in reply to Re^3: Rapid text searches ( O(1) space and time)
in thread Rapid text searches

I don't understand why you think I'm doing something different?

Oh, I don't really. The OP said "hash" and "single entry" so I felt like a Berkeley was the right answer (if the up front cost of building it is acceptable). Your solution is more flexible and really cool. I just thought my suggestion wasn't a "sledgehammer" but possibly a perfect fit. :)

Update: I'll see if I can post a 1,000 searches match too; I have to rebuild the thing so it might be later.

Update to the update: it was about 9 times slower to do 1,000 look-ups, clocking in at .15 for the most part (1.42 GHz G4 Mac).

  • Comment on Re^4: Rapid text searches ( O(1) space and time)