Coroutines using Redis and Python

At the 2015 Open Repositories conference in Indianapolis, I was approached by Mark Matienzo, Director of Technology at the Digital Public Library of America, to join a team for pitching ideas at a contest sponsored by the conference. Our pitch was for a Linked Data Fragments Server with caching that would enable people and organizations to provide a simple and well-understood service to query and get back RDF triples from a graph database instead of supporting full resource-heavy SPARQL endpoints that even for the largest website is difficult to keep and run for users. Linked Data Fragments, first proposed by Ruben Verborgh at Ghent University in Belgium, constructs a triple pattern fragment made up of a subjects, predicates, and object statements that combine to query a Linked Data store and returns a Linked Data Fragment made up of triples matching the query along with metadata and paging information.

Although our team did not win the pitch contest, with...