At present filename is not unique. I see that one user has uploaded "fusion.csv" at least 3 times, and presumably the latest file overwrites, since they're going to the same folder? So this is a problem if we need to access those files.
This simple-minded fix puts the timestamp onto the start of the filename. (Neater would be to use the submission ID number, but that is not known until after the db commit.)
To ensure unique filenames.
At present
filename
is not unique. I see that one user has uploaded "fusion.csv" at least 3 times, and presumably the latest file overwrites, since they're going to the same folder? So this is a problem if we need to access those files.This simple-minded fix puts the timestamp onto the start of the filename. (Neater would be to use the submission ID number, but that is not known until after the db commit.)