This is a place where marcel's python grounding is really useful. Reading json, marcel generates a stream of nested dicts and lists. That stream can be turned b

This is a place where marcel's python grounding is really useful. Reading json, marcel generates a stream of nested dicts and lists. That stream can be turned back to json for output. Similar for CSV.