Scenario: I have a directory full of code files (in this case, 51 of them) that each need a find-and-replace executed; however, it is adding a line to part of the code, so replace one line with several. This means that SED cannot do the trick; you’ll get errors like, > > > sed: -e expression #1, char 49: unterminated `s' command Rather than spending the time figuring out how to encode things for SED, the simplest solution was just to use emacs inimitable dired.

I start with a list with a bunch of entries like this: newsletter W02 newsletter W02 newsletter W02 newsletter W02 EDIT: New Solutions It turns out that emacs automatically initializes a variable that tracks how many interations your replacement function has made, and in lisp mode you can get to it with \#. This way, the single line necessary to number all my newsletters becomes this: W\(02\) → W\,(format "%02d" (+ 4 \#))) and, of course, you can adjust that 4 to whatever you need it to be for your starting digit.