[
https://issues.apache.org/jira/browse/LUCENE-6218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14305609#comment-14305609
]
Robert Muir commented on LUCENE-6218:
-------------------------------------
Well, i'm not sure merging docsenum and docsandpositionsenum is related. The
scoring api still has no concept of positions with that change, and there is no
optimization to implement in any of the scorers (yet), so I think we can defer
whether the boolean needs to be upgraded to an int, in other issues, when we
have real use cases to use it?
In this case we are just saying up-front that we don't need the score to speed
up queries we have today.
And we have a few use cases right now that can be sped up by the change, e.g.
sorting by something other than score, boolean MUST_NOT, constant score
queries, filters, etc. I only implemented the optimization for terms and exact
phrases, but we could do it for sloppy phrases, spans, etc too.
> don't decode freqs or enumerate all positions, when scores are not needed
> -------------------------------------------------------------------------
>
> Key: LUCENE-6218
> URL: https://issues.apache.org/jira/browse/LUCENE-6218
> Project: Lucene - Core
> Issue Type: Bug
> Reporter: Robert Muir
> Attachments: LUCENE-6218.patch
>
>
> Today if you don't call score() some things are faster, we won't invoke
> similarity or read the norm for the document or other things.
> On the other hand, its sad in this case that we are decompressing twice as
> many packed integers as we need (freqs can be skipped over, and our postings
> lists supports that) and walking all positions in phrase matching to
> determine the number of times the phrase matched (1 is enough, then we can
> stop).
> When scoring is not needed, things can be optimized in other cases too (e.g.
> thats the whole concept of filters).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
