DocumentExtractor is a traverser over the full-text results (EntityInfo)
This operation is as lazy as possible:
- the query is executed eagerly
- results are not retrieved until actually requested
#getFirstIndex and #getMaxIndex define the boundaries available to #extract.
DocumentExtractor objects *must* be closed when the results are no longer traversed.
Emmanuel Bernard, John Griffin, Hardy Ferentschik, Sanne Grinovero (C) 2011 Red Hat Inc.