[
https://issues.apache.org/jira/browse/LUCENE-8038?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16240182#comment-16240182
]
Alan Woodward commented on LUCENE-8038:
---------------------------------------
bq. It might be useful to let PayloadDecoder return null for missing payloads
I think I'd rather go back to the lucene 5.0 behaviour of returning 1.0 - this
is all being done within the scoring loop, so we want to keep boxing and object
creation to an absolute minimum.
bq. But what is the story on more complex custom payload functions?
There's definitely more to do here! But yes, let's defer that to a later
issue. This one is mainly to enable LUCENE-8014.
> Decouple payload decoding from Similarity
> -----------------------------------------
>
> Key: LUCENE-8038
> URL: https://issues.apache.org/jira/browse/LUCENE-8038
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Alan Woodward
> Assignee: Alan Woodward
> Attachments: LUCENE-8038-master.patch, LUCENE-8038.patch
>
>
> PayloadScoreQuery is the only place that currently uses
> SimScorer.computePayloadFactor(), and as discussed on LUCENE-8014, this seems
> like the wrong place for it. We should instead add a PayloadDecoder
> abstraction that is passed to PayloadScoreQuery.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
