[
https://issues.apache.org/jira/browse/MAHOUT-1112?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13495646#comment-13495646
]
Andrew Janowczyk commented on MAHOUT-1112:
------------------------------------------
i hear ya, i believe i only used it in ClusterLabels, though i'd be very open
to suggestions : )
i found that the problem was that the function isDeleted was removed, making
iteration over all documents a difficult since you don't know if it is or isn't
there.
the AtomicReader assures a snapshot view of the repository at reader creation
time, so that problem is circumvented.
i didn't really see any special note in the migration text about how to deal
with this issue, but looking back i think i have a few ideas.
if its a problem, create me a ticket and i'll implement my ideas there in a
patch while you review this one. for the time being i think it should be okay
to at least get things rolling.
fyi the approach i'm thinking may also work is:
{code}
Bits liveDocs = MultiFields.getLiveDocs(indexReader);
if (liveDocs.get(docID)) {
// document is not deleted...
}
{code}
but haven't had time to test it
> Migrate code from Lucene / Solr 3.6 to 4.0.0
> --------------------------------------------
>
> Key: MAHOUT-1112
> URL: https://issues.apache.org/jira/browse/MAHOUT-1112
> Project: Mahout
> Issue Type: Task
> Affects Versions: 0.8
> Reporter: Andrew Janowczyk
> Assignee: Grant Ingersoll
> Fix For: 0.8
>
> Attachments: LuceneIterator.java.patch,
> mahout-lucene-3.6-to-4.0.0-patch
>
> Original Estimate: 20h
> Remaining Estimate: 20h
>
> Lucene/Solr 4.0.0 was released on October 12, 2012
> Many of the functionalities have changed, so it wasn't an easy migration.
> I'm including a patch, which passes all junit tests, for review.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
