Make downloading faster and more efficient

[ZeevoX]

• Use multiprocessing for significantly faster downloads that actually max out the available bandwidth
• Add persistent caching using diskcache library to skip unnecessary requests (this is useful when downloading a massive collection of photos for archival and the connection is interrupted or the program exits due to an error; this feature would need to be disabled if fetching new content from the collection of a user that has previously been downloaded)
• Minor bug fixes, reformatting, refactoring...

NB: I made these changes to the code for myself, and only later thought of submitting a pull request. Perhaps this does not fit with the author's vision of the project and that is absolutely fine. I leave it here for those interested users with a similar use-case to mine.

[ohare93]

Can confirm: works like a charm :+1:

[ohare93]

[I would raise this as an issue on your repo @ZeevoX but it seems that issues are not enabled]

Though the caching needs a bit of work and can be confusing. I ran it without the session cookie the first time, to see what it could see publicly. After adding the session key it still only found the same public albums (due to the caching). Deleting the cache folder solved this, but it was rather confusing.