Would it be possible somehow to make skip list (postings) RLE compressed
without affecting performance in cases where RLE cannot identify longer runs?
we have an unusual (?) case where we have an opportunity to sort documents on
category field before indexing. this order gets slightly disturbed
[
https://issues.apache.org/jira/browse/LUCENE-550?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Karl Wettin updated LUCENE-550:
---
Attachment: LUCENE-550_20070804_no_core_changes.txt
This is a small and completely isolated version o
On Saturday 04 August 2007 09:36, eks dev wrote:
> Would it be possible somehow to make skip list (postings) RLE compressed
without affecting performance in cases where RLE cannot identify longer runs?
I suppose you mean run length encoding? (I missed the first posting about
this.)
>
> we have
Hi Paul, thanks for feedback
>I suppose you mean run length encoding? (I missed the first posting about
this.)
You are right, that is what I meant by RLE. This was the first posting. I am
just trying get some feedback to see if there are some knock-out conditions
disqualifying this idea.
a b
Eks,
Two things:
In the latest Matcher patches I forgot to consider your implementation of
a Matcher on OpenBitSet. In case I don't mention this there, please holler
there when things start moving. My implementation of that in the latest
patches is really no more than a first attempt, and mostly
