Open tissole opened 3 years ago
Calculating the hash of the files would imply a significant disk and cpu usage. Research is needed.
Obviously, there's no need here for a heavily cryptographic hash algorithm because security is not a concern. A lighter non-cryptographic algorithm can be used and I found one blazing fast xxHash. It has a python implementation too.
Of course @tissole , I was thinking about sha1 or sha256.
I wish that the app could detect duplicate files and skipping them. Maybe storing the names and hashes of files uploaded in a database that keeps track of them. This will still not prevent uploading the files that were already in a channel, uploaded by other means.
So, a solution could be indexing the files from the channel in a database too, for future reference. This option could be turned on/off if the user wishes so.
This feature can be implemented?