Hi I have a .dat file named say file.dat it contains something like this Build_ver=CTM_4_0_0 Build_No=101 Patch=1 i want to open this .dat file from shell script and i want the values to be assigned to another 3 diff variables? how should i do it Appreciate your help

Use nawk: VAR1=`nawk -F= '{if(NR==1)print $2}' info.dat` VAR2=`nawk -F= '{if(NR==2)print $2}' info.dat` VAR3=`nawk -F= '{if(NR==3)print $2}' info.dat` echo $VAR1 --> CTM_4_0_0 echo $VAR2 --> 101 echo $VAR3 --> 1 -jim

Don't need nawk. Just put them in an array. I put your example data in a file called junk.dat. Jerry #!/bin/ksh exec 3< junk.dat counter=1 while [ $counter -le 3 ] do read -u3 Line var[$counter]=$Line ((counter += 1)) done counter=1 while [ $counter -le 3 ] do print ${var[$counter]} (( counter += 1 )) done

That last solution does not reassign the variables to different variable names. It loads entire lines into an array, then prints those entire lines from the array. The same results can be accomplished with: cat junk.dat I offer two solutions. The first, same as jimbo, reassigns to sequentially numbered var1, var2, etc, in whatever order the input is in. But instead of using awk, I use read like Jerry, except that I delimit the incoming fields with an equal-sign so I can get to the second field: #!/bin/ksh k=1 cat file.dat | while IFS="=" read word1 word2 do eval var$k=$word2 ((k=k+1)) done print var1=$var1 print var2=$var2 print var3=$var3 exit 0 But this second solution should be more useful. It looks for the old variable names of interest, then reassigns the assigned value to corresponsing new variable names. Lines not of interest will be ignored: #!/bin/sh cat file.dat | while IFS="=" read word1 word2 do case $word1 in Build_ver) myver=$word2;; Build_No) myno=$word2;; Patch) mypatch=$word2;; esac done print myver=$myver print myno=$myno print mypatch=$mypatch exit 0

Would have been a trival exercise to put them into different variables. Thought the array was sufficient. There is absolutely nothing wrong with the provided solution.