[
https://issues.apache.org/jira/browse/LUCENE-3505?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Robert Muir updated LUCENE-3505:
--------------------------------
Attachment: LUCENE-3505.patch
Updated patch: the specialization of ConjunctionTermsScorer broke this
navigation API as well, because it doesnt have subscorers, instead accessing
the docsenums and shit directly.
But there is a way worse bug in this: if even one term is omitTF in your term
conjunction it treats the whole conjunction as match-only, which is wrong
scoring!!!!!!
I fixed both of these in the patch. I think we can improve/refactor/generalize
this conjunction stuff even better, but this issue is already out of control.
I'm just trying to fix bugs.
> BooleanScorer2.freq() doesnt work unless you call score() first.
> ----------------------------------------------------------------
>
> Key: LUCENE-3505
> URL: https://issues.apache.org/jira/browse/LUCENE-3505
> Project: Lucene - Java
> Issue Type: Bug
> Reporter: Robert Muir
> Fix For: 4.0, 5.0, 3.6.2
>
> Attachments: LUCENE-3505.patch, LUCENE-3505.patch, LUCENE-3505.patch,
> LUCENE-3505.patch, LUCENE-3505.patch, LUCENE-3505.patch
>
>
> its 0, the freq() is then calculated as a side effect of score()... we should
> at least document this or throw UOE for freq() instead.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
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]
