Shell Programming and Scripting

The line
print "YES : $(cat count.xls|grep -i '~Y or YES~' | wc -l)">>final_count.xls

is working correctly for me as i have data delimited by ~ and I am looking for occurances of field Y or YES.

No, you are not. This grep command will only match lines that contain the exact string ~Y or YES~. If you want to match any line that contains ~Y, ~y, YES~, YEs~, YeS~, Yes~, yES~, yEs~, yeS~, or yes~ you need something much more like what protocomm suggested (but for EREs with alternation, you need egrep instead of grep).

Please describe in English exactly what you are hoping to match with your grep commands. Note also since you apparently aren't matching complete fields, your search for NO will also match the lines containing No Resonse. I.e., the lines containing ~NO~ and the lines containing ~No Response~ will both be counted as matching ~NO when doing case insensitive matches. Is that really what you want?

Hi Don,

I need 4 different counts with their match for exact values as :
YES : ~Y or YES~
No : ~N or NO~
No Response : ~No Response~
Junk : Other than above 3

OK.

So this is a completely different problem than what you described in the 1st post in this thread. All of the lines you showed us in that post:

Code:

01~12345~Y~YES~aaaaa~can
02~23456~N~NO~bbbbb~can
.
.
.
99~23

are now to be counted as Junk and the counts of all of the other three categories (using your sample input) are to be zero.

So, the following totally untested code (since we have no samples to test) might do what you want:

Code:

awk -F'~' '
$1 == 99 { tr += $2 }
index("~Y or YES~") { y++;next }
index("~N or NO~") { n++;next }
index("~No Response~") { nr++;next }
{j++}
END {   printf("Total Records : %d\n\n", tr)
        printf("YES : %d\n\nNo : %d\n\nNo Response : %d\n\nJunk : %d\n", y, n, nr, j)
}
' $(cat file.txt) > final_count.txt
if [ $? -ne 0 ]
then    echo failed >&2
        exit 12
fi

Notes: