I've been intrigued by the new Lucene FulltextIndex capabilities. It looks great when running through neo4J procedures, but I'd love to get access to some of the Lucene capabilities that are not exposed directly. Thus, I've been playing around with using the Lucene 5.5 API to connect to the index directly.
This has not gone well. I'm able to connect, but I cannot get queries to return any results, even given queries that work fine from the neo4j web console.
Is this fundamentally a bad idea? Is there a better way to get at the Lucene information (I'm particularly interested in doc /term frequency info)? Are there some specific ways of using the Lucene index that I should know about?