lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Shai Erera (JIRA)" <>
Subject [jira] Commented: (LUCENE-1593) Optimizations to TopScoreDocCollector and TopFieldCollector
Date Sun, 12 Apr 2009 20:46:14 GMT


Shai Erera commented on LUCENE-1593:

This sounds like it should work, but since I'm not fully familiar with that code, I can't
back you up :). I guess a test case will clarify it. In the meantime, I read BooleanScorer
and BooleanScorer2 code, and came across several possible optimizations:
* BS.score() and score(HC) check if coordFactors is null on every call (this actually is problematic
for score() only). I think we can init coordFactors in the ctor, as well as after every call
to add(Scorer)? That is not called during query execution, but score() is.
* Same for BS2.score() and score(HC) - initCountingSumScorer?
* Cleanup BS.add() code a bit. For example, it checks 'if (prohibited)', does something and
then 'if (!prohibited)'. Maybe merge all the required/prohibited/both cases together?
* BS2 declares 'defaultSimilarity' and instantiates it to new DefaultSimilarity(). Two things
here: (1) can the field be defined final (2) use Similarity.getDefault()?
* BS2.SingleMatchScorer's score() method looks a bit suspicious. It checks if doc() >=
lastScoredDoc and if so updates lastScoredDoc and increments coordinator.nrMatchers. It then
calls scorer.score() regardless. So it looks as if this method is expected to be called for
the same doc several times. When 1575 is committed, I think we should wrap the input scorer
with the new ScoreCachingWrappingScorer so that the actual score will not be computed over
and over? Also, I think doc() needs to be saved in a local variable.

> Optimizations to TopScoreDocCollector and TopFieldCollector
> -----------------------------------------------------------
>                 Key: LUCENE-1593
>                 URL:
>             Project: Lucene - Java
>          Issue Type: Improvement
>          Components: Search
>            Reporter: Shai Erera
>             Fix For: 2.9
> This is a spin-off of LUCENE-1575 and proposes to optimize TSDC and TFC code to remove
unnecessary checks. The plan is:
> # Ensure that IndexSearcher returns segements in increasing doc Id order, instead of
> # Change TSDC and TFC's code to not use the doc id as a tie breaker. New docs will always
have larger ids and therefore cannot compete.
> # Pre-populate HitQueue with sentinel values in TSDC (score = Float.NEG_INF) and remove
the check if reusableSD == null.
> # Also move to use "changing top" and then call adjustTop(), in case we update the queue.
> # some methods in Sort explicitly add SortField.FIELD_DOC as a "tie breaker" for the
last SortField. But, doing so should not be necessary (since we already break ties by docID),
and is in fact less efficient (once the above optimization is in).
> # Investigate PQ - can we deprecate insert() and have only insertWithOverflow()? Add
a addDummyObjects method which will populate the queue without "arranging" it, just store
the objects in the array (this can be used to pre-populate sentinel values)?
> I will post a patch as well as some perf measurements as soon as I have them.

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message