Or, or just install the
Unicode::Tussle package from CPAN. I forgot; that gets them all, including an OED2 version of
look
that knows about parts of speech and does O(log N) lookups on prefix matches very very quickly, does best-guess fuzzy matches, or full regexes. I think I call it
unilook. Rumour has it that includes the full OED2 word list, plus quarterly supplements since 2000.