
26 Jan
2010
26 Jan
'10
8:11 a.m.
Matthew Phillips-5 wrote:
I also found it to to be very slow.
My variant: http://a-ejeli-tak.livejournal.com/1326.html Spellchecker in Haskell String version runs in 2.5 sec, ByteString in 1.2 sec (just for one word e.g. just to build the tree). 8 sec to check input of 400 words (copied from Norvig's example). I think laziness helps here to avoid unnecessary checks (once the first match is found). Haven't tried it on a larger data sets neither tried to optimize it. Cheated on dictionary parsing though... -- View this message in context: http://old.nabble.com/Spelling-checker-exercise-tp27269320p27322382.html Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.