This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
import perldelta from eb83ed8 into release branch
authorRicardo Signes <rjbs@cpan.org>
Thu, 10 May 2012 13:38:02 +0000 (09:38 -0400)
committerRicardo Signes <rjbs@cpan.org>
Thu, 10 May 2012 15:11:42 +0000 (11:11 -0400)
pod/perldelta.pod

index b355569..535604b 100644 (file)
@@ -1823,6 +1823,16 @@ string (0 is the return), and overlong UTF-8 sequences.
 
 =item *
 
+The returned code point from the function C<utf8n_to_uvuni()>
+when the input is malformed UTF-8, malformations are allowed, and
+C<utf8> warnings are off is now the Unicode REPLACEMENT CHARACTER
+whenever the malformation is such that no well-defined code point can be
+computed.  Previously the returned value was essentially garbage.  The
+only malformations that have well-defined values are a zero-length
+string (0 is the return), and overlong UTF-8 sequences.
+
+=item *
+
 Padlists are now marked C<AvREAL>; i.e., reference-counted.  They have
 always been reference-counted, but were not marked real, because F<pad.c>
 did its own clean-up, instead of using the usual clean-up code in F<sv.c>.