[
https://issues.apache.org/jira/browse/LUCENE-5545?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael McCandless updated LUCENE-5545:
---------------------------------------
Attachment: LUCENE-5545.patch
New patch, improving the tests, which found a problem (infinite loop!)
with QueryRescorer, which I fixed by removing entirely the
OnlyDocIDsFilter. Now I just interact directly w/ the Scorer per
segment instead, and it also addresses the TODO about not creating
lots of objects.
I think it's ready.
> Add ExpressionRescorer
> ----------------------
>
> Key: LUCENE-5545
> URL: https://issues.apache.org/jira/browse/LUCENE-5545
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Michael McCandless
> Assignee: Michael McCandless
> Fix For: 4.8, 5.0
>
> Attachments: LUCENE-5545.patch, LUCENE-5545.patch, LUCENE-5545.patch,
> LUCENE-5545.patch
>
>
> In LUCENE-5489 we added QueryRescorer, to rescore first-pass hits using
> scores from a (usually) more expensive second-pass query.
> I think we should also add ExpressionRescorer, to compute the second pass
> score using an arbitrary JS expression.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
