The patches are to convert UTF-8 to an 8-bit encoding, which doesn't work -- the real problem goes much deeper since Perl::Tidy breaks down line input into bytes and processes things byte-by-byte (using maps for the "known 256 chars).... Such maps don't scale so well when you move up to the number of chars in Unicode. It looked to be close to a rewrite of the core code.

It's been an open bug since 2008 and doesn't look easy to fix so it hasn't been.

Expect to see it handle unicode sometime right after perlmonks does. ;-)