[
https://issues.apache.org/jira/browse/LUCENE-2348?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12935835#action_12935835
]
Robert Muir commented on LUCENE-2348:
-------------------------------------
{quote}
I think Robert's comments possibly stem from the misconception that the
duplicate filter somehow works like field collapsing. I wrote a test just to
illustrate how it actually behaves, just to make sure I wasn't confused myself
(since he seemed to think I was...)
{quote}
No, I understand exactly how this filter works. which is why my patch, that
uses SlowMultiReaderWrapper and forces the index to appear as if it were a
single segment, fixes the issue.
{quote}
Actually I think Filter is the natural fit for this functionality.
...
But I don't consider this a show stopper - it'd be simple to change
DuplicateFilter to receive the top IR up front, and pre-compute and
cache the bit set for all segments.
{quote}
So now you contradict yourself. the only way is like you said, in the ctor,
in other words, its forcefully cached. This is *unnatural* !
We should deprecate this functionality.
If someone wants to make a "DuplicateBitSetBuilder" that is a factory for
creating a BitSet,
to me that is more natural and obvious as to what is going on.
if its not doing the work in getdocidset, it shouldn't extend Filter!
> DuplicateFilter incorrectly handles multiple calls to getDocIdSet for segment
> readers
> -------------------------------------------------------------------------------------
>
> Key: LUCENE-2348
> URL: https://issues.apache.org/jira/browse/LUCENE-2348
> Project: Lucene - Java
> Issue Type: Bug
> Components: contrib/*
> Affects Versions: 2.9.2
> Reporter: Trejkaz
> Attachments: LUCENE-2348.patch, LUCENE-2348.patch
>
>
> DuplicateFilter currently works by building a single doc ID set, without
> taking into account that getDocIdSet() will be called once per segment and
> only with each segment's local reader.
--
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]
