lars nygaard <[EMAIL PROTECTED]> writes:

> I'm having trouble getting unicode from apache(2.0.48) and perl (5.8.0)
> via LWP(5.76). Here's an illustrative cgi script:
>  
> print "Content-type: text/plain; charset=utf-8\n\n"; 
> print "Det bÃÂr gÃÂ";
> 
> And it prints nicely in Mozilla. However, this program:
>  
> my $ua = LWP::UserAgent->new;
> my $response = $ua->request(GET 'http://localhost/cgi-bin/test.cgi');
> my $content = $response->content;
> print $content;
> 
> prints "Det bÃÂÃÂr gÃÂÃÂ".

You end up with doubly encoded UTF-8 here.  What I think is going on
here is that you are running under some UTF-8 locale (default in
Redhat 8 and 9) and perl-5.8.0 will then auto UTF-8 encode strings
on STDOUT.  The command 'locale' will print out information about the
current locale.

Upgrading to perl-5.8.2 should actually fix this problem as it does
not go into UTF-8 mode for file handles by default.

Another way to to tell perl that content is utf8().  This can be
achieved with:

    use Encode qw(decode_utf8);

    ....
    my $content = decode_utf8($response->content);
    print $content;

Yet another way is to turn off the UTF-8 locale.  Try setting the LANG
environment variable to 'C' before you run your client program.

Regards,
Gisle

Reply via email to