[
https://issues.apache.org/jira/browse/LUCENE-5486?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael McCandless resolved LUCENE-5486.
----------------------------------------
Resolution: Fixed
> query time join should override Weight.scoresDocsOutOfOrder
> -----------------------------------------------------------
>
> Key: LUCENE-5486
> URL: https://issues.apache.org/jira/browse/LUCENE-5486
> Project: Lucene - Core
> Issue Type: Bug
> Components: modules/join
> Reporter: Michael McCandless
> Assignee: Michael McCandless
> Priority: Minor
> Fix For: 4.8, 5.0
>
> Attachments: LUCENE-5486.patch
>
>
> The TermsIncludingScoreQuery seems to have optimized impls when the collector
> can accept out-of-order docIDs. BTW, as far as I know, besides
> BooleanScorer, this is the only query scorer we have that can "take
> advantage" of out-of-order docIDs during collection.
> However, its Weight class fails to override scoresDocsOutOfOrder, which then
> means that for the common case where apps use IndexSeacher.search to sort by
> score or field, these impls will never be used.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
