bash script to extract string from last output line

Question

This bash script tries to get the string "ami-774b7314" from the output of the following command so that it can be used as input to another command in the same script:

aws ec2 describe-images --region ap-southeast-2 --owners amazon --output text --query 'Images[].{A1name:Name,A2id:ImageId}' --filters Name=is-public,Values=true,Name=name,Values=amzn-ami-*.*.*-amazon-ecs-optimized | sort

amzn-ami-2016.03.i-amazon-ecs-optimized ami-22a49541
amzn-ami-2016.03.j-amazon-ecs-optimized ami-862211e5
amzn-ami-2016.09.a-amazon-ecs-optimized ami-73407d10
amzn-ami-2016.09.b-amazon-ecs-optimized ami-5781be34
amzn-ami-2016.09.c-amazon-ecs-optimized ami-774b7314 <===== this line

and to verify it is in the format ami-and-a-mix-of-alphanumeric else echo "bad string" and exit the script

How can it be done?

Answer 1

The tail command may be used for this. It gives you the last few (10 by default) lines of input as output (the "tail" of it).

With the -n flag you may specify exactly how much of the tail you'd like to have:

aws ec2 ... | sort | tail -n 1

To verify that it follows the right format, you may do

line="$( aws ec2 ... | sort | tail -n 1 )"

if [[ ! "$line" =~ ^amzn-ami-[0-9]{4}\.[0-9]{2}\.[a-z]-amazon-ecs-optimized\ ami-[0-9a-z]{8}$ ]]; then
    echo "bad string"
    exit 1
fi

If you only want the last bit of the last line, use cut to cut it out:

line="$( aws ec2 ... | sort | tail -n 1 | cut -d ' ' -f 2 )"

The -d ' ' -f 2 bit says "use space as the field delimiter and give me the second field".

Then the regular expression becomes shorter too:

if [[ ! "$line" =~ ^ami-[0-9a-z]{8}$ ]]; then
    echo "bad string"
    exit 1
fi