Enable the download of raw data files

[dalonsoa]

The file cannot be downloaded because Django in production, by default, prevents from downloading raw data files for security reasons. They need to be served in a different way. If downloading the raw data files is a thing, I'll create an issue about that.

Originally posted by @dalonsoa in #332 (comment)

[ICHydro]

There is no need for the user to download the raw data file. In fact, there is probably no need to keep the raw file after the data have been processed. Similarly, there is probably no need for the user to be able to change the data file of the import after it is created; they can just create a new import.

[dalonsoa]

If the raw data file is not to be kept, is there any reason for not deleting the whole data import object altogether once the import is successful? I don’t think it has much use if there’s no file.

[ICHydro]

Indeed, I am reconsidering this. There is probably value in keeping the file to enable future reprocessing, and it can also serve as a place for users to keep the raw data files to ensure that they do not get lost, which enhances reproducibility and traceability. So let's keep this as an enhancement, including the possibility for users to download the original file.

[dalonsoa]

If we are going to keep the raw data and enable the user to download it, we might want to consider using a proper external service for this rather than storing the data in the VM, something like AWS S3 buckets or Azure Blob storage or something along those lines.

@ICHydro I'd suggest we leave this on hold until we can discuss the options and implications.