Recently, I was stumped by "strange black boxes" appearing in a Win32 application textfield, whose contents are retrieved from a Unicode database field.
I used the following snippet to help me identify what was a "bad" mix sequence of <CR> and <LF> characters.
use Unicode::CharName qw/uname/;
use Unicode::UCD qw/charinfo/;
my $value = <what you are looking at>
my $c;
print "Unicode Charname:\tHexValue:\tCharCode\n";
for (split //, $value) {
$c = ord($_);
print uname($c) .":\t" . charinfo($c)->{code} . ":\t$c\n";
}