Thanks for the contribution. I'll take a more in-depth look shortly, though looks generally good. My one concern is that we now don't differentiate between a jsonPath being present and empty, and a jsonPath not being present at all. The existing solution is probably not ideal, but I don't feel we should lose detail that was previously available (however badly presented)