[eluser]pickupman[/eluser]
Good point Jelmer. I tried to plug your library above. I haven't used it yet though. I was testing this out on a data set that I had. I was surprised by the results. Using the MATCH...AGAINST table where there was about 100 rows of similar data, yielded nothing because there would be too many duplicates. However, I ran another test on a table of 42k US zipcodes searching by city name. The column is index, and using the LIKE %city_name% returned in .07 secs. Using MATCH(city) AGAINST('city_name') returned in .004 secs. About 40% faster. I guess it depends on your data.
