> .. correct, you need a per-language table, which (as far as I know) > does not actually exist.
It does, it's SpecialCasing.txt in the Unicode Character Database (http://www.unicode.org/Public/6.0.0/ucd/SpecialCasing.txt); see sections 3.13 and 5.18 of the Unicode Standard for more details (http://www.unicode.org/versions/Unicode6.0.0/{ch03,ch05}.pdf). It actually has much more information than language-specific casing exceptions, of which there are rather few (mostly for Turkish and Azeri, and Lithuanian). Arthur
