Would it be more efficient to have Document as HashMap rather than list of Field, especially if Document does comprise of large number of fields?