Literal strings

When dumping literal strings in your code, use the utf8 pragma when Unicode strings are desired and don’t use it or disable it when byte strings are desired. The pragma may also be lexically enabled or disabled.

    use utf8;
    {
        no utf8;
        say dump_hex('Ĝis! ☺');  # C4 9C 69 73 21 20 E2 98 BA
    }
    say dump_hex('Ĝis! ☺');  # 11C 69 73 21 20 263A

Command-line input and filehandles

The simplest way to ensure that you’re working with Unicode strings from all of your basic sources of input is to use the utf8::all pragma. This extends the utf8 pragma to automatically convert command-line arguments provided by @ARGV, user-defined filehandles, as well as standard filehandles like "STDIN", among others.

Other sources of input

To handle strings provided by other sources of input, such as from network protocols or web requests, decode the byte string to a Unicode string via Encode::decode using the source encoding name as the first argument.

    use Encode qw( decode );
    say dump_hex( decode('UTF-8', $string) );