Question

I’m working on a CGI script that is called from a piece of software (that I can’t change). The variables submitted by the software are giving me problems, because if they contain non-ascii characters they look like this:

ÿFFFFDEetta er texti meÿFFFFF0 ÿFFFFEDslenskum stÿFFFFF6fum

instead of

Þetta er texti með íslenskum stöfum.

I’ve tried messing with the Encode::decode function but nothing has come of it – all I’ve gotten it to do is change how the ÿ gets represented.

So yeah, I’m kind of stumped. What do I do to change all the ÿFFFFDEs into Þs and so on, without resorting to replacing each non-ascii character individually (which is not a solution because this needs to work for languages I don’t even speak)?

Answer 1

use Encode qw(decode);
use Encode::Escape qw();

$_ = 'ÿFFFFDEetta er texti meÿFFFFF0 ÿFFFFEDslenskum stÿFFFFF6fum';
s/ÿFFFF/\\x/g;
decode('iso-8859-1', decode('unicode-escape', $_));
# returns 'Þetta er texti með íslenskum stöfum'