I want get help to write a unix script to count "|" pipe in the flat file. The first row is header and the record count starts from 2nd row. I want to count the "|" in 2nd row only. Thanks for your help

awk -F"|" 'NR==2{print NF-1}' file next time, try to do it on your own first before posting. that will indicate you put in effort, even if its for homework. GNU win32 packages | Gawk

Thanks for the response. I already tried in the following way, but somehow it goes into endless loop. $ cat 1.txt abc|pqr|ttt|fff|rrr ctr=1 while true do var=`cat 1.txt | head -1 | cut -d'|' -f$ctr | wc -c` if [ $var -eq 1 ] break fi ctr=`expr $ctr + 1` done #number of field delimeters echo $ctr

that's a messy way to do what you want. learn how to use awk, as i have it to you. GNU win32 packages | Gawk

If you played with this on the command line, you'd see that variable var never equals 1 so the while loop cycles forever: cat 1.txt | head -1 | cut -d'|' -f 4 | wc -c I do not understand your logic and what you are trying to do. If you aren't going to use ghostdog's excellent awk one-liner, then consider cycling thru the file a line at a time and counting the pipe symbols: #!/bin/ksh pipecnt=0 while read line do count=0 linelen=`echo "$line"|wc -c` if [ $linelen -gt 0 ] then # don't do anything if line is zero len until [ $count -ge $linelen ] do count=`expr $count + 1` char=`echo "$line"|cut -b"$count"` if [ "$char" = "|" ] then # check if char is pipe pipecnt=`expr $pipecnt + 1` fi done fi done < 1.txt echo $pipecnt # end script I'm using ksh so you could use the ksh arithmetic method: ((pipecnt+=1)) # add 1 to pipecnt

well, if OP wants to do it with the shell, something like this may suffice while read line do x=${line//[!|]/} echo ${#x} done <file GNU win32 packages | Gawk