Hi again :)
 
  Hits versus HitCollector

  I implemented HitCollector, but it didn't seem much faster. I'm 
guessing its because Hits doesn't seem to run all the way through the 
resultset before finishing whereas HitsCollector does ?

  So, What I'd really like to know, is, does a Hits combined with 
HitCollector make sense ?

  Looking at the code, there seems to be a magical getMoreDocs, which 
calls a private method  that returns the topN results (without 
enumerating the whole list?)
 
  I also think I need to be able to externalise the whole Document 
concept -- I don't need the data stored in the Index -- just the 
DocID would suffice, then I could reference this from a big array (I 
have a couple of Gigs of RAM to play with).


I'm down to about 200msecs response time for 8 million short records. 
I need it below 50 (ideally around 5 :)).

  Cheers,
   Winton

Winton Davies
Lead Engineer, Overture (NSDQ: OVER)
1820 Gateway Drive, Suite 360
San Mateo, CA 94404
work: (650) 403-2259
cell: (650) 867-1598
http://www.overture.com/

--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to