[
https://issues.apache.org/jira/browse/LUCENE-7736?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15905422#comment-15905422
]
Adrien Grand commented on LUCENE-7736:
--------------------------------------
I like the termFreq and reader constant values sources, I can imagine them to
be useful to users. Can you add equals/methods to these new impls
(LUCENE-7723)? (I'm thinking that one easy way to make equals/hashcode work for
them could be to simply make them singletons)
bq. DoubleValuesSource (or LongValuesSource, in the case of norms)
Why are only norms a LongValuesSource? Should termFreq, docFreq, maxDoc and
numDocs be longs as well?
I'm curious whether you have a use-case for the {{joinDocFreq}} values sources,
it looks very expensive and a bit esoteric to me. I'm also unsure how useful
exposing raw norms is given that the value is an implementation detail of the
similarity?
bq. The IndexReaderFunctions class is currently in the queries module, but it
could easily be in core instead.
+1 to keep in queries for now
> Expose some IndexReader stats via DoubleValuesSources
> -----------------------------------------------------
>
> Key: LUCENE-7736
> URL: https://issues.apache.org/jira/browse/LUCENE-7736
> Project: Lucene - Core
> Issue Type: New Feature
> Reporter: Alan Woodward
> Assignee: Alan Woodward
> Priority: Minor
> Attachments: LUCENE-7736.patch
>
>
> We have a number of ValueSource implementations that expose IndexReader stats
> (numDocs, termFreq, etc). We should re-implement these as
> DoubleValuesSources, allowing them to be used in FunctionScoreQuery, etc.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
