Write Flickr apidata gatherer

[jonasob]

Write a handler for Flickr photos for the hasher. It can re-use the same image hasher queue, but it needs a specific queue and handler for apidata. From a db perspective, the "url" field would most likely be populated with photo-id's from flickr, as the unique identifier, and handler with the string "flickr".

Enqueue could then do the same as for WMC, to pull in works that are loaded and waiting to be processed, and queue them - one at a time - on a flickr.apidata queue. The handler for that queue then retrieves information from the Flickr API (credentials on the Hackpad) an stores in apidata, as well as places a medium resolution (at least 640px) file URL onto the wmc.hash queue.