#214 wrong encoding used for unrecognized words

[Arno Teigseth]

nb_NO uses ISO-8859-1 encoding for its .dic file.
I use utf8 for my terminal.

If run as "hunspell -d nb_NO -m" and checking a word with hunspell, all is ok if it's a recognized word:
øyelege
øyelege st:øyelege po:noun:m

But when writing an unrecognized word it's messing up:
øvingsteam
øvingsteam