[
https://issues.apache.org/jira/browse/LUCENE-5860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14079942#comment-14079942
]
Uwe Schindler commented on LUCENE-5860:
---------------------------------------
Hi, looks interesting, although I am not sure if the added complexity,
especially in NRQ is woth the small speed improvement. I have to carefully
apply the patch and review it, maybe tomorrow.
Without a good benchmark result that brings a good speed improvemenmt, I would
-1 it for NRQ (especially as NRQ only looks at few terms at all - at least with
my preferred precisionSteps, I still disagree with the new defaults). For NRQ
it is also not sure, if the additional min/max lookup cost is useful (because
the min/max term in the term dictionary does not help for the NRQ case, you
need the min/max shift=0 term).
Go for it for TRQ! Strong +1
> Use Terms.getMin/Max to speed up range queries/filters
> ------------------------------------------------------
>
> Key: LUCENE-5860
> URL: https://issues.apache.org/jira/browse/LUCENE-5860
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Michael McCandless
> Assignee: Michael McCandless
> Fix For: 5.0, 4.10
>
> Attachments: LUCENE-5860.patch
>
>
> As of LUCENE-5610, Lucene's Terms API now exposes min and max terms in
> each field. I think we can use this in our term/numeric range
> query/filters to avoid visiting a given segment by detecting up front
> that the terms in the segment don't overlap with the query's range.
> Even though block tree avoids disk seeks in certain cases when the
> term cannot exist on-disk, I think this change would further avoid
> disk seeks in additional cases because the min/max term has
> more/different information than the in-memory FST terms index.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
