Today I have a script that inside of it has a query that does a SELECT of the column name and the count of it. It can list more than 50+ different lines.
Ex.
BOOK 12
SHELF 2
DESK 3
Outline code:
[query] | while read unk_mtype_volume ; do
echo "$unk_mtype_volume"
done
I need to SUM every row number, to have a total.
This is a simple job for awk, no expensive iteration over the lines of the file/input needed. Just add up the second fields, and finally print the value:
awk '{sum+=$2} END{print sum}' file.txt
If the input is coming from STDIN:
... | awk '{sum+=$2} END{print sum}'
Example:
% cat file.txt
BOOK 12
SHELF 2
DESK 3
% awk '{sum+=$2} END{print sum}' file.txt
17
You can (and I would) use awk for the job — as shown by heemayl in their answer. You can also do it in pure Bash:
sum=0
while read location number
do
sum=$(($sum + $number))
done
echo "Total: $sum"
Note that this is fragile if any of the 'locations' consists of more than one word. You can code around that more easily in awk than Bash.
