Comment by vmfunction

Or wouldn't COPY from CSV be much faster?