Sed and awk have a lot of overlap. Another thing to consider is that neither might be the right choice if you are inside a bash script since spawning a new process in a tight loop can be very expensive and bash's built-in regex operations can have much better performance in those situations.