When I run your code from Komodo IDE which understands UTF8 it prints:

Content-Type:text/html; charset=utf-8
Content-Language: utf8;

<p>ä    <span class="sy">ä</span>, <span class="sy">Ä</span><span clas
+s="posg pos">Substantiv, Neutrum, das</span><span class="vg v"> Ä  &#
+603;&#720; das ä; Genitiv: des ä (umgangssprachlich: -s), ä (umgangss
+prachlich: -s) </span></p>
[download]

Is that not what you expected to see? Maybe the terminal you are using doesn't understand UTF8?

Update: note that you don't need use utf8;. That is only required if you want to use UTF8 in your source code. You don't do that, you create a string containing UTF8 characters, but the source code is pure 7 bit ASCII.

Yes, my code does send that to the browser, as I noted above. I just haven't found a way to send the converted text to a file. Unfortunately, the file sizes I'm working with are well above 100 MB, and it's simply not practical to try to run it all through the browser, click on "view source," copy it out, and paste it into a text file. In fact, even TextWrangler chokes on these file sizes already. I'm trying to convert them before pumping them into a database.

Blessings,

~Polyglot~

What is actually stored in your files? The literal text you provided in the string, or something else? If it is the text in the string then you can:

use strict;
use warnings;
use Encode;

binmode STDOUT, ':utf8';

print "Content-Type:text/html; charset=utf-8\n";
print "Content-Language: utf8;\n\n";

my $asText = do {local $/; <DATA>};

$asText =~ s!\\x(..)!chr(hex($1))!ge;

my $uCode;
my $newcode = decode('utf8', $asText);

print "<p>$newcode</p>\n";

__DATA__
\xc3\xa4    <span class="sy">\xc3\xa4</span>,
<span class="sy">\xc3\x84</span>
<span class="posg pos">Substantiv, Neutrum, das</span>
<span class="vg v"> \xc3\x84  \xc9\x9b\xcb\x90 das \xc3\xa4; Genitiv: 
+des \xc3\xa4 (umgangssprachlich: -s), \xc3\xa4 (umgangssprachlich: -s
+) </span>
[download]

Prints:

Content-Type:text/html; charset=utf-8
Content-Language: utf8;

<p>ä    <span class="sy">ä</span>,
<span class="sy">Ä</span>
<span class="posg pos">Substantiv, Neutrum, das</span>
<span class="vg v"> Ä  &#603;&#720; das ä; Genitiv: des ä (umgangsspra
+chlich: -s), ä (umgangssprachlich: -s) </span>
</p>
[download]

Optimising for fewest key strokes only makes sense transmitting to Pluto or beyond

The question is how \xc3\xa4 is actually represented in your file.

If you write qq|\xc3\xa4|, then Perl interprets the eight characters as two bytes with the hexadecimal values of c3 and a4, respectively. But if you read \xc3\xa4 from a file, this interpretation doesn't take place: These are eight individual ASCII characters. What you can do, of course, is do the interpretation yourself:

use Encode;
my $ucode = q/\xc3\xa4/; # note the use of 'q', not 'qq'
my $newcode = decode('utf8',$ucode =~ s/\\x([a-fA-F0-9]{2})/chr hex($1
+)/egr);
[download]

Yes, it appears you have the same solution that GrandFather posted, but have managed to crystallize it into a one-liner. Thank you. I'm sure future readers will appreciate this additional clarity.

NOTE: Yes, the file does have those literal characters, and they were being processed as eight ascii characters by Perl.

Blessings,

~Polyglot~