[
https://issues.apache.org/jira/browse/LUCENE-2468?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12868673#action_12868673
]
Michael McCandless commented on LUCENE-2468:
--------------------------------------------
bq. This is problematic, since a filter can be used somewhere in the query
tree, and wrapped for caching
Right so the issue here is that ConstantScoreQuery's scorer does not check
deleted docs when it runs -- it just relies entirely on what the filter said
was set.
But I expect this is the exception not the rule.
Ie most uses of a filter will see to it that deleted docs are already removed.
It's as if, somehow, when a caller wants the scorer or DocIdSet, it should
express whether it's OK that deleted docs are not removed... I think this'd be
another boolean arg (mustContainDeletions or some such) to scorer and
getDocIdSet.
Or, for a less invasive change, we could that you tell ConstantScoreQuery that
it must fully enforce deletions (if your app runs a query tree that has a path
involving ConstantScoreQuery not AND'd with some other query that'd enforce
deletions).
> reopen on NRT reader should share readers w/ unchanged segments
> ---------------------------------------------------------------
>
> Key: LUCENE-2468
> URL: https://issues.apache.org/jira/browse/LUCENE-2468
> Project: Lucene - Java
> Issue Type: Bug
> Reporter: Yonik Seeley
> Assignee: Michael McCandless
> Attachments: CacheTest.java, LUCENE-2468.patch
>
>
> A repoen on an NRT reader doesn't seem to share readers for those segments
> that are unchanged.
> http://search.lucidimagination.com/search/document/9f0335d480d2e637/nrt_and_caching_based_on_indexreader
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
