I think this probably needs a vignette to cover how to use it. I'll add that before it gets merged.

Things to note in vignette:

• I tried really hard not to break the existing API, and all of the previous tests work with no alterations, so I think I managed it
• You can only read from folders, cloud_put() could be updated but I don't think it's a priority
• To use you just pass the full URL as the path argument to cloud_get(), and the password if the URL needs one.
• Because the returned data doesn't include the filename, it can't be automatically opened with {rio}. You get a warning if open_file = TRUE (the default), and it will just return the file path instead.
• It works for folders too, but is not recursive (it will download only files at the top level). Downloads all files it finds and returns a vector of file paths.
• It will detect if a URL is for a file or a folder, you don't need to specify.

TODO:

• Vignette
• Update version & news