[KinoSearch] Possible Phrase Query Bug

[Nathan Kurz]

> What I actually committed was slightly different, and attempts to
> explain the algorithm more effectively:

I checked out the current version, and what you did makes sense.   You
might consider moving the new check to the outer loop, though, since
you don't have to check the initial conditions for every potential
anchor.  But apart from a couple extra compares, your version seems
fine.

> function-call overhead.  I'm not certain whether that was the right
> trade to make historically, but I am nearly certain that there's no
> way to do what we want to do with expanded position scoring without
> keeping all positions loaded.

I think you are right, and your version makes much more sense to me
than the Lucene version you describe.  One needs full access to the
positions.  My instinct involves pushing things even farther in the
direction you took.

Nathan Kurz
nate at verse.com

_______________________________________________
KinoSearch mailing list
KinoSearch at rectangular.com
http://www.rectangular.com/mailman/listinfo/kinosearch