[
https://issues.apache.org/jira/browse/LUCENE-8613?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16724039#comment-16724039
]
Tristan Stevens commented on LUCENE-8613:
-----------------------------------------
So I'm trying to avoid having to have a separate field as it's cumbersome and
error-prone to maintain. I think de-duplicated is fine, right? - if it's a set
it's a set. if it's a list it's a list. One would expect a set to be
de-duplicated.
I sort of made a start here:
https://github.com/apache/lucene-solr/compare/master...tmgstevens:LUCENE-8613?expand=1
but at the moment the tests don't work and I'm running out of Lucene knowledge
with which to debug, hence I'd welcome any support.
> CoveringQuery to use multi-valued field length as minimum match
> ---------------------------------------------------------------
>
> Key: LUCENE-8613
> URL: https://issues.apache.org/jira/browse/LUCENE-8613
> Project: Lucene - Core
> Issue Type: Improvement
> Affects Versions: 7.5
> Reporter: Tristan Stevens
> Priority: Minor
>
> Currently CoveringQuery uses a long field's value for the minimumMatch for
> each document.
> There are use cases where this value may be derived from the length of a
> multi-valued field (including docValues).
> This request is to support an alternative constructor and implementation for
> CoveringQuery to take a field name that can be used for the per-document
> minimum match value.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
