Question

So I have a file of “keys”, for example:

key1
key2
key3

and I have a file of key:value pairs:

key1:value1
key2:value2
key3:value3

I want to replace the keys in my file of keys with their corresponding values in the key:value file. So the file of keys will look like this when complete:

value1
value2
value3
...

What is the best way to do this in bash? Note that a key may appear more than once in the keys file, but should only appear once in the key:values file.

Answer 1

if the join command is available in your environment, the following should work. The addition of an index via the awk command is needed to restore original key order (via a Schwartzian transform).

join -o 1.1,2.2 -t':' -1 2 -2 1 <(awk '{print(NR":"$0)}' key_file | sort -k2,2 -t':') <(sort -k1,1 -t':' key_values_file) | sort -k1,1 -t':' | cut -f2 -d':'