If the file can be read entirely and there's no more data to parse after that, I'd expect that it would be more efficient to do as your coworker suggested. The amount of gain would depend on how ...