[KinoSearch] filtering search results

[Gavin Estey]

Marvin Humphrey wrote:

> At present, there are two approaches: one hack and one kludge.   
> There's the BitVector hack (which uses private API), and then there's  
> caching the field values in a Perl array.
>
I got the latter working, it seems to work nicely and reduces my test 
query from about 10 seconds to 1/10 of a second. My only issue now is 
building the cache in the first place, I can't seem to find a method of 
iterating over every doc in the index.

I don't understand the BitVector method, do you have an example of this?

Gavin.