lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Uwe Schindler (Commented) (JIRA)" <>
Subject [jira] [Commented] (LUCENE-3623) SegmentReader.getFieldNames ignores FieldOption.DOC_VALUES
Date Wed, 07 Dec 2011 14:30:40 GMT


Uwe Schindler commented on LUCENE-3623:

bq. MultiSource should just override asSortedSource() and throw an unsupportedOperationException.

This means you cannot use PKIndexSplitter and MultiPassIndexSplitter with docValues? We should
open another issue to make it work per-segment (by implementing a per-segment FilteredReader),
which is possible. Currently it wraps the source index by a SlowMultiReaderWrapper.
> SegmentReader.getFieldNames ignores FieldOption.DOC_VALUES
> ----------------------------------------------------------
>                 Key: LUCENE-3623
>                 URL:
>             Project: Lucene - Java
>          Issue Type: Bug
>    Affects Versions: 4.0
>            Reporter: Robert Muir
>             Fix For: 4.0
>         Attachments: LUCENE-3623.patch, LUCENE-3623.patch, LUCENE-3623.patch, LUCENE-3623_test.patch
> we use this getFieldNames api in segmentmerger if we merge something that isn't a SegmentReader
(e.g. FilterIndexReader)
> it looks to me that if you use a FilterIndexReader, call addIndexes(Reader...) the docvalues
will be simply dropped.
> I dont think its enough to just note that the field has docvalues either right? We need
to also set the type 
> correctly in the merged field infos? This would imply that instead of FieldOption.DOCVALUES,
we need to have a 
> FieldOption for each ValueType so that we correctly update the type.
> But looking at FI.update/setDocValues, it doesn't look like we 'type-promote' here anyway?

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:!default.jspa
For more information on JIRA, see:


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message