Replacing part of a string with another part of that same string [closed]

Question

How would I go about replacing part of a string with another part of that same string, in a bash script, using sed?

use the following shortcuts : tldp.org/LDP/abs/html/refcards.html#AEN22664 — PersianGulf, Mar 23 '13 at 21:40
Do you have the string in a variable? Or in a file? What are you doing with the string? — Gilles, Mar 23 '13 at 21:45
This needs a lot more detail to be able to give a reasoned answer. — vonbrand, Mar 23 '13 at 22:24

Stéphane Chazelas · Answer 1 · 2013-03-23 21:45:31Z

Your question is a bit too generic here.

An example:

sed 's/..\(.*\(..\)\)/\2\1/'

Would replace the first two characters of each line with the two last ones (only if there are at least 4 characters on those lines).

$ echo abcdef | sed 's/..\(.*\(..\)\)/\2\1/'
efcdef

The trick is to use \(...\) to capture what is being matched and use it in the replacement part as \1, \2...

closed as not a real question by Gilles, vonbrand, jasonwryan, rahmu, manatwork Mar 24 '13 at 10:32