This will likely required us to eval the code for each hint 2x, once in the context of the eval statement, and once outside with the code contents available for regex checking.
I initially avoided using the FCC evaluator just because this initial pass was so simple, but it would be nice to support more challenge types. Keeping this open while we wait for feedback on the project.
We currently have the runner working, but it fails to take into account a couple of edge cases. We should rework it to suppor the following: