Hi,
I've contributed a JCR Directory implementation for Lucene.
What it does: allows the storage of indexes in a JCR repository
Why: single place for content and index storage, transaction support (if
the JCR implementation support them),
Performance: JCR does not support random access (only plain Inputstreams),
but my implementation splits the files into fixed-size buffers to help
performance. It will also depend on the persistence mechanism used with JCR.
Status: beta
You can retrieve the code there:
http://issues.apache.org/jira/browse/LUCENE-484
I welcome your comments.
Regards,
Nicolas Bélisle
Laval University Library
