[
https://issues.apache.org/jira/browse/LUCENE-2541?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12889355#action_12889355
]
Uwe Schindler commented on LUCENE-2541:
---------------------------------------
After going to sleep I found out the following (in my dreams...):
Your patch fixes the *bug* but it breaks NRQ for such bounds. Whenever the "+/-
diff" operation overflows the long, it stops creating new sub ranges. This
leads to the fact that NRQ will enumerate *all* terms and not use lower
precisions! So as soon as the min or max bound is near +/- MAX_VALUE, the NRQ
behaves like TRQ - bummer.
The easy solution would be to use BigIntegers with 65 bits :-) Maybe thats the
easiest solution to correctly calculate this crazy stuff.
> NumericRangeQuery errors with endpoints near long min and max values
> --------------------------------------------------------------------
>
> Key: LUCENE-2541
> URL: https://issues.apache.org/jira/browse/LUCENE-2541
> Project: Lucene - Java
> Issue Type: Bug
> Components: Search
> Affects Versions: 2.9
> Reporter: Koji Sekiguchi
> Assignee: Uwe Schindler
> Attachments: LUCENE-2541-uwe.patch, LUCENE-2541.patch,
> LUCENE-2541.patch, TestNumericRangeQuery.java
>
>
> This problem first reported in Solr:
> http://lucene.472066.n3.nabble.com/range-query-on-TrieLongField-strange-result-tt970974.html#a970974
--
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: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
