I have a script called foo.sh that contains something like
exec myapp -o size=100m -f
Any idea how to create another script that parses foo.sh and retrieves the value of size? One can assume myapp only appears once in foo.sh, but the order of size argument can appear anywhere in the argument list
Thanks
With grep in a shell :
$ grep -oP 'myapp.*?size=\K\d+m' foo.sh
100m
With awk in a shell :
$ awk -F'size=' '{sub(/ -f/, "");print $2}' foo.sh
100m
or
$ awk '{print gensub(/.*size=([0-9]+m).*/, "\\1", $0)}' foo.sh
100m
With perl in a shell :
$ perl -lne 'print $1 if /exec.*?size=(\d+m)/' foo.sh
100m
Or using a shell funny trick :
$ declare $(grep -oP "\b\w+=\w+\b" foo.sh)
$ echo $size
100m
cat foo.sh | egrep -o 'size=[[:digit:]]+' | awk -F= '{print $2}'
cat | grep | awk is a useless use of cat & grep. partmaps.org/era/unix/award.html#catA variation on the sed approach. This one short-circuits once it's found the line. Useful if the file is very long and the target is likely near the beginning.
sed -ne '/exec myapp -o size=/{s/[^0-9]*\([m0-9]*\).*/\1/;p;q;}'
Once it finds the correct line, it extracts the size value, prints it, and then quits.
