From: Erik Veenstra Date: 2008-05-15T20:24:42+09:00 Subject: Re: 40 million levenshtein distances for two long strings > I have to do a lot of comparisons, about 40 million. 40 million or 400 million?... :} I've once written a Levenshtein implementation in C, because of this speed issue. It compares 2 20 KB files in 3 seconds. And it does come with a threshold. Interested? gegroet, Erik V. - http://www.erikveen.dds.nl/ ---------------------------------------------------------------- $ ll lev[12] -rw-r--r-- 1 erik erik 19275 May 15 12:10 lev1 -rw-r--r-- 1 erik erik 19275 May 15 12:12 lev2 $ cat lev.rb require "ev/levenshtein" s1 = File.read("lev1") s2 = File.read("lev2") p EV.levenshtein(s1, s2) $ time ruby lev.rb 12 real 0m3.105s user 0m2.843s sys 0m0.281s ----------------------------------------------------------------