algorithm to autodetect (japanese) encodings..

[Skip Montanaro]

    Andy> Nearly -) Get hold of Tamito Kajiyama's excellent Japanese codecs,
    Andy> which plug into Pythoin's unicode conversion system.  You can just
    Andy> try converting from each (Shift-JIS, EUC, UTF8) to Unicode in
    Andy> turn, ad see which raise exceptions.

Wow...  They are all that different that you're likely to only get a
successful conversion when you select the correct encoding?  I wish the
various Latin-1 variants were that distinct. ;-)

Skip