[
https://issues.apache.org/jira/browse/LUCENE-1791?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12743036#action_12743036
]
Mark Miller commented on LUCENE-1791:
-------------------------------------
bq. Sweet! ... Even the random boolean query test? (i see changes to
ReqExclScorer in your patch, but i don't relaly see how that could solve the
problem)
Yes - the issue was again how it was dealing with the sub readers - my original
fix was not fully correct - it just happened to work with the previous tests.
By the way - any changes to non tests should have been pulled - sorry - I'll
put up another patch.
bq. I'm clearly missing something ... why does ItemizedFilter need to change? I
know it's an abomination and you would never want to use something like that in
real code, but in this test the docIds passed to it on construction shouldn't
change - even when we wrap the reader/searcher in other searchers, i specificly
only put empty indexes (with no deletions) in front of the original index when
bulding the multireader/searcher so the docIds will be the same. why doesn't
the existing implementation work?
Because it uses external ids - that only works if you have a single reader. The
filter is applied per sub-reader - so if you use a filter that just sets doc 0
- every sub reader will match its first doc. You could do this in the past,
because it just worked on the top level reader - in that case, external ids
match completely against the reader - but with the sub-reader approach, each
sub-reader has an id space that starts at 0.
> Enhance QueryUtils and CheckHIts to wrap everything they check in
> MultiReader/MultiSearcher
> -------------------------------------------------------------------------------------------
>
> Key: LUCENE-1791
> URL: https://issues.apache.org/jira/browse/LUCENE-1791
> Project: Lucene - Java
> Issue Type: Test
> Reporter: Hoss Man
> Fix For: 2.9
>
> Attachments: LUCENE-1791.patch, LUCENE-1791.patch, LUCENE-1791.patch,
> LUCENE-1791.patch, LUCENE-1791.patch, LUCENE-1791.patch, LUCENE-1791.patch
>
>
> methods in CheckHits & QueryUtils are in a good position to take any Searcher
> they are given and not only test it, but also test MultiReader &
> MultiSearcher constructs built around them
--
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: java-dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-dev-h...@lucene.apache.org
