in reply to Unicode words match and catch
Wouldn't HTML::Entities fit the bill already, without the recognition of the particular alphabets?