[
https://issues.apache.org/jira/browse/LUCENE-3126?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Shai Erera updated LUCENE-3126:
-------------------------------
Attachment: LUCENE-3126.patch
Patch includes:
* Modify CFW to not assume the files it 'merges' into a CFS exist in the same
Directory as it was initialized with + test
* Modify SegmentMerger following CFW updates + test
* Modify IW to copy a segment into a CFS + test.
NOTES:
* I did not cover 'unrolling' a CFS if the MP does not support CFS. I don't
think it's so critical (left a TODO in TestAddIndexes)
* About shared doc stores -- I'd appreciate a review of this. Since we don't do
shared doc stores anymore, I don't know how to simulate it for the test.
> IndexWriter.addIndexes can make any incoming segment into CFS if it isn't
> already
> ---------------------------------------------------------------------------------
>
> Key: LUCENE-3126
> URL: https://issues.apache.org/jira/browse/LUCENE-3126
> Project: Lucene - Java
> Issue Type: Improvement
> Components: core/index
> Reporter: Shai Erera
> Assignee: Shai Erera
> Priority: Minor
> Fix For: 3.2, 4.0
>
> Attachments: LUCENE-3126.patch
>
>
> Today, IW.addIndexes(Directory) does not modify the CFS-mode of the incoming
> segments. However, if IndexWriter's MP wants to create CFS (in general),
> there's no reason why not turn the incoming non-CFS segments into CFS. We
> anyway copy them, and if MP is not against CFS, we should create a CFS out of
> them.
> Will need to use CFW, not sure it's ready for that w/ current API (I'll need
> to check), but luckily we're allowed to change it (@lucene.internal).
> This should be done, IMO, even if the incoming segment is large (i.e., passes
> MP.noCFSRatio) b/c like I wrote above, we anyway copy it. However, if you
> think otherwise, speak up :).
> I'll take a look at this in the next few days.
--
This message is automatically generated by JIRA.
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]
