> I used the "for each line" in a recent project that evaluates 500,000 > lines > and takes just a couple of seconds to complete. It is VERY fast! > > SKIP > I guess the reason it's fast is because it implements the equivalent C loop in a very raw manner... jbv