Re: UTF-8 and browsers - Update

[dakkar]

Bug in Firefox. It should work as you describe.

As for the composition: first of all, work on characters, or at least or codepoints, not on utf-8 bytes. Second, you want Unicode Normal Form C (see Unicode::Normalize), so that you can write:

use Unicode::Normalize;
use charnames ':full'; # this is just to make things easier in this ex
+ample

binmode(STDOUT,':utf8'); # this to make 'print' output utf-8 bytes

my $a="O\N{COMBINING DIAERESIS}";
my $b=NFC($a);

print length($a),$a,"\n";
print length($b),$b,"\n";

Will print:

2Ö
1Ö

(more or less, depending on PM's escaping mechanisms)