[
https://issues.apache.org/jira/browse/LUCENE-3083?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13030761#comment-13030761
]
Uwe Schindler commented on LUCENE-3083:
---------------------------------------
To check this we can readd the LTC.newMergePolicy() call in my patch of
LUCENE-3082 (once this is committed).
> MockRandomMergePolicy optimizes segments not in the Set passed in
> -----------------------------------------------------------------
>
> Key: LUCENE-3083
> URL: https://issues.apache.org/jira/browse/LUCENE-3083
> Project: Lucene - Java
> Issue Type: Bug
> Reporter: Uwe Schindler
> Assignee: Michael McCandless
> Attachments: LUCENE-3083.patch
>
>
> The test class MockRandomMergePolicy shuffles the whole SegmentInfos passed
> to the optimize callback and returns random segments for optimizing. This is
> fine, but it also returns segments, that are not listed in the
> Set<SegmentInfo> that is also passed in, containing the subset of segments to
> optimize.
> This bug was found when writing a testcase for LUCENE-3082: The wrapper
> MergePolicy (when wrapped around MockRandomMergePolicy) only passes a subset
> of the segments to the delegate (the ones that are in old index format). But
> MockRandom created OneMerge in its return MergeSpecification having segments
> outside this set.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
