exchange two words using sed in Linux

Question

guys! I am trying to exchange two words in a line but it doesn\'t work. For example: \"Today is my first day of university\" should be \"my is Today first day of university\"

Answer 1

You are not accounting for whitespace.

use [ \t]+ between words.

Answer 2

Try this:

sed -rn 's/(\w+\s)(\w+\s)(\w+\s)(.*)/\3\2\1\4/p' filename.txt

-n suppress automatic printing of pattern space

-r use extended regular expressions in the script

\s for whitespace

Answer 3

I start to make it with \s which means any whitespaces chars.
I use it for match every words with [^\s]*which match with everything but not spaces.
And I had \s* for match withspaces between words. And don't forget to rewrite a space in replacement.

Look a this for an example:

sed 's#\([^ ]*\)\s+#\1 #'

( I use # instead of /)

Answer 4

This might work for you (GNU sed):

sed -r 's/(\S+)\s+(\S+)\s+(\S+)/\3 \2 \1/' file 

Answer 5

sed -r 's/^(\w+)(\s+\w+\s+)(\w+)(.*)/\3\2\1\4/'

with your example:

kent$  echo "Today is my first day of university"|sed -r 's/^(\w+)(\s+\w+\s+)(\w+)(.*)/\3\2\1\4/'                                                                           
my is Today first day of university

for your problem, awk is more straightforward:

awk '{t=$1;$1=$3;$3=t}1'

same input:

kent$  echo "Today is my first day of university"|awk '{t=$1;$1=$3;$3=t}1'                       
my is Today first day of university