After downloading the corresponding metadata for each dataset #15: we'll need to: - extract attribute definitions and units from metadata and associate them with the correct entity/attributes - extract publication dates and timestamps to help with versioning and checking for new updates (relates to #6) - perhaps extract code definitions to help with flagged data (can be a separate sheet, can be comments, idk) #9