Proposal: Scan directory before pull

As far as I understand it, when a user pulls a dat, if some of the files exist already, they are simply overwritten. It should be possible to hash the data in the directory and compare it to the dat before pulling to avoid duplicate downloads.