Question

I have a problem finding a sed script that works when the regex looks exactly the same but I only want to substitute one.

The file looks like this:

something someting something
this 
something someting something
something someting something
this 
someting something
someting something
someting something
this
someting something
someting something
this

I want to substitute the third “this” to something else. I have tried with:

sed '3,/this.*/s/this.*/something/'

and various similar attempts, but it doesn’t work.

Answer 1

This isn’t easy in sed but awk can do it

awk '/this/{count++;if(count==3){sub("this","something")}}1'