Question

I often need to run various scripts which edit the data inside a CSV file, but restricted to only a particular column within that file. Redesigning all of the scripts so that they only manipulate a particular column is sometimes complicated. I think the simplest method is to output the particular column to a temporary file, such as temp.csv, run the scripts on temp.csv, then place this edited data back into the same column in the CSV file.

I know that I can retrieve just the data from a particular column of a CSV file using this:

awk -F@ '{print $3}' file.csv >> temp.csv

After editing temp.csv with the scripts, what is the simplest way to return this data back to the same column in the CSV file?

Answer 1

You can modify $3 in place. E.g.:

awk -F@ 'BEGIN { OFS= FS }
         { # do whatever with $3, like
           $3 += 2
           $3 = $3 "appended text"
           # and so on, then finally
           print $0
         }' INPUT.CSV > OUTPUT.CSV