Hello there,
I'm a solr newbie but i've used lucene for some complex
IR projects before.
Can someone please help me understand the extent to which solr allows
access to lucene?
To elaborate, say, i'm considering the use of solr for all its wonderful
properties like scaling,
distributed search, ease of updates,etc. I've a corpus of data that i'd
like lucene to index.
Further, I'm working on some graph research, where i'd like to disjunctively
query keyword
terms and use the indepent result sets as entry points into my graph of
documents.
I have my own data structures (in java) that handle efficient graph
walks,etc and eventually
apply a whole bunch of math to re-rank results/result trees.
In a more traditional setting, i can imagine using lucene as an external jar
dependency, hook
it up with the rest of my code in java and ship it off into Tomcat.
Is this doable with solr? Please help with comments on the specifc mechanics
of hooking up
custom java application logic with lucene before integrating with the rest
of the tomcat ecosystem.
Thank you very much.
Nick.
