Notes:

The files are in one large directory with a csv file parsed to identify the files and what they are attached to. There is a seperate csv mapping the reference_id from the file to the particular page and metadata on it