使用SED在bash中交换列而不使用循环

Question

I'm new to Sed, I'm trying to learn some pattern using Sed. I got a filenamne.txt that has the following entry:

ppp/jjj qqq/kkk rrr/lll

My goal is to swap the word before the slash and the word after the slash in each of the three word1/word2 columns:

jjj/ppp kkk/qqq lll/rrr

I tried using sed –re 's!(.*)(/)(.*)!\\1\\2\\!' filename.txt sed –re 's!(.*)(/)(.*)!\\1\\2\\!' filename.txt , but it didn't work. Any idea how can I go about it?

Answer 1

$ echo "ppp/jjj qqq/kkk rrr/lll" | sed -e 's/$/ /' -e 's!\([^/]*\)/\([^ ]*\) !\2/\1 !g'
jjj/ppp kkk/qqq lll/rrr 

Answer 2

Use replacement in perl command-line is a lot more straight-forward :-

perl -pe 's/(\w+)\/(\w+)/$2\/$1/g' file
jjj/ppp kkk/qqq lll/rrr

Answer 3

$ sed 's#\([^ ]*\)/\([^ ]*\)#\2/\1#g' file
jjj/ppp kkk/qqq lll/rrr