Cache all hashes of files into database

[rudolphos]

To prevent rehashing the same file over again and thus improving the efficiency I propose adding a basic database to store hashes of each hashed image tied to their full path. This would be especially useful if there are several 10,000s of images.

[martok]

Speed wouldn't really be an argument (I can run 100k files in 30 minutes on my 10 year old laptop).

If anything, I would probably only have a file content map SHA256->DHash-Tuple so that moving files doesn't recomupte either. This would also catch exactly identical files for free.