Add dataset hashing #8
Description
To make pipeline running more efficient, maintain a single file with a hash for each dataset and add a before_node_run hook that checks the hashes. Is there a cached dataset that would work for this? Or should it just be a file?