Over time, my criterion benchmarks have ended up following some common conventions. And where there is regularity, there is an opportunity for automation.
This program leverages the regularity of my benchmark naming conventions to automatically produce (for now) throughput plots vs input size.