From #22 line notes: Since this uses the lookup protocol you can do a batch of work at a time. Look at MetadataWranglerCoverageProvider in circulation/coverage.py. I bet you could even refactor a lot of code into a common subclass.

Mimic the MetadataWrangler provider where possible and create a subclass if useful.

┆Issue is synchronized with a GitHub issue
┆Repository Name: metadata_wrangler
┆Issue Number: 26