perldelta for #118297
authorFather Chrysostomos <sprout@cpan.org>
Tue, 18 Jun 2013 15:26:12 +0000 (08:26 -0700)
committerFather Chrysostomos <sprout@cpan.org>
Tue, 18 Jun 2013 15:28:35 +0000 (08:28 -0700)
copied from f4612955aa389 on the maint-5.18 branch

pod/perldelta.pod

index 82c2b74..06a276e 100644 (file)
@@ -654,6 +654,19 @@ by the C<syntax> category.  On the other hand, the C<exec> category
 controlled its fatal-ness.  It is now entirely handled by the C<exec>
 category.
 
+=item *
+
+Perl v5.18 inadvertently introduced a bug whereby interpolating mixed up-
+and down-graded UTF-8 strings in a regex could result in malformed UTF-8
+in the pattern: specifically if a downgraded character in the range
+C<\x80..\xff> followed a UTF-8 string, e.g.
+
+    utf8::upgrade(  my $u = "\x{e5}");
+    utf8::downgrade(my $d = "\x{e5}");
+    /$u$d/
+
+[RT #118297]
+
 =back
 
 =head1 Known Problems