Don't assume on encoding

[EvanCarroll]

Though the project we forked from assumes a specific encoding, we shouldn't. It's not safe to assume that LibreOffice is using anyone encoding. They have many.

This assumption won't hold

ENCODING="ISO8859-2"

pg_updatedicts does this right, they parse the files. Though ideally this would be fixed upstream. There is no reason for any of these to be anything but UTF-8.

LibreOffice/dictionaries#7