This is an umbrella ticket for tracking of lucene 2.9 upgrade.

Things to keep in mind:

Segment readers is now exposed, and by default, Lucene is loading (FieldCache) and searching in segments. BoboIndexReader being a FilteredIndexReader may not be correct anymore, perhaps BoboSegmentReader?

A patch make bobo trunk can be built with lucene 2.9 (but fails the tests).

Changes: * DocIdSet.iterator throws IOException now; * Weight is an abstract class, and its scorer method has too more args now; * createWeight goes public; * Include issue [24]: support multi segmentreaders.

Just make compilable, incorrect behaviors are not fixed.

I have the same problem,can't use with lucene 2.9

are you guys looking at the right branch. Trunk is not 2.9 compatible. We are working on 2.9 integration on:BR_DEV_LUCENE_2.9

Issue 17 has been merged into this issue.

Issue 13 has been merged into this issue.

Patched bobo-solr for solr 1.5 dev trunk. Attached patch file. SOLR 1.5 Rev: 894477

I managed to make this lucene 2.9 branch work with solr 1.5 dev by patching bobo-solr

I added bobo-solr to the solr trunk and built the solr war with bobo inside - this seemed easier to me

Had to make sure the same lucene library used by solr was used by bobo-browse but it seems to work on a quick test.

Has bobo-browse been fixed to work on lucene 2.9 - as this bug looks like its still open - what are the remaining issues?

Thanks Abdul!

The remainder issues:

1) one area we can optimize on performance with facet merges for each segment. We want to do this before the official release. 2) We want to have it running in production in LinkedIn.com for a while before releasing to make sure everything is smooth both in functional and performance.

-John

Abdul:

 I looked at the patch. The problem is that if you call

BoboIndexReader.getInstance per request, that is a significant performance penalty. What was the problem to let the IndexReaderFactory do it?

-John

Hey,

The original code wouldn't compile with the solr head(1.5 -dev) and I I couldn't get the IndexReaderFactory to work but then I really didn't look into it in much detail as I ran out of time.

-Abdul

i found that is we set Filter,DocIdSet.iterator throws IOException;

HI Yumwiming:

 We have moved to Github: http://github.com/javasoze/bobo.

 And issue management has moved to jira: http://snaprojects.jira.com/browse/BOBO

 Please find the latest stuffs there.

Thanks

-John