Elegant, thanks!

What I'm doing is editing a bunch of files on a web server to reflect the new area code - so I had planned to read each line of the input file, make the replacement, and then save the new version (and a backup of the old version). Then I thought about the line-break issue.

So I think Roy's solution of grabbing a couple lines together, checking them, and them splitting them again before writing to the new file is what I need to do, in order to catch numbers that split over two lines. I see some stuff in your example that can help me with that.