674        (a) A lone ] character is dis-allowed (Perl treats it as data).
2676        (b) A back reference to an unmatched subpattern matches an empty string
2677            (Perl fails the current match path).
2679        (c) A data ] in a character class must be notated as \] because if the
2680            first data character in a class is ], it defines an empty class. (In
2681            Perl it is not possible to have an empty class.) The empty class []
2682            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2683            The negative empty class [^] matches any one character, independently
2684            of the DOTALL setting.
2686    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2687        non-existent subpattern following a character class starting with ']' and
2688        containing () gave an internal compiling error instead of "reference to
2689        non-existent subpattern". Fortunately, when the pattern did exist, the
2690        compiled code was correct. (When scanning forwards to check for the
2691        existence of the subpattern, it was treating the data ']' as terminating
2692        the class, so got the count wrong. When actually compiling, the reference
2693        was subsequently set up correctly.)
2695    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2696        it was being rejected as not supported by pcre_dfa_exec(), even though
2697        other assertions are supported. I have made pcre_dfa_exec() support
2698        (*FAIL).
2700    16. The implementation of 13c above involved the invention of a new opcode,
2701        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2702        cannot be changed at match time, I realized I could make a small
2703        improvement to matching performance by compiling OP_ALLANY instead of
2704        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2705        on the OP_ANY path.
2707    17. Compiling pcretest on Windows with readline support failed without the
2708        following two fixes: (1) Make the unistd.h include conditional on
2709        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2711    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2712        ncurses library to be included for pcretest when ReadLine support is
2713        requested, but also to allow for it to be overridden. This patch came from
2714        Daniel Bergström.
2716    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2717        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2718        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2719        spotting this.
2722    Version 7.6 28-Jan-08
2723    ---------------------
2725    1.  A character class containing a very large number of characters with
2726        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2727        overflow.
2729    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2730        HAVE_LONG_LONG is not defined.
2732    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2733        bring them up to date and include new features. This patch includes:
2735        - Fixed PH's badly added libz and libbz2 support.
2736        - Fixed a problem with static linking.
2737        - Added pcredemo. [But later removed - see 7 below.]
2738        - Fixed dftables problem and added an option.
2739        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2740            HAVE_LONG_LONG.
2741        - Added readline support for pcretest.
2742        - Added an listing of the option settings after cmake has run.
2744    4.  A user submitted a patch to Makefile that makes it easy to create
2745        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2746        Makefile.am that cause it to include this special target, without
2747        affecting anything else. Note that the same mingw target plus all
2748        the other distribution libraries and programs are now supported
2749        when configuring with CMake (see 6 below) instead of with
2750        Configure/Make.
2752    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2753        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2754        exported in the Windows port". It has not yet been confirmed that the patch
2755        solves the problem, but it does no harm.
2757    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2758        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2759        with CMake, and also correct the comment about stack recursion.
2761    7.  Remove the automatic building of pcredemo from the ./configure system and
2762        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2763        of a program that users should build themselves after PCRE is installed, so
2764        building it automatically is not really right. What is more, it gave
2765        trouble in some build environments.
2767    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2770    Version 7.5 10-Jan-08
2771    ---------------------
2773    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2774        values in parens when parsing an RE using the C++ wrapper."
2776    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2777        Characters greater than 255 were excluded from the class instead of being
2778        included.
2780    3.  The same bug as (2) above applied to negated POSIX classes such as
2781        [:^space:].
2783    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2784        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2785        I have changed it.
2787    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2788        first named subpattern) and a construct such as (?&a) would reference the
2789        first named subpattern whose name started with "a" (in other words, the
2790        length check was missing). Both these problems are fixed. "Subpattern name
2791        expected" is now given for (?&) (a zero-length name), and this patch also
2792        makes it give the same error for \k'' (previously it complained that that
2793        was a reference to a non-existent subpattern).
2795    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2796        this is right because (?- can be followed by option settings as well as by
2797        digits. I have, however, made the messages clearer.
2799    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2800        than the number used in the conditional) now cause a compile-time error.
2801        This is actually not compatible with Perl, which accepts such patterns, but
2802        treats the conditional as always being FALSE (as PCRE used to), but it
2803        seems to me that giving a diagnostic is better.
2805    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2806        and messages.
2808    9.  Fix two occurrences of "backslash" in comments that should have been
2809        "backspace".
2811    10. Remove two redundant lines of code that can never be obeyed (their function
2812        was moved elsewhere).
2814    11. The program that makes PCRE's Unicode character property table had a bug
2815        which caused it to generate incorrect table entries for sequences of
2816        characters that have the same character type, but are in different scripts.
2817        It amalgamated them into a single range, with the script of the first of
2818        them. In other words, some characters were in the wrong script. There were
2819        thirteen such cases, affecting characters in the following ranges:
2821          U+002b0 - U+002c1
2822          U+0060c - U+0060d
2823          U+0061e - U+00612
2824          U+0064b - U+0065e
2825          U+0074d - U+0076d
2826          U+01800 - U+01805
2827          U+01d00 - U+01d77
2828          U+01d9b - U+01dbf
2829          U+0200b - U+0200f
2830          U+030fc - U+030fe
2831          U+03260 - U+0327f
2832          U+0fb46 - U+0fbb1
2833          U+10450 - U+1049d
2835    12. The -o option (show only the matching part of a line) for pcregrep was not
2836        compatible with GNU grep in that, if there was more than one match in a
2837        line, it showed only the first of them. It now behaves in the same way as
2838        GNU grep.
2840    13. If the -o and -v options were combined for pcregrep, it printed a blank
2841        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2842        does the same. The return code can be used to tell if there were any
2843        non-matching lines.
2845    14. Added --file-offsets and --line-offsets to pcregrep.
2847    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2848        infinitely looping recursion. The bug was that positive lookaheads were not
2849        being skipped when checking for a possible empty match (negative lookaheads
2850        and both kinds of lookbehind were skipped).
2852    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2853        inclusion of <windows.h> to before rather than after the definition of
2854        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2856    17. Specifying a possessive quantifier with a specific limit for a Unicode
2857        character property caused pcre_compile() to compile bad code, which led at
2858        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2859        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2860        caused the error; without that there was no problem.
2862    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2864    19. Added --enable-pcretest-libreadline.
2866    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2867        RE::GlobalReplace(). As a result, the number of replacements returned was
2868        double what it should be. I removed one of the increments, but Craig sent a
2869        later patch that removed the other one (the right fix) and added unit tests
2870        that check the return values (which was not done before).
2872    21. Several CMake things:
2874        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2875            the names libpcre and libpcreposix, not just pcre and pcreposix.
2877        (2) The above change means that pcretest and pcregrep are now correctly
2878            linked with the newly-built libraries, not previously installed ones.
2882    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2883        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2884        UTF-8 newline character). The key issue is that the pattern starts .*;
2885        this means that the match must be either at the beginning, or after a
2886        newline. The bug was in the code for advancing after a failed match and
2887        checking that the new position followed a newline. It was not taking
2888        account of UTF-8 characters correctly.
2890    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2891        character classes. PCRE was not treating the sequence [:...:] as a
2892        character class unless the ... were all letters. Perl, however, seems to
2893        allow any characters between [: and :], though of course it rejects as
2894        unknown any "names" that contain non-letters, because all the known class
2895        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2896        for example, whereas PCRE did not - it did not recognize a POSIX character
2897        class. This seemed a bit dangerous, so the code has been changed to be
2898        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2899        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2900        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2901        Perl does, and where it didn't before.
2903    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2904        Windows environments %n is disabled by default.
2907    Version 7.4 21-Sep-07
2908    ---------------------
2910    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2911        means that a class such as [\s] counted as "explicit reference to CR or
2912        LF". That isn't really right - the whole point of the change was to try to
2913        help when there was an actual mention of one of the two characters. So now
2914        the change happens only if \r or \n (or a literal CR or LF) character is
2915        encountered.
2917    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2918        of both had grown to the point where there were only 3 bits left.
2919        Fortunately, there was spare space in the data structure, and so I have
2920        moved the internal flags into a new 16-bit field to free up more option
2921        bits.
2923    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2924        but did not set the internal JCHANGED flag - either of these is enough to
2925        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2926        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2927        start sets both bits.
2929    4.  Added options (at build time, compile time, exec time) to change \R from
2930        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2932    5.  doc/pcresyntax.html was missing from the distribution.
2934    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2935        compatibility, even though it is no longer used.
2937    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2938        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2939        windows.h file is present (where different names are used). [This was
2940        reversed later after testing - see 16 below.]
2942    8.  Changed all #include <config.h> to #include "config.h". There were also
2943        some further <pcre.h> cases that I changed to "pcre.h".
2945    9.  When pcregrep was used with the --colour option, it missed the line ending
2946        sequence off the lines that it output.
2948    10. It was pointed out to me that arrays of string pointers cause lots of
2949        relocations when a shared library is dynamically loaded. A technique of
2950        using a single long string with a table of offsets can drastically reduce
2951        these. I have refactored PCRE in four places to do this. The result is
2952        dramatic:
2954          Originally:                          290
2955          After changing UCP table:            187
2956          After changing error message table:   43
2957          After changing table of "verbs"       36
2958          After changing table of Posix names   22
2960        Thanks to the folks working on Gregex for glib for this insight.
2962    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2963        unicode-properties was also set.
2965    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2967    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2968        checked only for CRLF.
2970    14. Added casts to pcretest.c to avoid compiler warnings.
2972    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2974    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2975        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2976        entirely. This removes changes made in 7 above.
2978    17. The CMake files have been updated, and there is now more information about
2979        building with CMake in the NON-UNIX-USE document.
2982    Version 7.3 28-Aug-07
2983    ---------------------
2985     1. In the rejigging of the build system that eventually resulted in 7.1, the
2986        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2987        brackets there is not right, since it causes compilers to look for an
2988        installed pcre.h, not the version that is in the source that is being
2989        compiled (which of course may be different). I have changed it back to:
2991          #include "pcre.h"
2993        I have a vague recollection that the change was concerned with compiling in
2994        different directories, but in the new build system, that is taken care of
2995        by the VPATH setting the Makefile.
2997     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2998        when the subject happened to end in the byte 0x85 (e.g. if the last
2999        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
3000        characters but of course it shouldn't be taken as a newline when it is part
3001        of another character. The bug was that, for an unlimited repeat of . in
3002        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
3003        characters when looking for a newline.
3005     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
3007     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
3008        in debug output.
3010     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
3011        long printing in the pcrecpp unittest when running under MinGW.
3013     6. ESC_K was left out of the EBCDIC table.
3015     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
3016        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
3017        limit also applies to "virtual nesting" when a pattern is recursive, and in
3018        this case 1000 isn't so big. I have been able to remove this limit at the
3019        expense of backing off one optimization in certain circumstances. Normally,
3020        when pcre_exec() would call its internal match() function recursively and
3021        immediately return the result unconditionally, it uses a "tail recursion"
3022        feature to save stack. However, when a subpattern that can match an empty
3023        string has an unlimited repetition quantifier, it no longer makes this
3024        optimization. That gives it a stack frame in which to save the data for
3025        checking that an empty string has been matched. Previously this was taken
3026        from the 1000-entry workspace that had been reserved. So now there is no
3027        explicit limit, but more stack is used.
3029     8. Applied Daniel's patches to solve problems with the import/export magic
3030        syntax that is required for Windows, and which was going wrong for the
3031        pcreposix and pcrecpp parts of the library. These were overlooked when this
3032        problem was solved for the main library.
3034     9. There were some crude static tests to avoid integer overflow when computing
3035        the size of patterns that contain repeated groups with explicit upper
3036        limits. As the maximum quantifier is 65535, the maximum group length was
3037        set at 30,000 so that the product of these two numbers did not overflow a
3038        32-bit integer. However, it turns out that people want to use groups that
3039        are longer than 30,000 bytes (though not repeat them that many times).
3040        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
3041        made it possible to implement the integer overflow checks in a much more
3042        dynamic way, which I have now done. The artificial limitation on group
3043        length has been removed - we now have only the limit on the total length of
3044        the compiled pattern, which depends on the LINK_SIZE setting.
3046    10. Fixed a bug in the documentation for get/copy named substring when
3047        duplicate names are permitted. If none of the named substrings are set, the
3048        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
3049        empty string.
3051    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
3052        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
3053        because the ] is interpreted as the first data character and the
3054        terminating ] is not found. PCRE has been made compatible with Perl in this
3055        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
3056        cause memory overwriting.
3058    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
3059        string has been matched (to stop an infinite loop). It was not recognizing
3060        a conditional subpattern that could match an empty string if that
3061        subpattern was within another subpattern. For example, it looped when
3062        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
3063        condition was not nested. This bug has been fixed.
3065    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
3066        past the start of the subject in the presence of bytes with the top bit
3067        set, for example "\x8aBCD".
3069    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
3070        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
3072    14. Optimized (?!) to (*FAIL).
3074    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
3075        This restricts code points to be within the range 0 to 0x10FFFF, excluding
3076        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
3077        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
3078        does: it's just the validity check that is more restrictive.
3080    16. Inserted checks for integer overflows during escape sequence (backslash)
3081        processing, and also fixed erroneous offset values for syntax errors during
3082        backslash processing.
3084    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
3085        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
3087    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
3088        caused an overrun.
3090    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
3091        something other than just ASCII characters) inside a group that had an
3092        unlimited repeat caused a loop at compile time (while checking to see
3093        whether the group could match an empty string).
3095    20. Debugging a pattern containing \p or \P could cause a crash. For example,
3096        [\P{Any}] did so. (Error in the code for printing property names.)
3098    21. An orphan \E inside a character class could cause a crash.
3100    22. A repeated capturing bracket such as (A)? could cause a wild memory
3101        reference during compilation.
3103    23. There are several functions in pcre_compile() that scan along a compiled
3104        expression for various reasons (e.g. to see if it's fixed length for look
3105        behind). There were bugs in these functions when a repeated \p or \P was
3106        present in the pattern. These operators have additional parameters compared
3107        with \d, etc, and these were not being taken into account when moving along
3108        the compiled data. Specifically:
3110        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
3111            length.
3113        (b) An item such as \pL+ within a repeated group could cause crashes or
3114            loops.
3116        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
3117            "reference to non-existent subpattern" error.
3119        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
3121    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
3122        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
3124    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
3126    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
3127        character were causing crashes (broken optimization).
3129    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
3130        \p or \P) caused a compile-time loop.
3132    28. More problems have arisen in unanchored patterns when CRLF is a valid line
3133        break. For example, the unstudied pattern [\r\n]A does not match the string
3134        "\r\nA" because change 7.0/46 below moves the current point on by two
3135        characters after failing to match at the start. However, the pattern \nA
3136        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
3137        the same is true. There doesn't seem any very clean way out of this, but
3138        what I have chosen to do makes the common cases work: PCRE now takes note
3139        of whether there can be an explicit match for \r or \n anywhere in the
3140        pattern, and if so, 7.0/46 no longer applies. As part of this change,
3141        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
3142        pattern has explicit CR or LF references.
3144    29. Added (*CR) etc for changing newline setting at start of pattern.
3147    Version 7.2 19-Jun-07
3148  ---------------------  ---------------------
3150   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
# Line 21  Version 7.2 01-May-07 Line 3164  Version 7.2 01-May-07
3164      stack recursion. This gives a massive performance boost under BSD, but just      stack recursion. This gives a massive performance boost under BSD, but just
3165      a small improvement under Linux. However, it saves one field in the frame      a small improvement under Linux. However, it saves one field in the frame
3166      in all cases.      in all cases.
3168   6. Added more features from the forthcoming Perl 5.10:   6. Added more features from the forthcoming Perl 5.10:
3170      (a) (?-n) (where n is a string of digits) is a relative subroutine or      (a) (?-n) (where n is a string of digits) is a relative subroutine or
3171          recursion call. It refers to the nth most recently opened parentheses.          recursion call. It refers to the nth most recently opened parentheses.
3173      (b) (?+n) is also a relative subroutine call; it refers to the nth next      (b) (?+n) is also a relative subroutine call; it refers to the nth next
3174          to be opened parentheses.          to be opened parentheses.
3176      (c) Conditions that refer to capturing parentheses can be specified      (c) Conditions that refer to capturing parentheses can be specified
3177          relatively, for example, (?(-2)... or (?(+3)...          relatively, for example, (?(-2)... or (?(+3)...
3179        (d) \K resets the start of the current match so that everything before
3180            is not part of it.
3182        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
3184        (f) \g{name} is another synonym - part of Perl 5.10's unification of
3185            reference syntax.
3187        (g) (?| introduces a group in which the numbering of parentheses in each
3188            alternative starts with the same number.
3190        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
3192     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
3195     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
3196        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
3197        for detecting groups that can match an empty string.
3199     9. A pattern with a very large number of alternatives (more than several
3200        hundred) was running out of internal workspace during the pre-compile
3201        phase, where pcre_compile() figures out how much memory will be needed. A
3202        bit of new cunning has reduced the workspace needed for groups with
3203        alternatives. The 1000-alternative test pattern now uses 12 bytes of
3204        workspace instead of running out of the 4096 that are available.
3206    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
3208    11. Applied patch from Google to remove an optimization that didn't quite work.
3209        The report of the bug said:
3211          pcrecpp::RE("a*").FullMatch("aaa") matches, while
3212          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
3213          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
3215    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
3216        it matched the wrong number of bytes.
3219  Version 7.1 24-Apr-07  Version 7.1 24-Apr-07
3220  ---------------------  ---------------------

Removed from v.167  
changed lines
  Added in v.1609

  ViewVC Help
Powered by ViewVC 1.1.5