/[pcre]/code/trunk/ChangeLog
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 1543 by ph10, Tue Apr 7 15:52:11 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.2 01-May-07  Version 8.37 xx-xxx-2015
5    ------------------------
6    
7    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
8        for those parentheses to be closed with whatever has been captured so far.
9        However, it was failing to mark any other groups between the hightest
10        capture so far and the currrent group as "unset". Thus, the ovector for
11        those groups contained whatever was previously there. An example is the
12        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
13    
14    2.  If an assertion condition was quantified with a minimum of zero (an odd
15        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
16    
17    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
18        unrecognized modifier, a crash could occur.
19    
20    4.  An attempt to do global matching in pcretest with a zero-length ovector
21        caused a crash.
22    
23    5.  Fixed a memory leak during matching that could occur for a subpattern
24        subroutine call (recursive or otherwise) if the number of captured groups
25        that had to be saved was greater than ten.
26    
27    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
28        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
29        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
30    
31    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
32        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
33    
34    8. When a pattern is compiled, it remembers the highest back reference so that
35       when matching, if the ovector is too small, extra memory can be obtained to
36       use instead. A conditional subpattern whose condition is a check on a
37       capture having happened, such as, for example in the pattern
38       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
39       setting the highest backreference number. This mattered only if pcre_exec()
40       was called with an ovector that was too small to hold the capture, and there
41       was no other kind of back reference (a situation which is probably quite
42       rare). The effect of the bug was that the condition was always treated as
43       FALSE when the capture could not be consulted, leading to a incorrect
44       behaviour by pcre2_match(). This bug has been fixed.
45    
46    9. A reference to a duplicated named group (either a back reference or a test
47       for being set in a conditional) that occurred in a part of the pattern where
48       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
49       to be incorrectly calculated, leading to overwriting.
50    
51    10. A mutually recursive set of back references such as (\2)(\1) caused a
52        segfault at study time (while trying to find the minimum matching length).
53        The infinite loop is now broken (with the minimum length unset, that is,
54        zero).
55    
56    11. If an assertion that was used as a condition was quantified with a minimum
57        of zero, matching went wrong. In particular, if the whole group had
58        unlimited repetition and could match an empty string, a segfault was
59        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
60        assertions to be quantified, but not if they are being used as conditions,
61        so the above pattern is faulted by Perl. PCRE has now been changed so that
62        it also rejects such patterns.
63    
64    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
65        failed to allow the zero-repeat case if pcre2_exec() was called with an
66        ovector too small to capture the group.
67    
68    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
69        Red Hat Product Security:
70    
71        (a) A crash if /K and /F were both set with the option to save the compiled
72        pattern.
73    
74        (b) Another crash if the option to print captured substrings in a callout
75        was combined with setting a null ovector, for example \O\C+ as a subject
76        string.
77    
78    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
79        forward reference repeated a large (but limited) number of times within a
80        repeated outer group that has a zero minimum quantifier, caused incorrect
81        code to be compiled, leading to the error "internal error:
82        previously-checked referenced subpattern not found" when an incorrect
83        memory address was read. This bug was reported as "heap overflow",
84        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
85        CVE-2015-2325.
86    
87    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
88        call within a group that also contained a recursive back reference caused
89        incorrect code to be compiled. This bug was reported as "heap overflow",
90        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
91        number CVE-2015-2326.
92    
93    24. Computing the size of the JIT read-only data in advance has been a source
94        of various issues, and new ones are still appear unfortunately. To fix
95        existing and future issues, size computation is eliminated from the code,
96        and replaced by on-demand memory allocation.
97    
98    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
99        adjacent to the end of the range, and the range contained characters with
100        more than one other case, caused incorrect behaviour when compiled in UTF
101        mode. In that example, the range a-j was left out of the class.
102    
103    26. Fix JIT compilation of conditional blocks, which assertion
104        is converted to (*FAIL). E.g: /(?(?!))/.
105    
106    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
107        discovered by the LLVM fuzzer.
108    
109    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
110        when this assertion was used as a condition, for example (?(?!)a|b). In
111        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
112        error about an unsupported item.
113    
114    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
115        possessification code could take exponential time to complete. A recursion
116        depth limit of 1000 has been imposed to limit the resources used by this
117        optimization.
118    
119    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
120        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
121        because \S ensures they are all in the class. The code for doing this was
122        interacting badly with the code for computing the amount of space needed to
123        compile the pattern, leading to a buffer overflow. This bug was discovered
124        by the LLVM fuzzer.
125    
126    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
127        other kinds of group caused stack overflow at compile time. This bug was
128        discovered by the LLVM fuzzer.
129    
130    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
131        between a subroutine call and its quantifier was incorrectly compiled,
132        leading to buffer overflow or other errors. This bug was discovered by the
133        LLVM fuzzer.
134    
135    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
136        assertion after (?(. The code was failing to check the character after
137        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
138        was discovered by the LLVM fuzzer.
139    
140    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
141        a fixed maximum following a group that contains a subroutine reference was
142        incorrectly compiled and could trigger buffer overflow. This bug was
143        discovered by the LLVM fuzzer.
144    
145    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
146        caused a stack overflow instead of the diagnosis of a non-fixed length
147        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
148    
149    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
150        (e.g. /(?<=\Ka)/) could make pcregrep loop.
151    
152    
153    Version 8.36 26-September-2014
154    ------------------------------
155    
156    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
157        -Wmissing-field-initializers and -Wunused-parameter.
158    
159    2.  The tests for quantifiers being too big (greater than 65535) were being
160        applied after reading the number, and stupidly assuming that integer
161        overflow would give a negative number. The tests are now applied as the
162        numbers are read.
163    
164    3.  Tidy code in pcre_exec.c where two branches that used to be different are
165        now the same.
166    
167    4.  The JIT compiler did not generate match limit checks for certain
168        bracketed expressions with quantifiers. This may lead to exponential
169        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
170        issue should be resolved now.
171    
172    5.  Fixed an issue, which occures when nested alternatives are optimized
173        with table jumps.
174    
175    6.  Inserted two casts and changed some ints to size_t in the light of some
176        reported 64-bit compiler warnings (Bugzilla 1477).
177    
178    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
179        an empty string, which sometimes were behaving incorrectly in the
180        interpreter (though correctly in the JIT matcher). This pcretest input is
181        an example:
182    
183          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
184          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
185    
186        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
187        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
188        for an empty string was breaking the inner loop and carrying on at a lower
189        level, when possessive repeated groups should always return to a higher
190        level as they have no backtrack points in them. The empty string test now
191        occurs at the outer level.
192    
193    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
194        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
195    
196    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
197        doesn't).
198    
199    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
200        to hold a pattern to be incorrectly computed (too small) when there were
201        named back references to duplicated names. This could cause "internal
202        error: code overflow" or "double free or corruption" or other memory
203        handling errors.
204    
205    11. When named subpatterns had the same prefixes, back references could be
206        confused. For example, in this pattern:
207    
208          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
209    
210        the reference to 'Name' was incorrectly treated as a reference to a
211        duplicate name.
212    
213    12. A pattern such as /^s?c/mi8 where the optional character has more than
214        one "other case" was incorrectly compiled such that it would only try to
215        match starting at "c".
216    
217    13. When a pattern starting with \s was studied, VT was not included in the
218        list of possible starting characters; this should have been part of the
219        8.34/18 patch.
220    
221    14. If a character class started [\Qx]... where x is any character, the class
222        was incorrectly terminated at the ].
223    
224    15. If a pattern that started with a caseless match for a character with more
225        than one "other case" was studied, PCRE did not set up the starting code
226        unit bit map for the list of possible characters. Now it does. This is an
227        optimization improvement, not a bug fix.
228    
229    16. The Unicode data tables have been updated to Unicode 7.0.0.
230    
231    17. Fixed a number of memory leaks in pcregrep.
232    
233    18. Avoid a compiler warning (from some compilers) for a function call with
234        a cast that removes "const" from an lvalue by using an intermediate
235        variable (to which the compiler does not object).
236    
237    19. Incorrect code was compiled if a group that contained an internal recursive
238        back reference was optional (had quantifier with a minimum of zero). This
239        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
240        caused segmentation faults because of stack overflows at compile time.
241    
242    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
243        group that is quantified with an indefinite repeat, caused a compile-time
244        loop which used up all the system stack and provoked a segmentation fault.
245        This was not the same bug as 19 above.
246    
247    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
248        Patch by Mike Frysinger.
249    
250    
251    Version 8.35 04-April-2014
252    --------------------------
253    
254    1.  A new flag is set, when property checks are present in an XCLASS.
255        When this flag is not set, PCRE can perform certain optimizations
256        such as studying these XCLASS-es.
257    
258    2.  The auto-possessification of character sets were improved: a normal
259        and an extended character set can be compared now. Furthermore
260        the JIT compiler optimizes more character set checks.
261    
262    3.  Got rid of some compiler warnings for potentially uninitialized variables
263        that show up only when compiled with -O2.
264    
265    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
266        of a match later then the end of the match. The pcretest program was not
267        handling the case sensibly - it was outputting from the start to the next
268        binary zero. It now reports this situation in a message, and outputs the
269        text from the end to the start.
270    
271    5.  Fast forward search is improved in JIT. Instead of the first three
272        characters, any three characters with fixed position can be searched.
273        Search order: first, last, middle.
274    
275    6.  Improve character range checks in JIT. Characters are read by an inprecise
276        function now, which returns with an unknown value if the character code is
277        above a certain threshold (e.g: 256). The only limitation is that the value
278        must be bigger than the threshold as well. This function is useful when
279        the characters above the threshold are handled in the same way.
280    
281    7.  The macros whose names start with RAWUCHAR are placeholders for a future
282        mode in which only the bottom 21 bits of 32-bit data items are used. To
283        make this more memorable for those maintaining the code, the names have
284        been changed to start with UCHAR21, and an extensive comment has been added
285        to their definition.
286    
287    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
288        to the export list in Makefile.am (they were accidentally omitted from the
289        8.34 tarball).
290    
291    9.  The informational output from pcretest used the phrase "starting byte set"
292        which is inappropriate for the 16-bit and 32-bit libraries. As the output
293        for "first char" and "need char" really means "non-UTF-char", I've changed
294        "byte" to "char", and slightly reworded the output. The documentation about
295        these values has also been (I hope) clarified.
296    
297    10. Another JIT related optimization: use table jumps for selecting the correct
298        backtracking path, when more than four alternatives are present inside a
299        bracket.
300    
301    11. Empty match is not possible, when the minimum length is greater than zero,
302        and there is no \K in the pattern. JIT should avoid empty match checks in
303        such cases.
304    
305    12. In a caseless character class with UCP support, when a character with more
306        than one alternative case was not the first character of a range, not all
307        the alternative cases were added to the class. For example, s and \x{17f}
308        are both alternative cases for S: the class [RST] was handled correctly,
309        but [R-T] was not.
310    
311    13. The configure.ac file always checked for pthread support when JIT was
312        enabled. This is not used in Windows, so I have put this test inside a
313        check for the presence of windows.h (which was already tested for).
314    
315    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
316        The algorithm provides a way to skip certain starting offsets, and usually
317        faster than linear prefix searches.
318    
319    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
320        as for 'fr_FR' and 'french'. For some reason, however, it then used the
321        Windows-specific input and output files, which have 'french' screwed in.
322        So this could never have worked. One of the problems with locales is that
323        they aren't always the same. I have now updated RunTest so that it checks
324        the output of the locale test (test 3) against three different output
325        files, and it allows the test to pass if any one of them matches. With luck
326        this should make the test pass on some versions of Solaris where it was
327        failing. Because of the uncertainty, the script did not used to stop if
328        test 3 failed; it now does. If further versions of a French locale ever
329        come to light, they can now easily be added.
330    
331    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
332        there was a message during ./configure, but it did not stop. This now
333        provokes an error. The invalid example in README has been corrected.
334        If a value less than the minimum is given, the minimum value has always
335        been used, but now a warning is given.
336    
337    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
338        was a bug in the test system, which is now fixed. Also, the list of various
339        configurations that are tested for each release did not have one with both
340        16/32 bits and --enable-bar-anycrlf. It now does.
341    
342    18. pcretest was missing "-C bsr" for displaying the \R default setting.
343    
344    19. Little endian PowerPC systems are supported now by the JIT compiler.
345    
346    20. The fast forward newline mechanism could enter to an infinite loop on
347        certain invalid UTF-8 input. Although we don't support these cases
348        this issue can be fixed by a performance optimization.
349    
350    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
351        not take account if existing stack usage. There is now a new global
352        variable called pcre_stack_guard that can be set to point to an external
353        function to check stack availability. It is called at the start of
354        processing every parenthesized group.
355    
356    22. A typo in the code meant that in ungreedy mode the max/min qualifier
357        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
358        not match "ab".
359    
360    23. When UTF was disabled, the JIT program reported some incorrect compile
361        errors. These messages are silenced now.
362    
363    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
364        compiler.
365    
366    25. Change all the temporary files used in RunGrepTest to be different to those
367        used by RunTest so that the tests can be run simultaneously, for example by
368        "make -j check".
369    
370    
371    Version 8.34 15-December-2013
372    -----------------------------
373    
374    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
375        executable memory. Patch inspired by Carsten Klein.
376    
377    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
378        this macro is never tested and has no effect, because the work to support
379        coverage involves only compiling and linking options and special targets in
380        the Makefile. The comment in config.h implied that defining the macro would
381        enable coverage support, which is totally false. There was also support for
382        setting this macro in the CMake files (my fault, I just copied it from
383        configure). SUPPORT_GCOV has now been removed.
384    
385    3.  Make a small performance improvement in strlen16() and strlen32() in
386        pcretest.
387    
388    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
389        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
390        these cases). There was also one in pcretest.c.
391    
392    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
393    
394    6.  In UTF mode, the code for checking whether a group could match an empty
395        string (which is used for indefinitely repeated groups to allow for
396        breaking an infinite loop) was broken when the group contained a repeated
397        negated single-character class with a character that occupied more than one
398        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
399        UTF-8 mode). The effect was undefined: the group might or might not be
400        deemed as matching an empty string, or the program might have crashed.
401    
402    7.  The code for checking whether a group could match an empty string was not
403        recognizing that \h, \H, \v, \V, and \R must match a character.
404    
405    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
406        an empty string. If it can, pcretest shows this in its information output.
407    
408    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
409        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
410        matched by pcre_exec() without using JIT:
411    
412        (a) If the rest of the pattern did not match after a maximal run of
413            grapheme clusters, the code for backing up to try with fewer of them
414            did not always back up over a full grapheme when characters that do not
415            have the modifier quality were involved, e.g. Hangul syllables.
416    
417        (b) If the match point in a subject started with modifier character, and
418            there was no match, the code could incorrectly back up beyond the match
419            point, and potentially beyond the first character in the subject,
420            leading to a segfault or an incorrect match result.
421    
422    10. A conditional group with an assertion condition could lead to PCRE
423        recording an incorrect first data item for a match if no other first data
424        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
425        first data item, and therefore matched "ca" after "c" instead of at the
426        start.
427    
428    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
429        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
430    
431    12. The source of pcregrep now includes z/OS-specific code so that it can be
432        compiled for z/OS as part of the special z/OS distribution.
433    
434    13. Added the -T and -TM options to pcretest.
435    
436    14. The code in pcre_compile.c for creating the table of named capturing groups
437        has been refactored. Instead of creating the table dynamically during the
438        actual compiling pass, the information is remembered during the pre-compile
439        pass (on the stack unless there are more than 20 named groups, in which
440        case malloc() is used) and the whole table is created before the actual
441        compile happens. This has simplified the code (it is now nearly 150 lines
442        shorter) and prepared the way for better handling of references to groups
443        with duplicate names.
444    
445    15. A back reference to a named subpattern when there is more than one of the
446        same name now checks them in the order in which they appear in the pattern.
447        The first one that is set is used for the reference. Previously only the
448        first one was inspected. This change makes PCRE more compatible with Perl.
449    
450    16. Unicode character properties were updated from Unicode 6.3.0.
451    
452    17. The compile-time code for auto-possessification has been refactored, based
453        on a patch by Zoltan Herczeg. It now happens after instead of during
454        compilation. The code is cleaner, and more cases are handled. The option
455        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
456        options in pcretest are provided to set it. It can also be set by
457        (*NO_AUTO_POSSESS) at the start of a pattern.
458    
459    18. The character VT has been added to the default ("C" locale) set of
460        characters that match \s and are generally treated as white space,
461        following this same change in Perl 5.18. There is now no difference between
462        "Perl space" and "POSIX space". Whether VT is treated as white space in
463        other locales depends on the locale.
464    
465    19. The code for checking named groups as conditions, either for being set or
466        for being recursed, has been refactored (this is related to 14 and 15
467        above). Processing unduplicated named groups should now be as fast at
468        numerical groups, and processing duplicated groups should be faster than
469        before.
470    
471    20. Two patches to the CMake build system, by Alexander Barkov:
472    
473          (1) Replace the "source" command by "." in CMakeLists.txt because
474              "source" is a bash-ism.
475    
476          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
477              without these the CMake build does not work on Solaris.
478    
479    21. Perl has changed its handling of \8 and \9. If there is no previously
480        encountered capturing group of those numbers, they are treated as the
481        literal characters 8 and 9 instead of a binary zero followed by the
482        literals. PCRE now does the same.
483    
484    22. Following Perl, added \o{} to specify codepoints in octal, making it
485        possible to specify values greater than 0777 and also making them
486        unambiguous.
487    
488    23. Perl now gives an error for missing closing braces after \x{... instead of
489        treating the string as literal. PCRE now does the same.
490    
491    24. RunTest used to grumble if an inappropriate test was selected explicitly,
492        but just skip it when running all tests. This make it awkward to run ranges
493        of tests when one of them was inappropriate. Now it just skips any
494        inappropriate tests, as it always did when running all tests.
495    
496    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
497        character types such as \d or \w, too many callouts were inserted, and the
498        data that they returned was rubbish.
499    
500    26. In UCP mode, \s was not matching two of the characters that Perl matches,
501        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
502        were matched by \h. The code has now been refactored so that the lists of
503        the horizontal and vertical whitespace characters used for \h and \v (which
504        are defined only in one place) are now also used for \s.
505    
506    27. Add JIT support for the 64 bit TileGX architecture.
507        Patch by Jiong Wang (Tilera Corporation).
508    
509    28. Possessive quantifiers for classes (both explicit and automatically
510        generated) now use special opcodes instead of wrapping in ONCE brackets.
511    
512    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
513        (because it's meaningless), this was not happening when PCRE_CASELESS was
514        set. Not wrong, but inefficient.
515    
516    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
517        properties for \w, \d, etc) is present in a test regex. Otherwise if the
518        test contains no characters greater than 255, Perl doesn't realise it
519        should be using Unicode semantics.
520    
521    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
522        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
523        does in Unicode mode.
524    
525    32. Added the "forbid" facility to pcretest so that putting tests into the
526        wrong test files can sometimes be quickly detected.
527    
528    33. There is now a limit (default 250) on the depth of nesting of parentheses.
529        This limit is imposed to control the amount of system stack used at compile
530        time. It can be changed at build time by --with-parens-nest-limit=xxx or
531        the equivalent in CMake.
532    
533    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
534        errors. Perl warns for these when in warning mode, but PCRE has no facility
535        for giving warnings.
536    
537    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
538        However, this was not working for (?!) because it is optimized to (*FAIL),
539        for which PCRE does not allow quantifiers. The optimization is now disabled
540        when a quantifier follows (?!). I can't see any use for this, but it makes
541        things uniform.
542    
543    36. Perl no longer allows group names to start with digits, so I have made this
544        change also in PCRE. It simplifies the code a bit.
545    
546    37. In extended mode, Perl ignores spaces before a + that indicates a
547        possessive quantifier. PCRE allowed a space before the quantifier, but not
548        before the possessive +. It now does.
549    
550    38. The use of \K (reset reported match start) within a repeated possessive
551        group such as (a\Kb)*+ was not working.
552    
553    40. Document that the same character tables must be used at compile time and
554        run time, and that the facility to pass tables to pcre_exec() and
555        pcre_dfa_exec() is for use only with saved/restored patterns.
556    
557    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
558        features for Builds with MSVC:
559    
560        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
561           builds). The .rc files can be used to set FileDescription and many other
562           attributes.
563    
564        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
565           This allows higher-level build scripts which want .pdb files to avoid
566           hard-coding the exact files needed."
567    
568    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
569        mean "start of word" and "end of word", respectively, as a transition aid.
570    
571    43. A minimizing repeat of a class containing codepoints greater than 255 in
572        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
573        compiled to use the heap for recursion.
574    
575    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
576        is configured.
577    
578    
579    Version 8.33 28-May-2013
580    ------------------------
581    
582    1.  Added 'U' to some constants that are compared to unsigned integers, to
583        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
584        variables that are added to signed variables, to ensure the result is
585        signed and can be negated.
586    
587    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
588        CMake config files.
589    
590    3.  Revise the creation of config.h.generic so that all boolean macros are
591        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
592        overriding via -D on the command line possible.
593    
594    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
595        to unsigned int is reported to make a quite noticeable speed difference in
596        a specific Windows environment. Testing on Linux did also appear to show
597        some benefit (and it is clearly not harmful). Also fixed the definition of
598        Xop which should be unsigned.
599    
600    5.  Related to (4), changing the definition of the intermediate variable cc
601        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
602        improvements.
603    
604    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
605        unnecessary spaces.
606    
607    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
608        and later.
609    
610    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
611    
612    9.  Optimizing fast_forward_start_bits in JIT.
613    
614    10. Adding support for callouts in JIT, and fixing some issues revealed
615        during this work. Namely:
616    
617        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
618    
619        (b) Minimum length was not checked before the matching is started.
620    
621    11. The value of capture_last that is passed to callouts was incorrect in some
622        cases when there was a capture on one path that was subsequently abandoned
623        after a backtrack. Also, the capture_last value is now reset after a
624        recursion, since all captures are also reset in this case.
625    
626    12. The interpreter no longer returns the "too many substrings" error in the
627        case when an overflowing capture is in a branch that is subsequently
628        abandoned after a backtrack.
629    
630    13. In the pathological case when an offset vector of size 2 is used, pcretest
631        now prints out the matched string after a yield of 0 or 1.
632    
633    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
634        Only supported by the JIT compiler at the moment.
635    
636    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
637    
638    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
639        offset of the starting point of the matching process, provided the offsets
640        vector is large enough.
641    
642    17. The \A escape now records a lookbehind value of 1, though its execution
643        does not actually inspect the previous character. This is to ensure that,
644        in partial multi-segment matching, at least one character from the old
645        segment is retained when a new segment is processed. Otherwise, if there
646        are no lookbehinds in the pattern, \A might match incorrectly at the start
647        of a new segment.
648    
649    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
650    
651    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
652        gives some modest performance improvement in 8-bit mode.
653    
654    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
655        be expressed in certain programming languages using Universal Character
656        Names.
657    
658    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
659        which points out that "non characters" are not "characters that may not
660        appear in Unicode strings" but rather "characters that are reserved for
661        internal use and have only local meaning".
662    
663    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
664        there was a conditional group that depended on an assertion, if the
665        assertion was false, the callout that immediately followed the alternation
666        in the condition was skipped when pcre_exec() was used for matching.
667    
668    23. Allow an explicit callout to be inserted before an assertion that is the
669        condition for a conditional group, for compatibility with automatic
670        callouts, which always insert a callout at this point.
671    
672    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
673        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
674    
675    25. (*PRUNE) is now supported by the JIT compiler.
676    
677    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
678    
679    27. Fix the case where there are two or more SKIPs with arguments that may be
680        ignored.
681    
682    28. (*SKIP) is now supported by the JIT compiler.
683    
684    29. (*THEN) is now supported by the JIT compiler.
685    
686    30. Update RunTest with additional test selector options.
687    
688    31. The way PCRE handles backtracking verbs has been changed in two ways.
689    
690        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
691        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
692        backtracking. In some cases this makes it more Perl-compatible, but Perl's
693        rather obscure rules do not always do the same thing.
694    
695        (2) Previously, backtracking verbs were confined within assertions. This is
696        no longer the case for positive assertions, except for (*ACCEPT). Again,
697        this sometimes improves Perl compatibility, and sometimes does not.
698    
699    32. A number of tests that were in test 2 because Perl did things differently
700        have been moved to test 1, because either Perl or PCRE has changed, and
701        these tests are now compatible.
702    
703    32. Backtracking control verbs are now handled in the same way in JIT and
704        interpreter.
705    
706    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
707        contained a forward subroutine reference caused a compile error.
708    
709    34. Auto-detect and optimize limited repetitions in JIT.
710    
711    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
712        blocking (*UTF) etc.
713    
714    36. In the interpreter, maximizing pattern repetitions for characters and
715        character types now use tail recursion, which reduces stack usage.
716    
717    37. The value of the max lookbehind was not correctly preserved if a compiled
718        and saved regex was reloaded on a host of different endianness.
719    
720    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
721        of the compiled pattern block, expand the flags field from 16 to 32 bits
722        because it was almost full.
723    
724    39. Try madvise first before posix_madvise.
725    
726    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
727        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
728        original change locked out all matches of empty strings. This has been
729        changed so that one match of an empty string per line is recognized.
730        Subsequent searches on the same line (for colouring or for --only-matching,
731        for example) do not recognize empty strings.
732    
733    41. Applied a user patch to fix a number of spelling mistakes in comments.
734    
735    42. Data lines longer than 65536 caused pcretest to crash.
736    
737    43. Clarified the data type for length and startoffset arguments for pcre_exec
738        and pcre_dfa_exec in the function-specific man pages, where they were
739        explicitly stated to be in bytes, never having been updated. I also added
740        some clarification to the pcreapi man page.
741    
742    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
743        a segmentation fault.
744    
745    
746    Version 8.32 30-November-2012
747    -----------------------------
748    
749    1.  Improved JIT compiler optimizations for first character search and single
750        character iterators.
751    
752    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
753        Patch by Daniel Richard G.
754    
755    3.  Single character iterator optimizations in the JIT compiler.
756    
757    4.  Improved JIT compiler optimizations for character ranges.
758    
759    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
760        Reported by Giuseppe D'Angelo.
761    
762    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
763        of a line, was being set incorrectly in cases where .* appeared inside
764        atomic brackets at the start of a pattern, or where there was a subsequent
765        *PRUNE or *SKIP.
766    
767    7.  Improved instruction cache flush for POWER/PowerPC.
768        Patch by Daniel Richard G.
769    
770    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
771        grep:
772    
773        (a) There is now no limit to the number of patterns to be matched.
774    
775        (b) An error is given if a pattern is too long.
776    
777        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
778            are now supported.
779    
780        (d) --exclude-from and --include-from (multiple use) have been added.
781    
782        (e) Exclusions and inclusions now apply to all files and directories, not
783            just to those obtained from scanning a directory recursively.
784    
785        (f) Multiple uses of -f and --file-list are now supported.
786    
787        (g) In a Windows environment, the default for -d has been changed from
788            "read" (the GNU grep default) to "skip", because otherwise the presence
789            of a directory in the file list provokes an error.
790    
791        (h) The documentation has been revised and clarified in places.
792    
793    9.  Improve the matching speed of capturing brackets.
794    
795    10. Changed the meaning of \X so that it now matches a Unicode extended
796        grapheme cluster.
797    
798    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
799        out POSIX threads when JIT support is configured.
800    
801    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
802    
803    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
804        structure could go wrong in environments where size_t is not the same size
805        as int.
806    
807    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
808        set.
809    
810    15. The EBCDIC support had decayed; later updates to the code had included
811        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
812        general tidy up of EBCDIC-related issues, and the documentation was also
813        not quite right. There is now a test that can be run on ASCII systems to
814        check some of the EBCDIC-related things (but is it not a full test).
815    
816    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
817        in a small tidy to the code.
818    
819    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
820    
821    18. If the --only-matching (-o) option in pcregrep is specified multiple
822        times, each one causes appropriate output. For example, -o1 -o2 outputs the
823        substrings matched by the 1st and 2nd capturing parentheses. A separating
824        string can be specified by --om-separator (default empty).
825    
826    19. Improving the first n character searches.
827    
828    20. Turn case lists for horizontal and vertical white space into macros so that
829        they are defined only once.
830    
831    21. This set of changes together give more compatible Unicode case-folding
832        behaviour for characters that have more than one other case when UCP
833        support is available.
834    
835        (a) The Unicode property table now has offsets into a new table of sets of
836            three or more characters that are case-equivalent. The MultiStage2.py
837            script that generates these tables (the pcre_ucd.c file) now scans
838            CaseFolding.txt instead of UnicodeData.txt for character case
839            information.
840    
841        (b) The code for adding characters or ranges of characters to a character
842            class has been abstracted into a generalized function that also handles
843            case-independence. In UTF-mode with UCP support, this uses the new data
844            to handle characters with more than one other case.
845    
846        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
847            whose other case is greater than 256 are now correctly matched
848            caselessly. Previously, the high codepoint matched the low one, but not
849            vice versa.
850    
851        (d) The processing of \h, \H, \v, and \ in character classes now makes use
852            of the new class addition function, using character lists defined as
853            macros alongside the case definitions of 20 above.
854    
855        (e) Caseless back references now work with characters that have more than
856            one other case.
857    
858        (f) General caseless matching of characters with more than one other case
859            is supported.
860    
861    22. Unicode character properties were updated from Unicode 6.2.0
862    
863    23. Improved CMake support under Windows. Patch by Daniel Richard G.
864    
865    24. Add support for 32-bit character strings, and UTF-32
866    
867    25. Major JIT compiler update (code refactoring and bugfixing).
868        Experimental Sparc 32 support is added.
869    
870    26. Applied a modified version of Daniel Richard G's patch to create
871        pcre.h.generic and config.h.generic by "make" instead of in the
872        PrepareRelease script.
873    
874    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
875        pcre_compile.c when checking for a zero character.
876    
877    28. Introducing a native interface for JIT. Through this interface, the compiled
878        machine code can be directly executed. The purpose of this interface is to
879        provide fast pattern matching, so several sanity checks are not performed.
880        However, feature tests are still performed. The new interface provides
881        1.4x speedup compared to the old one.
882    
883    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
884        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
885        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
886    
887    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
888        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
889        the "old" RFC 2279). Instead, it ended up passing a negative length to
890        pcre_exec().
891    
892    31. Add support for GCC's visibility feature to hide internal functions.
893    
894    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
895        "unknown -C option" after outputting 0 or 1.
896    
897    33. There is now support for generating a code coverage report for the test
898        suite in environments where gcc is the compiler and lcov is installed. This
899        is mainly for the benefit of the developers.
900    
901    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
902        unaddressable using valgrind annotations, allowing valgrind to detect
903        invalid memory accesses. This is mainly for the benefit of the developers.
904    
905    25. (*UTF) can now be used to start a pattern in any of the three libraries.
906    
907    26. Give configure error if --enable-cpp but no C++ compiler found.
908    
909    
910    Version 8.31 06-July-2012
911    -------------------------
912    
913    1.  Fixing a wrong JIT test case and some compiler warnings.
914    
915    2.  Removed a bashism from the RunTest script.
916    
917    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
918        to unsigned type, result still unsigned" that was given by an MS compiler
919        on encountering the code "-sizeof(xxx)".
920    
921    4.  Partial matching support is added to the JIT compiler.
922    
923    5.  Fixed several bugs concerned with partial matching of items that consist
924        of more than one character:
925    
926        (a) /^(..)\1/ did not partially match "aba" because checking references was
927            done on an "all or nothing" basis. This also applied to repeated
928            references.
929    
930        (b) \R did not give a hard partial match if \r was found at the end of the
931            subject.
932    
933        (c) \X did not give a hard partial match after matching one or more
934            characters at the end of the subject.
935    
936        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
937            a partial match for the string "\r".
938    
939        (e) When newline was set to CRLF, the metacharacter "." did not recognize
940            a partial match for a CR character at the end of the subject string.
941    
942    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
943        running pcretest, the text "(JIT)" added to the output whenever JIT is
944        actually used to run the match.
945    
946    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
947        or /S+[+] with a digit between 1 and 7.
948    
949    8.  OP_NOT now supports any UTF character not just single-byte ones.
950    
951    9.  (*MARK) control verb is now supported by the JIT compiler.
952    
953    10. The command "./RunTest list" lists the available tests without actually
954        running any of them. (Because I keep forgetting what they all are.)
955    
956    11. Add PCRE_INFO_MAXLOOKBEHIND.
957    
958    12. Applied a (slightly modified) user-supplied patch that improves performance
959        when the heap is used for recursion (compiled with --disable-stack-for-
960        recursion). Instead of malloc and free for each heap frame each time a
961        logical recursion happens, frames are retained on a chain and re-used where
962        possible. This sometimes gives as much as 30% improvement.
963    
964    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
965        call.
966    
967    14. As documented, (*COMMIT) is now confined to within a positive assertion.
968    
969    15. It is now possible to link pcretest with libedit as an alternative to
970        libreadline.
971    
972    16. (*COMMIT) control verb is now supported by the JIT compiler.
973    
974    17. The Unicode data tables have been updated to Unicode 6.1.0.
975    
976    18. Added --file-list option to pcregrep.
977    
978    19. Added binary file support to pcregrep, including the -a, --binary-files,
979        -I, and --text options.
980    
981    20. The madvise function is renamed for posix_madvise for QNX compatibility
982        reasons. Fixed by Giuseppe D'Angelo.
983    
984    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
985    
986    22. Changed the option for creating symbolic links for 16-bit man pages from
987        -s to -sf so that re-installing does not cause issues.
988    
989    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
990    
991    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
992        matches in certain environments (the workspace was not being correctly
993        retained). Also added to pcre_dfa_exec() a simple plausibility check on
994        some of the workspace data at the beginning of a restart.
995    
996    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
997        was not doing so when it should - probably a typo introduced by SVN 528
998        (change 8.10/14).
999    
1000    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1001        \w+ when the character tables indicated that \x{c4} was a word character.
1002        There were several related cases, all because the tests for doing a table
1003        lookup were testing for characters less than 127 instead of 255.
1004    
1005    27. If a pattern contains capturing parentheses that are not used in a match,
1006        their slots in the ovector are set to -1. For those that are higher than
1007        any matched groups, this happens at the end of processing. In the case when
1008        there were back references that the ovector was too small to contain
1009        (causing temporary malloc'd memory to be used during matching), and the
1010        highest capturing number was not used, memory off the end of the ovector
1011        was incorrectly being set to -1. (It was using the size of the temporary
1012        memory instead of the true size.)
1013    
1014    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1015        ovector size, it uses memory at the end of the block that it has got.
1016    
1017    29. Check for an overlong MARK name and give an error at compile time. The
1018        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1019    
1020    30. JIT compiler update.
1021    
1022    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1023        Rill for the patch.
1024    
1025    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1026    
1027    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1028    
1029    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1030        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1031        was enabled.
1032    
1033    35. Improve JIT code generation for greedy plus quantifier.
1034    
1035    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1036        1 to "aa" instead of to an empty string. The bug affected repeated groups
1037        that could potentially match an empty string.
1038    
1039    37. Optimizing single character iterators in JIT.
1040    
1041    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1042        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1043        codepoints that are too big for the mode are faulted, and in a UTF mode,
1044        disallowed codepoints are also faulted.
1045    
1046    39. If PCRE was compiled with UTF support, in three places in the DFA
1047        matcher there was code that should only have been obeyed in UTF mode, but
1048        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1049        processing when bytes with values greater than 127 were present. In 16-bit
1050        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1051        both cases the values are those that cannot be the first data item in a UTF
1052        character. The three items that might have provoked this were recursions,
1053        possessively repeated groups, and atomic groups.
1054    
1055    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1056        and pcregrep, because some OS require shared objects to be explicitly
1057        passed to ld, causing the link step to fail if they are not.
1058    
1059    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1060        mode, patterns that started with \h* or \R* might be incorrectly matched.
1061    
1062    
1063    Version 8.30 04-February-2012
1064    -----------------------------
1065    
1066    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1067        name is defined in ctype.h.
1068    
1069    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1070        only in quite long subpatterns.
1071    
1072    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1073        since it was replaced by pcre_fullinfo() in February 2000.
1074    
1075    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1076        match a (*MARK), and the match failed at the start of the subject, a
1077        reference to memory before the start of the subject could occur. This bug
1078        was introduced by fix 17 of release 8.21.
1079    
1080    5.  A reference to an unset group with zero minimum repetition was giving
1081        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1082        /(another)?(\1?)test/ matched against "hello world test". This bug was
1083        introduced in release 8.13.
1084    
1085    6.  Add support for 16-bit character strings (a large amount of work involving
1086        many changes and refactorings).
1087    
1088    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1089        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1090        from a file.
1091    
1092    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1093        rounding is not applied in this particular case).
1094    
1095    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1096        if they appear, or are escaped, in patterns.
1097    
1098    10. Get rid of a number of -Wunused-but-set-variable warnings.
1099    
1100    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1101        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1102        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1103        also returns the mark "x". This bug applied to capturing parentheses,
1104        non-capturing parentheses, and atomic parentheses. It also applied to some
1105        assertions.
1106    
1107    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1108        information out of configure.ac instead of relying on pcre.h.generic, which
1109        is not stored in the repository.
1110    
1111    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1112        -lreadline.
1113    
1114    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1115    
1116    15. Applied Graycode's patch to put the top-level frame on the stack rather
1117        than the heap when not using the stack for recursion. This gives a
1118        performance improvement in many cases when recursion is not deep.
1119    
1120    16. Experimental code added to "pcretest -C" to output the stack frame size.
1121    
1122    
1123    Version 8.21 12-Dec-2011
1124    ------------------------
1125    
1126    1.  Updating the JIT compiler.
1127    
1128    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1129        are added as well.
1130    
1131    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1132        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1133        calling _pcre_jit_exec. Some extra comments are added.
1134    
1135    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1136        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1137        was introduced by change 18 for 8.20.
1138    
1139    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1140        ECMA-262 standard.
1141    
1142    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1143        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1144        This bug was probably introduced by change 9 of 8.13.
1145    
1146    7.  While fixing 6 above, I noticed that a number of other items were being
1147        incorrectly rejected as "not fixed length". This arose partly because newer
1148        opcodes had not been added to the fixed-length checking code. I have (a)
1149        corrected the bug and added tests for these items, and (b) arranged for an
1150        error to occur if an unknown opcode is encountered while checking for fixed
1151        length instead of just assuming "not fixed length". The items that were
1152        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1153        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1154        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1155    
1156    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1157        being incorrectly compiled and would have given unpredicatble results.
1158    
1159    9.  A possessively repeated subpattern with minimum repeat count greater than
1160        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1161        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1162        the first (A) could occur when it should not.
1163    
1164    10. Add a cast and remove a redundant test from the code.
1165    
1166    11. JIT should use pcre_malloc/pcre_free for allocation.
1167    
1168    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1169        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1170        is anything other than /usr, -L is still shown).
1171    
1172    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1173    
1174    14. Perl does not support \N without a following name in a [] class; PCRE now
1175        also gives an error.
1176    
1177    15. If a forward reference was repeated with an upper limit of around 2000,
1178        it caused the error "internal error: overran compiling workspace". The
1179        maximum number of forward references (including repeats) was limited by the
1180        internal workspace, and dependent on the LINK_SIZE. The code has been
1181        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1182        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1183        of around 200,000 forward references. While doing this, I also speeded up
1184        the filling in of repeated forward references.
1185    
1186    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1187        incorrectly expecting the subject to contain another "a" after the start.
1188    
1189    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1190        in the match, the SKIP should be ignored. This was not happening; instead
1191        the SKIP was being treated as NOMATCH. For patterns such as
1192        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1193        tested.
1194    
1195    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1196        now much more compatible with Perl, in particular in cases where the result
1197        is a non-match for a non-anchored pattern. For example, if
1198        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1199        "m", where previously it did not return a name. A side effect of this
1200        change is that for partial matches, the last encountered mark name is
1201        returned, as for non matches. A number of tests that were previously not
1202        Perl-compatible have been moved into the Perl-compatible test files. The
1203        refactoring has had the pleasing side effect of removing one argument from
1204        the match() function, thus reducing its stack requirements.
1205    
1206    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1207        subsequent uses of /S (without +) incorrectly behaved like /S+.
1208    
1209    21. Retrieve executable code size support for the JIT compiler and fixing
1210        some warnings.
1211    
1212    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1213        not work when the shorter character appeared right at the end of the
1214        subject string.
1215    
1216    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1217        systems.
1218    
1219    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1220        output it when the /M option is used in pcretest.
1221    
1222    25. The CheckMan script was not being included in the distribution. Also, added
1223        an explicit "perl" to run Perl scripts from the PrepareRelease script
1224        because this is reportedly needed in Windows.
1225    
1226    26. If study data was being save in a file and studying had not found a set of
1227        "starts with" bytes for the pattern, the data written to the file (though
1228        never used) was taken from uninitialized memory and so caused valgrind to
1229        complain.
1230    
1231    27. Updated RunTest.bat as provided by Sheri Pierce.
1232    
1233    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1234    
1235    29. Computation of memory usage for the table of capturing group names was
1236        giving an unnecessarily large value.
1237    
1238    
1239    Version 8.20 21-Oct-2011
1240    ------------------------
1241    
1242    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1243        a POSIX class. After further experiments with Perl, which convinced me that
1244        Perl has bugs and confusions, a closing square bracket is no longer allowed
1245        in a POSIX name. This bug also affected patterns with classes that started
1246        with full stops.
1247    
1248    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1249        captured substring, but while checking the failing first alternative,
1250        substring 1 is temporarily captured. If the output vector supplied to
1251        pcre_exec() was not big enough for this capture, the yield of the function
1252        was still zero ("insufficient space for captured substrings"). This cannot
1253        be totally fixed without adding another stack variable, which seems a lot
1254        of expense for a edge case. However, I have improved the situation in cases
1255        such as /(a)(b)x|abc/ matched against "abc", where the return code
1256        indicates that fewer than the maximum number of slots in the ovector have
1257        been set.
1258    
1259    3.  Related to (2) above: when there are more back references in a pattern than
1260        slots in the output vector, pcre_exec() uses temporary memory during
1261        matching, and copies in the captures as far as possible afterwards. It was
1262        using the entire output vector, but this conflicts with the specification
1263        that only 2/3 is used for passing back captured substrings. Now it uses
1264        only the first 2/3, for compatibility. This is, of course, another edge
1265        case.
1266    
1267    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1268        main code base, and can be used by building with --enable-jit. When this is
1269        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1270        runtime --no-jit option is given.
1271    
1272    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1273        ovector, the return from the function was zero, implying that there were
1274        other matches that did not fit. The correct "exactly full" value is now
1275        returned.
1276    
1277    6.  If a subpattern that was called recursively or as a subroutine contained
1278        (*PRUNE) or any other control that caused it to give a non-standard return,
1279        invalid errors such as "Error -26 (nested recursion at the same subject
1280        position)" or even infinite loops could occur.
1281    
1282    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1283        computing the minimum length on reaching *ACCEPT, and so ended up with the
1284        wrong value of 1 rather than 0. Further investigation indicates that
1285        computing a minimum subject length in the presence of *ACCEPT is difficult
1286        (think back references, subroutine calls), and so I have changed the code
1287        so that no minimum is registered for a pattern that contains *ACCEPT.
1288    
1289    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1290        it was not handled as intended. [But see 16 below.]
1291    
1292    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1293        Sheri Pierce.
1294    
1295    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1296        the first byte in a match must be "a".
1297    
1298    11. Change 17 for 8.13 increased the recursion depth for patterns like
1299        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1300        pattern contains any instances of (*THEN). If it does not, the old
1301        optimizations are restored. It would be nice to do this on a per-group
1302        basis, but at the moment that is not feasible.
1303    
1304    12. In some environments, the output of pcretest -C is CRLF terminated. This
1305        broke RunTest's code that checks for the link size. A single white space
1306        character after the value is now allowed for.
1307    
1308    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1309        For "fr", it uses the Windows-specific input and output files.
1310    
1311    14. If (*THEN) appeared in a group that was called recursively or as a
1312        subroutine, it did not work as intended. [But see next item.]
1313    
1314    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1315        pattern fragments (but not containing any | characters). If A and B are
1316        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1317        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1318        D. In other words, Perl considers parentheses that do not contain any |
1319        characters to be part of a surrounding alternative, whereas PCRE was
1320        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1321        differently. PCRE now behaves in the same way as Perl, except in the case
1322        of subroutine/recursion calls such as (?1) which have in any case always
1323        been different (but PCRE had them first :-).
1324    
1325    16. Related to 15 above: Perl does not treat the | in a conditional group as
1326        creating alternatives. Such a group is treated in the same way as an
1327        ordinary group without any | characters when processing (*THEN). PCRE has
1328        been changed to match Perl's behaviour.
1329    
1330    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1331        RunGrepTest script failed.
1332    
1333    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1334        inevitable for groups that contain captures, but it can lead to a lot of
1335        stack use in large patterns. The old behaviour has been restored for atomic
1336        groups that do not contain any capturing parentheses.
1337    
1338    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1339        suppress the check for a minimum subject length at run time. (If it was
1340        given to pcre_exec() or pcre_dfa_exec() it did work.)
1341    
1342    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1343        fail to work when decoding hex characters in data strings in EBCDIC
1344        environments.
1345    
1346    21. It appears that in at least one Mac OS environment, the isxdigit() function
1347        is implemented as a macro that evaluates to its argument more than once,
1348        contravening the C 90 Standard (I haven't checked a later standard). There
1349        was an instance in pcretest which caused it to go wrong when processing
1350        \x{...} escapes in subject strings. The has been rewritten to avoid using
1351        things like p++ in the argument of isxdigit().
1352    
1353    
1354    Version 8.13 16-Aug-2011
1355    ------------------------
1356    
1357    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1358    
1359    2.  Two minor typos in pcre_internal.h have been fixed.
1360    
1361    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1362        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1363        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1364    
1365    4.  There were a number of related bugs in the code for matching backrefences
1366        caselessly in UTF-8 mode when codes for the characters concerned were
1367        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1368        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1369        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1370        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1371        2-byte code at the end of the subject (it thought there wasn't enough data
1372        left).
1373    
1374    5.  Comprehensive information about what went wrong is now returned by
1375        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1376        as the output vector has at least 2 elements. The offset of the start of
1377        the failing character and a reason code are placed in the vector.
1378    
1379    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1380        now returned is for the first byte of the failing character, instead of the
1381        last byte inspected. This is an incompatible change, but I hope it is small
1382        enough not to be a problem. It makes the returned offset consistent with
1383        pcre_exec() and pcre_dfa_exec().
1384    
1385    7.  pcretest now gives a text phrase as well as the error number when
1386        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1387        failure, the offset and reason code are output.
1388    
1389    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1390        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1391        back over a single character (\n). This seems wrong (because it treated the
1392        two characters as a single entity when going forwards), conflicts with the
1393        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1394        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1395        has been changed.
1396    
1397    9.  Some internal refactoring has changed the processing so that the handling
1398        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1399        time (the PCRE_DOTALL option was changed this way some time ago: version
1400        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1401        which was always a bit of a fudge. It also means that there is one less
1402        argument for the match() function, which reduces its stack requirements
1403        slightly. This change also fixes an incompatibility with Perl: the pattern
1404        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1405    
1406    10. More internal refactoring has drastically reduced the number of recursive
1407        calls to match() for possessively repeated groups such as (abc)++ when
1408        using pcre_exec().
1409    
1410    11. While implementing 10, a number of bugs in the handling of groups were
1411        discovered and fixed:
1412    
1413        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1414        (a|)*(?1) gave a compile-time internal error.
1415        ((a|)+)+  did not notice that the outer group could match an empty string.
1416        (^a|^)+   was not marked as anchored.
1417        (.*a|.*)+ was not marked as matching at start or after a newline.
1418    
1419    12. Yet more internal refactoring has removed another argument from the match()
1420        function. Special calls to this function are now indicated by setting a
1421        value in a variable in the "match data" data block.
1422    
1423    13. Be more explicit in pcre_study() instead of relying on "default" for
1424        opcodes that mean there is no starting character; this means that when new
1425        ones are added and accidentally left out of pcre_study(), testing should
1426        pick them up.
1427    
1428    14. The -s option of pcretest has been documented for ages as being an old
1429        synonym of -m (show memory usage). I have changed it to mean "force study
1430        for every regex", that is, assume /S for every regex. This is similar to -i
1431        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1432        using it. It makes it easier to run collections of tests with and without
1433        study enabled, and thereby test pcre_study() more easily. All the standard
1434        tests are now run with and without -s (but some patterns can be marked as
1435        "never study" - see 20 below).
1436    
1437    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1438        restoration of the capturing data to the outer values was not happening
1439        correctly.
1440    
1441    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1442        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1443        pattern had matched an empty string, and so incorrectly returned a no
1444        match.
1445    
1446    17. There was optimizing code for the last branch of non-capturing parentheses,
1447        and also for the obeyed branch of a conditional subexpression, which used
1448        tail recursion to cut down on stack usage. Unfortunately, now that there is
1449        the possibility of (*THEN) occurring in these branches, tail recursion is
1450        no longer possible because the return has to be checked for (*THEN). These
1451        two optimizations have therefore been removed. [But see 8.20/11 above.]
1452    
1453    18. If a pattern containing \R was studied, it was assumed that \R always
1454        matched two bytes, thus causing the minimum subject length to be
1455        incorrectly computed because \R can also match just one byte.
1456    
1457    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1458        was incorrectly computed.
1459    
1460    20. If /S is present twice on a test pattern in pcretest input, it now
1461        *disables* studying, thereby overriding the use of -s on the command line
1462        (see 14 above). This is necessary for one or two tests to keep the output
1463        identical in both cases.
1464    
1465    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1466        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1467    
1468    22. When an atomic group that contained a capturing parenthesis was
1469        successfully matched, but the branch in which it appeared failed, the
1470        capturing was not being forgotten if a higher numbered group was later
1471        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1472        group 1 to "a", when in fact it should be unset. This applied to multi-
1473        branched capturing and non-capturing groups, repeated or not, and also to
1474        positive assertions (capturing in negative assertions does not happen
1475        in PCRE) and also to nested atomic groups.
1476    
1477    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1478        subject after a captured substring, to make it easier to tell which of a
1479        number of identical substrings has been captured.
1480    
1481    24. The way atomic groups are processed by pcre_exec() has been changed so that
1482        if they are repeated, backtracking one repetition now resets captured
1483        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1484        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1485        "aaa". Previously, it would have been "a". As part of this code
1486        refactoring, the way recursive calls are handled has also been changed.
1487    
1488    25. If an assertion condition captured any substrings, they were not passed
1489        back unless some other capturing happened later. For example, if
1490        (?(?=(a))a) was matched against "a", no capturing was returned.
1491    
1492    26. When studying a pattern that contained subroutine calls or assertions,
1493        the code for finding the minimum length of a possible match was handling
1494        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1495        group 1 called group 2 while simultaneously a separate group 2 called group
1496        1). A stack overflow occurred in this case. I have fixed this by limiting
1497        the recursion depth to 10.
1498    
1499    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1500        Fortmann. This supports explicit test numbers on the command line, and has
1501        argument validation and error reporting.
1502    
1503    28. An instance of \X with an unlimited repeat could fail if at any point the
1504        first character it looked at was a mark character.
1505    
1506    29. Some minor code refactoring concerning Unicode properties and scripts
1507        should reduce the stack requirement of match() slightly.
1508    
1509    30. Added the '=' option to pcretest to check the setting of unused capturing
1510        slots at the end of the pattern, which are documented as being -1, but are
1511        not included in the return count.
1512    
1513    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1514        compiled something random. Now it gives a compile-time error (as does
1515        Perl).
1516    
1517    32. A *MARK encountered during the processing of a positive assertion is now
1518        recorded and passed back (compatible with Perl).
1519    
1520    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1521        had alternative anchored branches, the search for a second match in a line
1522        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1523        matched the line "0102" twice. The same bug affected patterns that started
1524        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1525        twice.
1526    
1527    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1528        does, and because of capturing effects, quantifying parenthesized
1529        assertions may at times be useful. Quantifiers are now allowed for
1530        parenthesized assertions.
1531    
1532    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1533    
1534    36. \g was being checked for fancy things in a character class, when it should
1535        just be a literal "g".
1536    
1537    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1538        appearance of a nested POSIX class supersedes an apparent external class.
1539        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1540        unescaped square brackets may also appear as part of class names. For
1541        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1542        more like Perl. (But see 8.20/1 above.)
1543    
1544    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1545        was because it thought it was \N{name}, which is not supported).
1546    
1547    39. Add minix to OS list not supporting the -S option in pcretest.
1548    
1549    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1550        cannot analyze patterns in sufficient detail to catch mutual recursions
1551        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1552        subgroup is called recursively as a subpattern for a second time at the
1553        same position in the subject string. In previous releases this might have
1554        been caught by the recursion limit, or it might have run out of stack.
1555    
1556    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1557        happen only once. PCRE was, however incorrectly giving a compile time error
1558        "recursive call could loop indefinitely" because it cannot analyze the
1559        pattern in sufficient detail. The compile time test no longer happens when
1560        PCRE is compiling a conditional subpattern, but actual runaway loops are
1561        now caught at runtime (see 40 above).
1562    
1563    42. It seems that Perl allows any characters other than a closing parenthesis
1564        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1565        has been changed to be the same.
1566    
1567    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1568        as not to get warnings when autogen.sh is called. Also changed
1569        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1570    
1571    44. To help people who use pcregrep to scan files containing exceedingly long
1572        lines, the following changes have been made:
1573    
1574        (a) The default value of the buffer size parameter has been increased from
1575            8K to 20K. (The actual buffer used is three times this size.)
1576    
1577        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1578            PCRE is built.
1579    
1580        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1581            to be set at run time.
1582    
1583        (d) Numerical values in pcregrep options can be followed by K or M, for
1584            example --buffer-size=50K.
1585    
1586        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1587            given and the return code is set to 2.
1588    
1589    45. Add a pointer to the latest mark to the callout data block.
1590    
1591    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1592        partial match of an empty string instead of no match. This was specific to
1593        the use of ".".
1594    
1595    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1596        complete match instead of a partial match. This bug was dependent on both
1597        the PCRE_UTF8 and PCRE_DOTALL options being set.
1598    
1599    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1600        starting byte set, because \b was not being ignored.
1601    
1602    
1603    Version 8.12 15-Jan-2011
1604    ------------------------
1605    
1606    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1607        checks for such things as part of the documentation building process.
1608    
1609    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1610        --match-limit and --recursion-limit options (added for 8.11). In
1611        particular, this made one of the standard tests fail. (The integer value
1612        went into the wrong half of a long int.)
1613    
1614    3.  If the --colour option was given to pcregrep with -v (invert match), it
1615        did strange things, either producing crazy output, or crashing. It should,
1616        of course, ignore a request for colour when reporting lines that do not
1617        match.
1618    
1619    4.  Another pcregrep bug caused similar problems if --colour was specified with
1620        -M (multiline) and the pattern match finished with a line ending.
1621    
1622    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1623        matched in multiline mode, the following line was shown as part of the
1624        match. This seems wrong, so I have changed it.
1625    
1626    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1627        the check for further matches in the same line (so they could be coloured)
1628        to overrun the end of the current line. If another match was found, it was
1629        incorrectly shown (and then shown again when found in the next line).
1630    
1631    7.  If pcregrep was compiled under Windows, there was a reference to the
1632        function pcregrep_exit() before it was defined. I am assuming this was
1633        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1634        reported by a user. I've moved the definition above the reference.
1635    
1636    
1637    Version 8.11 10-Dec-2010
1638    ------------------------
1639    
1640    1.  (*THEN) was not working properly if there were untried alternatives prior
1641        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1642        backtracked to try for "b" instead of moving to the next alternative branch
1643        at the same level (in this case, to look for "c"). The Perl documentation
1644        is clear that when (*THEN) is backtracked onto, it goes to the "next
1645        alternative in the innermost enclosing group".
1646    
1647    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1648        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1649        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1650        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1651        (*THEN).
1652    
1653    3.  If \s appeared in a character class, it removed the VT character from
1654        the class, even if it had been included by some previous item, for example
1655        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1656        of \s, but is part of the POSIX "space" class.)
1657    
1658    4.  A partial match never returns an empty string (because you can always
1659        match an empty string at the end of the subject); however the checking for
1660        an empty string was starting at the "start of match" point. This has been
1661        changed to the "earliest inspected character" point, because the returned
1662        data for a partial match starts at this character. This means that, for
1663        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1664        (previously it gave "no match").
1665    
1666    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1667        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1668        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1669        has an implication that the given string is incomplete (because a partial
1670        match is preferred over a full match). For this reason, these items now
1671        give a partial match in this situation. [Aside: previously, the one case
1672        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1673        match rather than a full match, which was wrong by the old rules, but is
1674        now correct.]
1675    
1676    6.  There was a bug in the handling of #-introduced comments, recognized when
1677        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1678        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1679        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1680        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1681        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1682        places in pcre_compile().
1683    
1684    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1685        comments when looking ahead for named forward references to subpatterns,
1686        the only newline sequence it recognized was NL. It now handles newlines
1687        according to the set newline convention.
1688    
1689    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1690        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1691        cater for a lack of strerror(). These oversights have been fixed.
1692    
1693    9.  Added --match-limit and --recursion-limit to pcregrep.
1694    
1695    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1696    
1697    11. When the -o option was used, pcregrep was setting a return code of 1, even
1698        when matches were found, and --line-buffered was not being honoured.
1699    
1700    12. Added an optional parentheses number to the -o and --only-matching options
1701        of pcregrep.
1702    
1703    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1704        can match an empty string. The code to do it in pcretest and pcredemo
1705        needed fixing:
1706    
1707        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1708            only one byte after an empty string match just before CRLF (this case
1709            just got forgotten; "any" and "anycrlf" were OK).
1710    
1711        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1712            mode when an empty string match preceded an ASCII character followed by
1713            a non-ASCII character. (The code for advancing by one character rather
1714            than one byte was nonsense.)
1715    
1716        (c) The pcredemo.c sample program did not have any code at all to handle
1717            the cases when CRLF is a valid newline sequence.
1718    
1719    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1720        as a starting offset was within the subject string. There is now a new
1721        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1722        negative or greater than the length of the string. In order to test this,
1723        pcretest is extended to allow the setting of negative starting offsets.
1724    
1725    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1726        starting offset points to the beginning of a UTF-8 character was
1727        unnecessarily clumsy. I tidied it up.
1728    
1729    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1730        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1731    
1732    17. Nobody had reported that the --include_dir option, which was added in
1733        release 7.7 should have been called --include-dir (hyphen, not underscore)
1734        for compatibility with GNU grep. I have changed it to --include-dir, but
1735        left --include_dir as an undocumented synonym, and the same for
1736        --exclude-dir, though that is not available in GNU grep, at least as of
1737        release 2.5.4.
1738    
1739    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1740        characters from a string of bytes) have been redefined so as not to use
1741        loops, in order to improve performance in some environments. At the same
1742        time, I abstracted some of the common code into auxiliary macros to save
1743        repetition (this should not affect the compiled code).
1744    
1745    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1746        compile-time error is now given if \c is not followed by an ASCII
1747        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1748        different, and any byte value is allowed.)
1749    
1750    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1751        START_OPTIMIZE option, which is now allowed at compile time - but just
1752        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1753        to pcregrep and other applications that have no direct access to PCRE
1754        options. The new /Y option in pcretest sets this option when calling
1755        pcre_compile().
1756    
1757    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1758        back references. Groups containing recursive back references were forced to
1759        be atomic by that change, but in the case of named groups, the amount of
1760        memory required was incorrectly computed, leading to "Failed: internal
1761        error: code overflow". This has been fixed.
1762    
1763    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1764        pcretest.c, to avoid build problems in some Borland environments.
1765    
1766    
1767    Version 8.10 25-Jun-2010
1768    ------------------------
1769    
1770    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1771        THEN.
1772    
1773    2.  (*ACCEPT) was not working when inside an atomic group.
1774    
1775    3.  Inside a character class, \B is treated as a literal by default, but
1776        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1777        causes the error). The code is unchanged, but I tidied the documentation.
1778    
1779    4.  Inside a character class, PCRE always treated \R and \X as literals,
1780        whereas Perl faults them if its -w option is set. I have changed PCRE so
1781        that it faults them when PCRE_EXTRA is set.
1782    
1783    5.  Added support for \N, which always matches any character other than
1784        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1785    
1786    6.  When compiling pcregrep with newer versions of gcc which may have
1787        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1788        declared with attribute warn_unused_result" were given. Just casting the
1789        result to (void) does not stop the warnings; a more elaborate fudge is
1790        needed. I've used a macro to implement this.
1791    
1792    7.  Minor change to pcretest.c to avoid a compiler warning.
1793    
1794    8.  Added four artifical Unicode properties to help with an option to make
1795        \s etc use properties (see next item). The new properties are: Xan
1796        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1797    
1798    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1799        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1800        this option. Modified pcretest to add /W to test this facility. Added
1801        REG_UCP to make it available via the POSIX interface.
1802    
1803    10. Added --line-buffered to pcregrep.
1804    
1805    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1806        studied, and the match started with a letter with a code point greater than
1807        127 whose first byte was different to the first byte of the other case of
1808        the letter, the other case of this starting letter was not recognized
1809        (#976).
1810    
1811    12. If a pattern that was studied started with a repeated Unicode property
1812        test, for example, \p{Nd}+, there was the theoretical possibility of
1813        setting up an incorrect bitmap of starting bytes, but fortunately it could
1814        not have actually happened in practice until change 8 above was made (it
1815        added property types that matched character-matching opcodes).
1816    
1817    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1818        possible starting bytes for non-anchored patterns.
1819    
1820    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1821        \R, and also a number of cases that involve Unicode properties, both
1822        explicit and implicit when PCRE_UCP is set.
1823    
1824    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1825        input, it could crash or give wrong results if characters with values
1826        greater than 0xc0 were present in the subject string. (Detail: it assumed
1827        UTF-8 input when processing these items.)
1828    
1829    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1830        size_t is 64-bit (#991).
1831    
1832    17. Added a check for running out of memory when PCRE is compiled with
1833        --disable-stack-for-recursion (#990).
1834    
1835    18. If the last data line in a file for pcretest does not have a newline on
1836        the end, a newline was missing in the output.
1837    
1838    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1839        less than 128) in its various bitmaps. However, there is a facility for
1840        generating tables according to the current locale when PCRE is compiled. It
1841        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1842        characters, are recognized by isspace() and therefore were getting set in
1843        these tables, and indeed these tables seem to approximate to ISO 8859. This
1844        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1845        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1846        which of course cannot start UTF-8 characters. I have changed the code so
1847        that only real ASCII characters (less than 128) and the correct starting
1848        bytes for UTF-8 encodings are set for characters greater than 127 when in
1849        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1850        altogether.)
1851    
1852    20. Added the /T option to pcretest so as to be able to run tests with non-
1853        standard character tables, thus making it possible to include the tests
1854        used for 19 above in the standard set of tests.
1855    
1856    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1857        reference to a subpattern the other side of a comment that contains an
1858        opening parenthesis caused either an internal compiling error, or a
1859        reference to the wrong subpattern.
1860    
1861    
1862    Version 8.02 19-Mar-2010
1863    ------------------------
1864    
1865    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1866    
1867    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1868        configured.
1869    
1870    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1871        original author of that file, following a query about its status.
1872    
1873    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1874        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1875    
1876    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1877        quantifier applied to a forward-referencing subroutine call, could compile
1878        incorrect code or give the error "internal error: previously-checked
1879        referenced subpattern not found".
1880    
1881    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1882        variables to point to external functions. For these systems, therefore,
1883        pcre_malloc etc. are now initialized to local functions that call the
1884        relevant global functions.
1885    
1886    7.  There were two entries missing in the vectors called coptable and poptable
1887        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1888        I've fixed the data, and added a kludgy way of testing at compile time that
1889        the lengths are correct (equal to the number of opcodes).
1890    
1891    8.  Following on from 7, I added a similar kludge to check the length of the
1892        eint vector in pcreposix.c.
1893    
1894    9.  Error texts for pcre_compile() are held as one long string to avoid too
1895        much relocation at load time. To find a text, the string is searched,
1896        counting zeros. There was no check for running off the end of the string,
1897        which could happen if a new error number was added without updating the
1898        string.
1899    
1900    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1901    
1902    11. \K was not working if it appeared in an atomic group or in a group that
1903        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1904        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1905        the assertion is positive, but not if it is negative.
1906    
1907    12. Change 11 fortuitously reduced the size of the stack frame used in the
1908        "match()" function of pcre_exec.c by one pointer. Forthcoming
1909        implementation of support for (*MARK) will need an extra pointer on the
1910        stack; I have reserved it now, so that the stack frame size does not
1911        decrease.
1912    
1913    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1914        item in branch that calls a recursion is a subroutine call - as in the
1915        second branch in the above example - was incorrectly given the compile-
1916        time error "recursive call could loop indefinitely" because pcre_compile()
1917        was not correctly checking the subroutine for matching a non-empty string.
1918    
1919    14. The checks for overrunning compiling workspace could trigger after an
1920        overrun had occurred. This is a "should never occur" error, but it can be
1921        triggered by pathological patterns such as hundreds of nested parentheses.
1922        The checks now trigger 100 bytes before the end of the workspace.
1923    
1924    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1925    
1926    
1927    Version 8.01 19-Jan-2010
1928    ------------------------
1929    
1930    1.  If a pattern contained a conditional subpattern with only one branch (in
1931        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1932        computed the wrong minimum data length (which is of course zero for such
1933        subpatterns). This could cause incorrect "no match" results.
1934    
1935    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1936        the pattern is reset in the first branch, pcre_compile() failed with
1937        "internal error: code overflow at offset...". This happened only when
1938        the reset was to the original external option setting. (An optimization
1939        abstracts leading options settings into an external setting, which was the
1940        cause of this.)
1941    
1942    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1943        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1944        assertion pattern did not match (meaning that the assertion was true), it
1945        was incorrectly treated as false if the SKIP had been reached during the
1946        matching. This also applied to assertions used as conditions.
1947    
1948    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1949        assertion subpattern, including such a pattern used as a condition,
1950        unpredictable results occurred, instead of the error return
1951        PCRE_ERROR_DFA_UITEM.
1952    
1953    5.  The C++ GlobalReplace function was not working like Perl for the special
1954        situation when an empty string is matched. It now does the fancy magic
1955        stuff that is necessary.
1956    
1957    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1958        removed. (These were left over from very, very early versions of PCRE.)
1959    
1960    7.  Some cosmetic changes to the code to make life easier when compiling it
1961        as part of something else:
1962    
1963        (a) Change DEBUG to PCRE_DEBUG.
1964    
1965        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1966            called "current" as "current_branch", to prevent a collision with the
1967            Linux macro when compiled as a kernel module.
1968    
1969        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1970            prevent a collision with the Linux macro when compiled as a kernel
1971            module.
1972    
1973    8.  In pcre_compile() there are some checks for integer overflows that used to
1974        cast potentially large values to (double). This has been changed to that
1975        when building, a check for int64_t is made, and if it is found, it is used
1976        instead, thus avoiding the use of floating point arithmetic. (There is no
1977        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1978        double.
1979    
1980    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1981        2005 (difference between two addresses compared to an unsigned value).
1982    
1983    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1984        custom one, because of the following reported problem in Windows:
1985    
1986          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1987              under Win32.
1988          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1989              therefore missing the function definition.
1990          - The compiler thus generates a "C" signature for the test function.
1991          - The linker fails to find the "C" function.
1992          - PCRE fails to configure if asked to do so against libbz2.
1993    
1994    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1995        messages were output:
1996    
1997          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1998          rerunning libtoolize, to keep the correct libtool macros in-tree.
1999          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
2000    
2001        I have done both of these things.
2002    
2003    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
2004        most of the time, it *can* run out if it is given a pattern that contains a
2005        runaway infinite recursion. I updated the discussion in the pcrestack man
2006        page.
2007    
2008    13. Now that we have gone to the x.xx style of version numbers, the minor
2009        version may start with zero. Using 08 or 09 is a bad idea because users
2010        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
2011        interpreted as invalid octal numbers. I've updated the previous comment in
2012        configure.ac, and also added a check that gives an error if 08 or 09 are
2013        used.
2014    
2015    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
2016        causing partial matching to fail when the end of the subject matched \W
2017        in a UTF-8 pattern where \W was quantified with a minimum of 3.
2018    
2019    15. There were some discrepancies between the declarations in pcre_internal.h
2020        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
2021        their definitions. The declarations used "const uschar *" and the
2022        definitions used USPTR. Even though USPTR is normally defined as "const
2023        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
2024        reported that: "This difference in casting confuses some C++ compilers, for
2025        example, SunCC recognizes above declarations as different functions and
2026        generates broken code for hbpcre." I have changed the declarations to use
2027        USPTR.
2028    
2029    16. GNU libtool is named differently on some systems. The autogen.sh script now
2030        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
2031        (FreeBSD).
2032    
2033    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
2034        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
2035        comment: "Figure out how to create a longlong from a string: strtoll and
2036        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
2037        instance, but it only takes 2 args instead of 3!"
2038    
2039    18. A subtle bug concerned with back references has been fixed by a change of
2040        specification, with a corresponding code fix. A pattern such as
2041        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
2042        refers, was giving matches when it shouldn't. For example, xa=xaaa would
2043        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
2044        same bug. Such groups have to be quantified to be useful, or contained
2045        inside another quantified group. (If there's no repetition, the reference
2046        can never match.) The problem arises because, having left the group and
2047        moved on to the rest of the pattern, a later failure that backtracks into
2048        the group uses the captured value from the final iteration of the group
2049        rather than the correct earlier one. I have fixed this in PCRE by forcing
2050        any group that contains a reference to itself to be an atomic group; that
2051        is, there cannot be any backtracking into it once it has completed. This is
2052        similar to recursive and subroutine calls.
2053    
2054    
2055    Version 8.00 19-Oct-09
2056    ----------------------
2057    
2058    1.  The table for translating pcre_compile() error codes into POSIX error codes
2059        was out-of-date, and there was no check on the pcre_compile() error code
2060        being within the table. This could lead to an OK return being given in
2061        error.
2062    
2063    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
2064        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
2065        in a Windows environment.
2066    
2067    3.  The pcregrep --count option prints the count for each file even when it is
2068        zero, as does GNU grep. However, pcregrep was also printing all files when
2069        --files-with-matches was added. Now, when both options are given, it prints
2070        counts only for those files that have at least one match. (GNU grep just
2071        prints the file name in this circumstance, but including the count seems
2072        more useful - otherwise, why use --count?) Also ensured that the
2073        combination -clh just lists non-zero counts, with no names.
2074    
2075    4.  The long form of the pcregrep -F option was incorrectly implemented as
2076        --fixed_strings instead of --fixed-strings. This is an incompatible change,
2077        but it seems right to fix it, and I didn't think it was worth preserving
2078        the old behaviour.
2079    
2080    5.  The command line items --regex=pattern and --regexp=pattern were not
2081        recognized by pcregrep, which required --regex pattern or --regexp pattern
2082        (with a space rather than an '='). The man page documented the '=' forms,
2083        which are compatible with GNU grep; these now work.
2084    
2085    6.  No libpcreposix.pc file was created for pkg-config; there was just
2086        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
2087    
2088    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
2089        when UCP support is not needed, by modifying the Python script that
2090        generates it from Unicode data files. This should not matter if the module
2091        is correctly used as a library, but I received one complaint about 50K of
2092        unwanted data. My guess is that the person linked everything into his
2093        program rather than using a library. Anyway, it does no harm.
2094    
2095    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
2096        was a minimum greater than 1 for a wide character in a possessive
2097        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
2098        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
2099        character. Chaos in the form of incorrect output or a compiling loop could
2100        result.
2101    
2102    9.  The restrictions on what a pattern can contain when partial matching is
2103        requested for pcre_exec() have been removed. All patterns can now be
2104        partially matched by this function. In addition, if there are at least two
2105        slots in the offset vector, the offset of the earliest inspected character
2106        for the match and the offset of the end of the subject are set in them when
2107        PCRE_ERROR_PARTIAL is returned.
2108    
2109    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
2110        synonymous with PCRE_PARTIAL, for backwards compatibility, and
2111        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
2112        and may be more useful for multi-segment matching.
2113    
2114    11. Partial matching with pcre_exec() is now more intuitive. A partial match
2115        used to be given if ever the end of the subject was reached; now it is
2116        given only if matching could not proceed because another character was
2117        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
2118        string "Z", which now yields "no match" instead of "partial match". In the
2119        case of pcre_dfa_exec(), "no match" is given if every matching path for the
2120        final character ended with (*FAIL).
2121    
2122    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
2123        if the pattern had a "must contain" character that was already found in the
2124        earlier partial match, unless partial matching was again requested. For
2125        example, with the pattern /dog.(body)?/, the "must contain" character is
2126        "g". If the first part-match was for the string "dog", restarting with
2127        "sbody" failed. This bug has been fixed.
2128    
2129    13. The string returned by pcre_dfa_exec() after a partial match has been
2130        changed so that it starts at the first inspected character rather than the
2131        first character of the match. This makes a difference only if the pattern
2132        starts with a lookbehind assertion or \b or \B (\K is not supported by
2133        pcre_dfa_exec()). It's an incompatible change, but it makes the two
2134        matching functions compatible, and I think it's the right thing to do.
2135    
2136    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
2137        so that the demonstration program is easily available in environments where
2138        PCRE has not been installed from source.
2139    
2140    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
2141        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
2142        library.
2143    
2144    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
2145        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
2146        is not the first non-POSIX option to be added. Clearly some people find
2147        these options useful.
2148    
2149    17. If a caller to the POSIX matching function regexec() passes a non-zero
2150        value for nmatch with a NULL value for pmatch, the value of
2151        nmatch is forced to zero.
2152    
2153    18. RunGrepTest did not have a test for the availability of the -u option of
2154        the diff command, as RunTest does. It now checks in the same way as
2155        RunTest, and also checks for the -b option.
2156    
2157    19. If an odd number of negated classes containing just a single character
2158        interposed, within parentheses, between a forward reference to a named
2159        subpattern and the definition of the subpattern, compilation crashed with
2160        an internal error, complaining that it could not find the referenced
2161        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
2162        [The bug was that it was starting one character too far in when skipping
2163        over the character class, thus treating the ] as data rather than
2164        terminating the class. This meant it could skip too much.]
2165    
2166    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
2167        /g option in pcretest when the pattern contains \K, which makes it possible
2168        to have an empty string match not at the start, even when the pattern is
2169        anchored. Updated pcretest and pcredemo to use this option.
2170    
2171    21. If the maximum number of capturing subpatterns in a recursion was greater
2172        than the maximum at the outer level, the higher number was returned, but
2173        with unset values at the outer level. The correct (outer level) value is
2174        now given.
2175    
2176    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
2177        PCRE did not set those parentheses (unlike Perl). I have now found a way to
2178        make it do so. The string so far is captured, making this feature
2179        compatible with Perl.
2180    
2181    23. The tests have been re-organized, adding tests 11 and 12, to make it
2182        possible to check the Perl 5.10 features against Perl 5.10.
2183    
2184    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
2185        pattern matches a fixed length string. PCRE did not allow this; now it
2186        does. Neither allows recursion.
2187    
2188    25. I finally figured out how to implement a request to provide the minimum
2189        length of subject string that was needed in order to match a given pattern.
2190        (It was back references and recursion that I had previously got hung up
2191        on.) This code has now been added to pcre_study(); it finds a lower bound
2192        to the length of subject needed. It is not necessarily the greatest lower
2193        bound, but using it to avoid searching strings that are too short does give
2194        some useful speed-ups. The value is available to calling programs via
2195        pcre_fullinfo().
2196    
2197    26. While implementing 25, I discovered to my embarrassment that pcretest had
2198        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
2199        study optimizations had never been tested with that matching function.
2200        Oops. What is worse, even when it was passed study data, there was a bug in
2201        pcre_dfa_exec() that meant it never actually used it. Double oops. There
2202        were also very few tests of studied patterns with pcre_dfa_exec().
2203    
2204    27. If (?| is used to create subpatterns with duplicate numbers, they are now
2205        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
2206        on the other side of the coin, they are no longer allowed to have different
2207        names, because these cannot be distinguished in PCRE, and this has caused
2208        confusion. (This is a difference from Perl.)
2209    
2210    28. When duplicate subpattern names are present (necessarily with different
2211        numbers, as required by 27 above), and a test is made by name in a
2212        conditional pattern, either for a subpattern having been matched, or for
2213        recursion in such a pattern, all the associated numbered subpatterns are
2214        tested, and the overall condition is true if the condition is true for any
2215        one of them. This is the way Perl works, and is also more like the way
2216        testing by number works.
2217    
2218    
2219    Version 7.9 11-Apr-09
2220    ---------------------
2221    
2222    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
2223        (pcretest), all targets were linked against these libraries. This included
2224        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
2225        libraries. This caused unwanted dependencies to be created. This problem
2226        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
2227        pcretest is linked with readline.
2228    
2229    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
2230        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
2231        moved outside it again, because FALSE and TRUE are already defined in AIX,
2232        but BOOL is not.
2233    
2234    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
2235        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
2236    
2237    4.  The pcregrep documentation said spaces were inserted as well as colons (or
2238        hyphens) following file names and line numbers when outputting matching
2239        lines. This is not true; no spaces are inserted. I have also clarified the
2240        wording for the --colour (or --color) option.
2241    
2242    5.  In pcregrep, when --colour was used with -o, the list of matching strings
2243        was not coloured; this is different to GNU grep, so I have changed it to be
2244        the same.
2245    
2246    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
2247        each matching line was coloured. Now it goes on to look for further matches
2248        of any of the test patterns, which is the same behaviour as GNU grep.
2249    
2250    7.  A pattern that could match an empty string could cause pcregrep to loop; it
2251        doesn't make sense to accept an empty string match in pcregrep, so I have
2252        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2253        seems to be how GNU grep behaves. [But see later change 40 for release
2254        8.33.]
2255    
2256    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2257        start or after a newline", because the conditional assertion was not being
2258        correctly handled. The rule now is that both the assertion and what follows
2259        in the first alternative must satisfy the test.
2260    
2261    9.  If auto-callout was enabled in a pattern with a conditional group whose
2262        condition was an assertion, PCRE could crash during matching, both with
2263        pcre_exec() and pcre_dfa_exec().
2264    
2265    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2266        used for matching.
2267    
2268    11. Unicode property support in character classes was not working for
2269        characters (bytes) greater than 127 when not in UTF-8 mode.
2270    
2271    12. Added the -M command line option to pcretest.
2272    
2273    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2274    
2275    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2276    
2277    16. Added comments and documentation about mis-use of no_arg in the C++
2278        wrapper.
2279    
2280    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2281        from Martin Jerabek that uses macro names for all relevant character and
2282        string constants.
2283    
2284    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2285        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2286        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2287        these, but not everybody uses configure.
2288    
2289    19. A conditional group that had only one branch was not being correctly
2290        recognized as an item that could match an empty string. This meant that an
2291        enclosing group might also not be so recognized, causing infinite looping
2292        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2293        with the subject "ab", where knowledge that the repeated group can match
2294        nothing is needed in order to break the loop.
2295    
2296    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2297        exec(), but without supplying a callout function, matching went wrong.
2298    
2299    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2300        leak if the size of the offset vector was greater than 30. When the vector
2301        is smaller, the saved offsets during recursion go onto a local stack
2302        vector, but for larger vectors malloc() is used. It was failing to free
2303        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2304        error, in fact).
2305    
2306    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2307        heapframe that is used only when UTF-8 support is enabled. This caused no
2308        problem, but was untidy.
2309    
2310    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2311        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2312        included within another project.
2313    
2314    24. Steven Van Ingelgem's patches to add more options to the CMake support,
2315        slightly modified by me:
2316    
2317          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2318              not building pcregrep.
2319    
2320          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2321              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2322    
2323    25. Forward references, both numeric and by name, in patterns that made use of
2324        duplicate group numbers, could behave incorrectly or give incorrect errors,
2325        because when scanning forward to find the reference group, PCRE was not
2326        taking into account the duplicate group numbers. A pattern such as
2327        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2328    
2329    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2330        the feature of a custom pointer more persuasive (as requested by a user).
2331    
2332    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2333        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2334        pre-defined.
2335    
2336    28. Added support for (*UTF8) at the start of a pattern.
2337    
2338    29. Arrange for flags added by the "release type" setting in CMake to be shown
2339        in the configuration summary.
2340    
2341    
2342    Version 7.8 05-Sep-08
2343    ---------------------
2344    
2345    1.  Replaced UCP searching code with optimized version as implemented for Ad
2346        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2347        stage table and inline lookup instead of a function, giving speed ups of 2
2348        to 5 times on some simple patterns that I tested. Permission was given to
2349        distribute the MultiStage2.py script that generates the tables (it's not in
2350        the tarball, but is in the Subversion repository).
2351    
2352    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2353        scripts.
2354    
2355    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2356        a group with a zero qualifier. The result of the study could be incorrect,
2357        or the function might crash, depending on the pattern.
2358    
2359    4.  Caseless matching was not working for non-ASCII characters in back
2360        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2361        It now works when Unicode Property Support is available.
2362    
2363    5.  In pcretest, an escape such as \x{de} in the data was always generating
2364        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2365        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2366        truncation.
2367    
2368    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2369    
2370    7.  Added two (int) casts to pcregrep when printing the difference of two
2371        pointers, in case they are 64-bit values.
2372    
2373    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2374        test 2 if it fails.
2375    
2376    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2377        and a #define of that name to empty if it is not externally set. This is to
2378        allow users of MSVC to set it if necessary.
2379    
2380    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2381        the convenience functions in the pcre_get.c source file.
2382    
2383    11. An option change at the start of a pattern that had top-level alternatives
2384        could cause overwriting and/or a crash. This command provoked a crash in
2385        some environments:
2386    
2387          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2388    
2389        This potential security problem was recorded as CVE-2008-2371.
2390    
2391    12. For a pattern where the match had to start at the beginning or immediately
2392        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2393        pcre_dfa_exec() could read past the end of the passed subject if there was
2394        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2395        pcretest so that it places the subject at the end of its malloc-ed buffer.
2396    
2397    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2398        exec() might read past the end of the data buffer in UTF-8 mode.
2399    
2400    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2401        the data contained the byte 0x85 as part of a UTF-8 character within its
2402        first line. This applied both to normal and DFA matching.
2403    
2404    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2405        /^[^d]*?$/8 failed to match "abc".
2406    
2407    16. Added a missing copyright notice to pcrecpp_internal.h.
2408    
2409    17. Make it more clear in the documentation that values returned from
2410        pcre_exec() in ovector are byte offsets, not character counts.
2411    
2412    18. Tidied a few places to stop certain compilers from issuing warnings.
2413    
2414    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2415        supplied by Stefan Weber. I made a further small update for 7.8 because
2416        there is a change of source arrangements: the pcre_searchfuncs.c module is
2417        replaced by pcre_ucd.c.
2418    
2419    
2420    Version 7.7 07-May-08
2421    ---------------------
2422    
2423    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2424        a string to a long long, pretend we don't even have a long long." This is
2425        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2426    
2427    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2428        pre-7.6 versions, which defined a global no_arg variable instead of putting
2429        it in the RE class. (See also #8 below.)
2430    
2431    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2432        Lopes.
2433    
2434    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2435    
2436        (1) The include/exclude patterns were being applied to the whole pathnames
2437            of files, instead of just to the final components.
2438    
2439        (2) If there was more than one level of directory, the subdirectories were
2440            skipped unless they satisfied the include/exclude conditions. This is
2441            inconsistent with GNU grep (and could even be seen as contrary to the
2442            pcregrep specification - which I improved to make it absolutely clear).
2443            The action now is always to scan all levels of directory, and just
2444            apply the include/exclude patterns to regular files.
2445    
2446    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2447        --exclude_dir in the tests to avoid scanning .svn directories.
2448    
2449    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2450        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2451        doesn't support NULs in patterns.
2452    
2453    7.  Added some missing "const"s to declarations of static tables in
2454        pcre_compile.c and pcre_dfa_exec.c.
2455    
2456    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2457        caused by fix #2  above. (Subsequently also a second patch to fix the
2458        first patch. And a third patch - this was a messy problem.)
2459    
2460    9.  Applied Craig's patch to remove the use of push_back().
2461    
2462    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2463        matching function regexec().
2464    
2465    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2466        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2467        references. PCRE supports relative numbers with this syntax (I don't think
2468        Oniguruma does).
2469    
2470    12. Previously, a group with a zero repeat such as (...){0} was completely
2471        omitted from the compiled regex. However, this means that if the group
2472        was called as a subroutine from elsewhere in the pattern, things went wrong
2473        (an internal error was given). Such groups are now left in the compiled
2474        pattern, with a new opcode that causes them to be skipped at execution
2475        time.
2476    
2477    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2478        to the way PCRE behaves:
2479    
2480        (a) A lone ] character is dis-allowed (Perl treats it as data).
2481    
2482        (b) A back reference to an unmatched subpattern matches an empty string
2483            (Perl fails the current match path).
2484    
2485        (c) A data ] in a character class must be notated as \] because if the
2486            first data character in a class is ], it defines an empty class. (In
2487            Perl it is not possible to have an empty class.) The empty class []
2488            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2489            The negative empty class [^] matches any one character, independently
2490            of the DOTALL setting.
2491    
2492    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2493        non-existent subpattern following a character class starting with ']' and
2494        containing () gave an internal compiling error instead of "reference to
2495        non-existent subpattern". Fortunately, when the pattern did exist, the
2496        compiled code was correct. (When scanning forwards to check for the
2497        existence of the subpattern, it was treating the data ']' as terminating
2498        the class, so got the count wrong. When actually compiling, the reference
2499        was subsequently set up correctly.)
2500    
2501    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2502        it was being rejected as not supported by pcre_dfa_exec(), even though
2503        other assertions are supported. I have made pcre_dfa_exec() support
2504        (*FAIL).
2505    
2506    16. The implementation of 13c above involved the invention of a new opcode,
2507        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2508        cannot be changed at match time, I realized I could make a small
2509        improvement to matching performance by compiling OP_ALLANY instead of
2510        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2511        on the OP_ANY path.
2512    
2513    17. Compiling pcretest on Windows with readline support failed without the
2514        following two fixes: (1) Make the unistd.h include conditional on
2515        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2516    
2517    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2518        ncurses library to be included for pcretest when ReadLine support is
2519        requested, but also to allow for it to be overridden. This patch came from
2520        Daniel Bergström.
2521    
2522    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2523        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2524        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2525        spotting this.
2526    
2527    
2528    Version 7.6 28-Jan-08
2529    ---------------------
2530    
2531    1.  A character class containing a very large number of characters with
2532        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2533        overflow.
2534    
2535    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2536        HAVE_LONG_LONG is not defined.
2537    
2538    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2539        bring them up to date and include new features. This patch includes:
2540    
2541        - Fixed PH's badly added libz and libbz2 support.
2542        - Fixed a problem with static linking.
2543        - Added pcredemo. [But later removed - see 7 below.]
2544        - Fixed dftables problem and added an option.
2545        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2546            HAVE_LONG_LONG.
2547        - Added readline support for pcretest.
2548        - Added an listing of the option settings after cmake has run.
2549    
2550    4.  A user submitted a patch to Makefile that makes it easy to create
2551        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2552        Makefile.am that cause it to include this special target, without
2553        affecting anything else. Note that the same mingw target plus all
2554        the other distribution libraries and programs are now supported
2555        when configuring with CMake (see 6 below) instead of with
2556        Configure/Make.
2557    
2558    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2559        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2560        exported in the Windows port". It has not yet been confirmed that the patch
2561        solves the problem, but it does no harm.
2562    
2563    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2564        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2565        with CMake, and also correct the comment about stack recursion.
2566    
2567    7.  Remove the automatic building of pcredemo from the ./configure system and
2568        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2569        of a program that users should build themselves after PCRE is installed, so
2570        building it automatically is not really right. What is more, it gave
2571        trouble in some build environments.
2572    
2573    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2574    
2575    
2576    Version 7.5 10-Jan-08
2577    ---------------------
2578    
2579    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2580        values in parens when parsing an RE using the C++ wrapper."
2581    
2582    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2583        Characters greater than 255 were excluded from the class instead of being
2584        included.
2585    
2586    3.  The same bug as (2) above applied to negated POSIX classes such as
2587        [:^space:].
2588    
2589    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2590        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2591        I have changed it.
2592    
2593    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2594        first named subpattern) and a construct such as (?&a) would reference the
2595        first named subpattern whose name started with "a" (in other words, the
2596        length check was missing). Both these problems are fixed. "Subpattern name
2597        expected" is now given for (?&) (a zero-length name), and this patch also
2598        makes it give the same error for \k'' (previously it complained that that
2599        was a reference to a non-existent subpattern).
2600    
2601    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2602        this is right because (?- can be followed by option settings as well as by
2603        digits. I have, however, made the messages clearer.
2604    
2605    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2606        than the number used in the conditional) now cause a compile-time error.
2607        This is actually not compatible with Perl, which accepts such patterns, but
2608        treats the conditional as always being FALSE (as PCRE used to), but it
2609        seems to me that giving a diagnostic is better.
2610    
2611    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2612        and messages.
2613    
2614    9.  Fix two occurrences of "backslash" in comments that should have been
2615        "backspace".
2616    
2617    10. Remove two redundant lines of code that can never be obeyed (their function
2618        was moved elsewhere).
2619    
2620    11. The program that makes PCRE's Unicode character property table had a bug
2621        which caused it to generate incorrect table entries for sequences of
2622        characters that have the same character type, but are in different scripts.
2623        It amalgamated them into a single range, with the script of the first of
2624        them. In other words, some characters were in the wrong script. There were
2625        thirteen such cases, affecting characters in the following ranges:
2626    
2627          U+002b0 - U+002c1
2628          U+0060c - U+0060d
2629          U+0061e - U+00612
2630          U+0064b - U+0065e
2631          U+0074d - U+0076d
2632          U+01800 - U+01805
2633          U+01d00 - U+01d77
2634          U+01d9b - U+01dbf
2635          U+0200b - U+0200f
2636          U+030fc - U+030fe
2637          U+03260 - U+0327f
2638          U+0fb46 - U+0fbb1
2639          U+10450 - U+1049d
2640    
2641    12. The -o option (show only the matching part of a line) for pcregrep was not
2642        compatible with GNU grep in that, if there was more than one match in a
2643        line, it showed only the first of them. It now behaves in the same way as
2644        GNU grep.
2645    
2646    13. If the -o and -v options were combined for pcregrep, it printed a blank
2647        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2648        does the same. The return code can be used to tell if there were any
2649        non-matching lines.
2650    
2651    14. Added --file-offsets and --line-offsets to pcregrep.
2652    
2653    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2654        infinitely looping recursion. The bug was that positive lookaheads were not
2655        being skipped when checking for a possible empty match (negative lookaheads
2656        and both kinds of lookbehind were skipped).
2657    
2658    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2659        inclusion of <windows.h> to before rather than after the definition of
2660        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2661    
2662    17. Specifying a possessive quantifier with a specific limit for a Unicode
2663        character property caused pcre_compile() to compile bad code, which led at
2664        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2665        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2666        caused the error; without that there was no problem.
2667    
2668    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2669    
2670    19. Added --enable-pcretest-libreadline.
2671    
2672    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2673        RE::GlobalReplace(). As a result, the number of replacements returned was
2674        double what it should be. I removed one of the increments, but Craig sent a
2675        later patch that removed the other one (the right fix) and added unit tests
2676        that check the return values (which was not done before).
2677    
2678    21. Several CMake things:
2679    
2680        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2681            the names libpcre and libpcreposix, not just pcre and pcreposix.
2682    
2683        (2) The above change means that pcretest and pcregrep are now correctly
2684            linked with the newly-built libraries, not previously installed ones.
2685    
2686        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2687    
2688    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2689        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2690        UTF-8 newline character). The key issue is that the pattern starts .*;
2691        this means that the match must be either at the beginning, or after a
2692        newline. The bug was in the code for advancing after a failed match and
2693        checking that the new position followed a newline. It was not taking
2694        account of UTF-8 characters correctly.
2695    
2696    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2697        character classes. PCRE was not treating the sequence [:...:] as a
2698        character class unless the ... were all letters. Perl, however, seems to
2699        allow any characters between [: and :], though of course it rejects as
2700        unknown any "names" that contain non-letters, because all the known class
2701        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2702        for example, whereas PCRE did not - it did not recognize a POSIX character
2703        class. This seemed a bit dangerous, so the code has been changed to be
2704        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2705        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2706        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2707        Perl does, and where it didn't before.
2708    
2709    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2710        Windows environments %n is disabled by default.
2711    
2712    
2713    Version 7.4 21-Sep-07
2714    ---------------------
2715    
2716    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2717        means that a class such as [\s] counted as "explicit reference to CR or
2718        LF". That isn't really right - the whole point of the change was to try to
2719        help when there was an actual mention of one of the two characters. So now
2720        the change happens only if \r or \n (or a literal CR or LF) character is
2721        encountered.
2722    
2723    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2724        of both had grown to the point where there were only 3 bits left.
2725        Fortunately, there was spare space in the data structure, and so I have
2726        moved the internal flags into a new 16-bit field to free up more option
2727        bits.
2728    
2729    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2730        but did not set the internal JCHANGED flag - either of these is enough to
2731        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2732        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2733        start sets both bits.
2734    
2735    4.  Added options (at build time, compile time, exec time) to change \R from
2736        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2737    
2738    5.  doc/pcresyntax.html was missing from the distribution.
2739    
2740    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2741        compatibility, even though it is no longer used.
2742    
2743    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2744        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2745        windows.h file is present (where different names are used). [This was
2746        reversed later after testing - see 16 below.]
2747    
2748    8.  Changed all #include <config.h> to #include "config.h". There were also
2749        some further <pcre.h> cases that I changed to "pcre.h".
2750    
2751    9.  When pcregrep was used with the --colour option, it missed the line ending
2752        sequence off the lines that it output.
2753    
2754    10. It was pointed out to me that arrays of string pointers cause lots of
2755        relocations when a shared library is dynamically loaded. A technique of
2756        using a single long string with a table of offsets can drastically reduce
2757        these. I have refactored PCRE in four places to do this. The result is
2758        dramatic:
2759    
2760          Originally:                          290
2761          After changing UCP table:            187
2762          After changing error message table:   43
2763          After changing table of "verbs"       36
2764          After changing table of Posix names   22
2765    
2766        Thanks to the folks working on Gregex for glib for this insight.
2767    
2768    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2769        unicode-properties was also set.
2770    
2771    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2772    
2773    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2774        checked only for CRLF.
2775    
2776    14. Added casts to pcretest.c to avoid compiler warnings.
2777    
2778    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2779    
2780    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2781        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2782        entirely. This removes changes made in 7 above.
2783    
2784    17. The CMake files have been updated, and there is now more information about
2785        building with CMake in the NON-UNIX-USE document.
2786    
2787    
2788    Version 7.3 28-Aug-07
2789    ---------------------
2790    
2791     1. In the rejigging of the build system that eventually resulted in 7.1, the
2792        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2793        brackets there is not right, since it causes compilers to look for an
2794        installed pcre.h, not the version that is in the source that is being
2795        compiled (which of course may be different). I have changed it back to:
2796    
2797          #include "pcre.h"
2798    
2799        I have a vague recollection that the change was concerned with compiling in
2800        different directories, but in the new build system, that is taken care of
2801        by the VPATH setting the Makefile.
2802    
2803     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2804        when the subject happened to end in the byte 0x85 (e.g. if the last
2805        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2806        characters but of course it shouldn't be taken as a newline when it is part
2807        of another character. The bug was that, for an unlimited repeat of . in
2808        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2809        characters when looking for a newline.
2810    
2811     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2812    
2813     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2814        in debug output.
2815    
2816     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2817        long printing in the pcrecpp unittest when running under MinGW.
2818    
2819     6. ESC_K was left out of the EBCDIC table.
2820    
2821     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2822        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2823        limit also applies to "virtual nesting" when a pattern is recursive, and in
2824        this case 1000 isn't so big. I have been able to remove this limit at the
2825        expense of backing off one optimization in certain circumstances. Normally,
2826        when pcre_exec() would call its internal match() function recursively and
2827        immediately return the result unconditionally, it uses a "tail recursion"
2828        feature to save stack. However, when a subpattern that can match an empty
2829        string has an unlimited repetition quantifier, it no longer makes this
2830        optimization. That gives it a stack frame in which to save the data for
2831        checking that an empty string has been matched. Previously this was taken
2832        from the 1000-entry workspace that had been reserved. So now there is no
2833        explicit limit, but more stack is used.
2834    
2835     8. Applied Daniel's patches to solve problems with the import/export magic
2836        syntax that is required for Windows, and which was going wrong for the
2837        pcreposix and pcrecpp parts of the library. These were overlooked when this
2838        problem was solved for the main library.
2839    
2840     9. There were some crude static tests to avoid integer overflow when computing
2841        the size of patterns that contain repeated groups with explicit upper
2842        limits. As the maximum quantifier is 65535, the maximum group length was
2843        set at 30,000 so that the product of these two numbers did not overflow a
2844        32-bit integer. However, it turns out that people want to use groups that
2845        are longer than 30,000 bytes (though not repeat them that many times).
2846        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2847        made it possible to implement the integer overflow checks in a much more
2848        dynamic way, which I have now done. The artificial limitation on group
2849        length has been removed - we now have only the limit on the total length of
2850        the compiled pattern, which depends on the LINK_SIZE setting.
2851    
2852    10. Fixed a bug in the documentation for get/copy named substring when
2853        duplicate names are permitted. If none of the named substrings are set, the
2854        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2855        empty string.
2856    
2857    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2858        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2859        because the ] is interpreted as the first data character and the
2860        terminating ] is not found. PCRE has been made compatible with Perl in this
2861        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2862        cause memory overwriting.
2863    
2864    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2865        string has been matched (to stop an infinite loop). It was not recognizing
2866        a conditional subpattern that could match an empty string if that
2867        subpattern was within another subpattern. For example, it looped when
2868        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2869        condition was not nested. This bug has been fixed.
2870    
2871    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2872        past the start of the subject in the presence of bytes with the top bit
2873        set, for example "\x8aBCD".
2874    
2875    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2876        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2877    
2878    14. Optimized (?!) to (*FAIL).
2879    
2880    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2881        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2882        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2883        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2884        does: it's just the validity check that is more restrictive.
2885    
2886    16. Inserted checks for integer overflows during escape sequence (backslash)
2887        processing, and also fixed erroneous offset values for syntax errors during
2888        backslash processing.
2889    
2890    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2891        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2892    
2893    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2894        caused an overrun.
2895    
2896    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2897        something other than just ASCII characters) inside a group that had an
2898        unlimited repeat caused a loop at compile time (while checking to see
2899        whether the group could match an empty string).
2900    
2901    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2902        [\P{Any}] did so. (Error in the code for printing property names.)
2903    
2904    21. An orphan \E inside a character class could cause a crash.
2905    
2906    22. A repeated capturing bracket such as (A)? could cause a wild memory
2907        reference during compilation.
2908    
2909    23. There are several functions in pcre_compile() that scan along a compiled
2910        expression for various reasons (e.g. to see if it's fixed length for look
2911        behind). There were bugs in these functions when a repeated \p or \P was
2912        present in the pattern. These operators have additional parameters compared
2913        with \d, etc, and these were not being taken into account when moving along
2914        the compiled data. Specifically:
2915    
2916        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2917            length.
2918    
2919        (b) An item such as \pL+ within a repeated group could cause crashes or
2920            loops.
2921    
2922        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2923            "reference to non-existent subpattern" error.
2924    
2925        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2926    
2927    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2928        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2929    
2930    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2931    
2932    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2933        character were causing crashes (broken optimization).
2934    
2935    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2936        \p or \P) caused a compile-time loop.
2937    
2938    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2939        break. For example, the unstudied pattern [\r\n]A does not match the string
2940        "\r\nA" because change 7.0/46 below moves the current point on by two
2941        characters after failing to match at the start. However, the pattern \nA
2942        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2943        the same is true. There doesn't seem any very clean way out of this, but
2944        what I have chosen to do makes the common cases work: PCRE now takes note
2945        of whether there can be an explicit match for \r or \n anywhere in the
2946        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2947        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2948        pattern has explicit CR or LF references.
2949    
2950    29. Added (*CR) etc for changing newline setting at start of pattern.
2951    
2952    
2953    Version 7.2 19-Jun-07
2954  ---------------------  ---------------------
2955    
2956   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 2970  Version 7.2 01-May-07
2970      stack recursion. This gives a massive performance boost under BSD, but just      stack recursion. This gives a massive performance boost under BSD, but just
2971      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
2972      in all cases.      in all cases.
2973    
2974   6. Added more features from the forthcoming Perl 5.10:   6. Added more features from the forthcoming Perl 5.10:
2975    
2976      (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
2977          recursion call. It refers to the nth most recently opened parentheses.          recursion call. It refers to the nth most recently opened parentheses.
2978    
2979      (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
2980          to be opened parentheses.          to be opened parentheses.
2981    
2982      (c) Conditions that refer to capturing parentheses can be specified      (c) Conditions that refer to capturing parentheses can be specified
2983          relatively, for example, (?(-2)... or (?(+3)...          relatively, for example, (?(-2)... or (?(+3)...
2984    
2985        (d) \K resets the start of the current match so that everything before
2986            is not part of it.
2987    
2988        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2989    
2990        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2991            reference syntax.
2992    
2993        (g) (?| introduces a group in which the numbering of parentheses in each
2994            alternative starts with the same number.
2995    
2996        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2997    
2998     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2999        PCRE_INFO_JCHANGED.
3000    
3001     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
3002        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
3003        for detecting groups that can match an empty string.
3004    
3005     9. A pattern with a very large number of alternatives (more than several
3006        hundred) was running out of internal workspace during the pre-compile
3007        phase, where pcre_compile() figures out how much memory will be needed. A
3008        bit of new cunning has reduced the workspace needed for groups with
3009        alternatives. The 1000-alternative test pattern now uses 12 bytes of
3010        workspace instead of running out of the 4096 that are available.
3011    
3012    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
3013    
3014    11. Applied patch from Google to remove an optimization that didn't quite work.
3015        The report of the bug said:
3016    
3017          pcrecpp::RE("a*").FullMatch("aaa") matches, while
3018          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
3019          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
3020    
3021    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
3022        it matched the wrong number of bytes.
3023    
3024    
3025  Version 7.1 24-Apr-07  Version 7.1 24-Apr-07
3026  ---------------------  ---------------------

Legend:
Removed from v.167  
changed lines
  Added in v.1543

  ViewVC Help
Powered by ViewVC 1.1.5