Question

Why does this script work with the old version of the file?

I want to remove the first line of a csv if the line contains the text KP_TITEL
If i test the script with an exit after sed everything works fine. But as soon as I want to test it with the while loop the file still contains the first line.

Any Idea?

sed -i '/KP_TITEL/d' "$1"

# read data
input=$1

while IFS=';' read -r f1 f2 f3 f4 f5 f6 f7 f8 f9 f10 f11 f12 f13
do

  other things

done < "$input"

Answer 1

I’m not sure what the option -i does for sed but usually, sed is a stream editor which processed the input and writes the result to stdout. In your case, the output is lost and then, the old file is read again by while. Try this instead:

sed '/KP_TITEL/d' "$1" | \
while ...
done

The pipe symbol will change the stdin of the while loop (and hence the read command).

I continue the line with \ since I find this more readable than:

sed '/KP_TITEL/d' "$1" | while ...

And to delete the first line, you should use either

sed 1d "$1"

or (cheaper)

tail --lines=+2 "$1"