Define a good PR review/stacking strategy for jkent

[albertisfu]

This is just to track the work Brennan is already doing to perform a full review of jkent before its release to PyPI.

[Brennan-Chesley-FLP]

1. The scraper definition interface
2. The database layer
3. The operational layer

[Brennan-Chesley-FLP]

These chunks ended up being too large, so I did a bit of shuffling into much smaller commits. The final PR has everything, the scraper definition interface is under review now. As that get's finished I'll just peel off the next commit into a PR to make things more manageable. The overall structure is still in place, it'll just be smaller pieces.