rehaul how drives are handled

[hildebro]

previously, the fields hdd_count and hdd_size where used. these had various problems:

• didn't account for SSDs at all
• didn't account for servers with varying disk sizes
• were just straight up wrong sometimes :D

so now, we use the more accurate array serverDiskData to fetch all types of disks.

also changed the offer formatting to account for the new types.

todo for another PR: add /set commands for ssds

one small issue with the PR: the new model offer_disk has an id property, because sqlalchemy doesn't like tables without a primary key. now on each call to update_offers, we toss away all existing entries and write new ones. this causes the id values to keep getting bigger over time.

[Nukesor]

FYI, there's a just format action, which should fix all issues that come up in the CI

You can check via just lint

[Nukesor]

Nice. Thanks for the contribution :blush: