GitLab

The current Pyonionoo returns only summary documents, which is a great start, but not enough to replace the Java Onionoo and get Pyonionoo some real usage. There's a single details/bandwidth/weights file per relay that changes at most once per hour together with the summary file. Pyonionoo would only have to read that file, either from disk or from its database (after importing it into that database, of course), and include the documents by fingerprint that match the user's query in the result.