Shell Programming and Scripting

I have a script that shows me the disk SPace used by different dir under my home dir:

Code:

#!/bin/ksh
cd /ednpdtu3/u01/pipe

p1=`df -g | tail -1 | tr -s " " | cut -d " " -f2`
echo "Total Disk Space of Home Dir is $p1 GB"
p2=`df -g | tail -1 | tr -s " " | cut -d " " -f3`
echo "Total Disk Space Free in Home Dir is $p2 GB"

count=`ls -ltr | grep ^d | wc -l`
echo "$count Dir found in HOME DIR"

a1=1
while [ $a1 -le $count ]
do
  b=`ls -ltr | grep ^d | head -$a1 | tail -1 | tr -s " " | cut -d " " -f9`
  #echo $b
  a1=`expr $a1 + 1`
  a2=`du -sk $b`
  echo "$a2" 
done

------------------------------------------------------------------------

Output i m getting:
28 ssh

89508 Outgoing

6393868 Incoming

4 Mail

----------------------------------------------------------

But i want to get only numeric value i.e. whenever it encounter space , it should remove spaces and watever is after spaces so that i can get onl;y numeric part. My o/p should be like below

28

89508

6393868

4
--------------------------------------------------------------

I know we can do it by SED and we have to implement this in a2=`du -sk $b` . I tried some commands with sed
,but not successfull

a2=`du -sk $b | `sed 's/ //g'-----> not working

I m pretty much sure we have to implement the logic in that part of the script i.e. a2=`du -sk $b` and using SED but not getting the correct way to use SED

Plz help me in this

Code:

sed 's/ .*//'

or:

Code:

awk '{print $1}'

Regards

thanks a lot got the answer by command du -sk $b | awk '{print $1}'