How can I reverse the first part of a sentence?

With sed:

sed 's/^/ /;:1 s/^\([^ ]*\) \([^ ]\)/\2\1 /;t1 s/ //' 

With perl:

perl -pe 's/\S+/reverse $&/e' 

With zsh:

string='a,b,c,d Access' setopt extendedglob result=${string/(#m)[^ ]#/${(j::)${(Oas::)MATCH}}} 

Using pee (pipe tee) from moreutils (apt-get install moreutils):

echo -n "a,b,c,d Access" | pee "cut -d ' ' -f 1 | rev" "cut -d ' ' -f 2" | tr '\n' ' '

The result is d,c,b,a Access with a trailing space, which comes from a translated newline.

Depending on Neven answer:

You can do this using this shell script:

#!/bin/bash secondPart=`echo "a,b,c,d Access" | awk '{print $2} '` firstPart=`echo "a,b,c,d Access" | awk '{print $1} '| awk -F , '{print $4","$3","$2","$1""}'` echo "$firstPart $secondPart" 

in command line just type:

sh yourShellscript.sh 

or using one command line as Jidder suggeted:

awk '$1=gensub(/(.),(.),(.),(.)/,"\\4,\\3,\\2,\\1","g",$1)' <<< "a,b,c,d Access" 

A perl solution:

$ echo "a,b,c,d Access" | perl -anle 'print ~~reverse($F[0])," @F[1..$#F]"' d,c,b,a Access 

Fair warning: this relies on your commas.

echo "a,b,c,d Access" | tr ,\ '[\n*]' | sed '1!G;$s/\n/,/g;s/\([^,]*\),\(.*\)/\2 \1/p;h;d' 

OUTPUT

d,c,b,a Access 

It's kinda long - longer than it needs to be. It's easier this way:

set -f ; IFS=,\ set -- ${0+a,b,c,d Access} while [ -n "${2+?}" ] do r="$1,$r" shift ; done echo "${r%,} $1" 

OUTPUT

d,c,b,a Access 

Still kinda long, but at least it makes sense. It get's a little easier with a function because you get a shell array. Like:

rev() ( i= set -f ; IFS=$1 ; shift ; set -- $* until [ "$((i=${i:=($#+1)}-1))" -eq 0 ] do [ "$i" -gt 1 ] && s=$IFS || s= eval printf '"%s$s"' \""\$$i"\" done ) rev , a,b,c,d && echo \ Access 

OUTPUT

d,c,b,a Access 

Or if you have it in a variable:

var="a,b,c,d Access" rev , "${var% *}" && echo " ${var##* }" 

OUTPUT

d,c,b,a Access