The OBDA flat file indexing specification adresses the need to provide a cross-platform indexing scheim for flat files that are reccord based. Typically these files have a primary ID per entry and potentially multiple secondary IDs in potentially multiple namespaces. Different entries will all have different primary IDs, but may shair seccondary IDs in a given namespace.
This package provides an API for managing these ID associations and retrieving blocks of files accordingly. Where ever possible, a sensible choice is made between linear scans and binary searches over ID. All index files use fixed-length reccords to make the binary searches as fast as is feasable.
Copyright © 2014 BioJava. All rights reserved.