[capitolweb] scraper and parser refactor

[will-horning]

This PR changes a bunch of things in the parser and scraper:

1. I removed the celery stuff, that would only really be useful if we were doing more large scale (and by necessity distributed) pre-compute jobs. Both scraper and parser are run through custom manage.py commands.
2. I separated the worker app into two apps: scraper and parser.
3. Added some thin test coverage for both scraper and parser.
4. Refactored the big parser function into a class that lazily computes each field using django's cached_property decorator.