Good to hear its kind of working for you!

But there will be a correct way for handling Swedish and Finnish unicode collation, so I wouldn't start switching between sort and collate in such cases until you've exhausted the "correct" way(s?) of doing this.

Perhaps you could ask SADAHIRO Tomoyuki, the Unicode::Collate author?

It would certainly be good if you can post any reply you get here, since this is the kind of stuff perl developers are going to have to know more and more about - I'm of course talking about the folks who don't know it already :)