I am trying to pack bytes into as few unicode characters as possible. Consider the Twitter 140 character limit, which allows byte-heavy unicode characters. I believe I can squeeze 3 8-bit ascii characters into a single 20-bit unicode character, which I am currently trying via the following:

$\="\n";
$_='Hello World!'; print;
$_=pack'(U)*',map{hex}unpack'(H5)*'; print;
$_=pack'(H5)*',map{sprintf"%05X",$_}unpack'(U)*'; print;
[download]

Hello World!
񈙖񬛲񗛷񬙂
He`lo Wopld

Note that when I attempt to pack'(H5)*' the same data I map/unpacked, (almost*) every third character is garbled, a symptom of dealing with an odd 5 hex characters, I'm sure. What is the appropriate way to do this without losing bits?

Packing/Unpacking with the normal 16-bits works as expected:

$\="\n";
$_='Hello World!'; print;
$_=pack'(U)*',map{hex}unpack'(H4)*'; print;
$_=pack'(H4)*',map{sprintf"%04X",$_}unpack'(U)*'; print;
[download]

Hello World!
䡥汬漠坯牬搡
Hello World!

...and even allows 'n*' packing, since n is a 16-bit unsigned short:

$\="\n";
$_='Hello World!'; print;
$_=pack'(U)*',map{hex}unpack'(H4)*'; print;
$_=pack'n*',unpack'(U)*'; print
[download]

Hello World!
䡥汬漠坯牬搡
Hello World!

Thanks for the help, and Cheers!