A new PlugOut that doesn't write any intermediate files (bar those directly associated with a document) and instead sends them to STDOUT in XML form (and InfoDBEntry elements) allowing proper parsing up in the Reduce phase in Hadoop.