[
https://issues.apache.org/jira/browse/LUCENE-5293?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13799492#comment-13799492
]
Paul Elschot commented on LUCENE-5293:
--------------------------------------
It felt like a big change when I started, but it was easier than I thought,
have a look at the patch of 18 Oct.
There are about 35 other places that directly extend DocIdSet or create a new
one from an inline subclass, I have not checked these yet.
This passes the current TestCachingWrapperFilter, but there are no tests for
this change yet.
For small segments, maxDoc() <= 256, this will use WAH8, would FBS better for
those cases?
The last choice for using the EF after the WAH8 was built is done using
sufficientlySmallerThanBitSet because that was available, but I'm not really
sure whether a smaller load factor should be used there.
> Also use EliasFanoDocIdSet in CachingWrapperFilter
> --------------------------------------------------
>
> Key: LUCENE-5293
> URL: https://issues.apache.org/jira/browse/LUCENE-5293
> Project: Lucene - Core
> Issue Type: Improvement
> Components: core/search
> Reporter: Paul Elschot
> Priority: Minor
> Attachments: LUCENE-5293.patch, LUCENE-5293.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.1#6144)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
