I know, sorry about that. Got distracted thinking about a more probable use for it. I did write something with parsename, which is very helpful and probably smarter than messing with left and right, though it also just captures the .com/net whatever part of the email reliably under specific conditions.

https://www.dropbox.com/s/y7pcvk1i7h7h9wb/emailx.xlsx

I replaced all letters and numbers with Xs, which for some reason took me a really long time to write, even cheating with Excel. So you can kind of see what I'm working with, as far as there being a lot of '.'s in my data. Some of the email addresses are a malformed nightmare, btw.

I guess my original thought was "cool nested left/right thing to whittle down data to what I want", but my brain drifted right back to the awful email validation case statement I had to mess with this weekend, over here. So now I'm wondering if it's applicable to that in any way, though I think not.