[WIP] #197 SCOTUS scraper

[ralexx]

From freelawproject/foresight#20 .

Notable

• Adds pymupdf as a dependency, as suggested.
• Includes log-then-raise handling for two characteristic host responses ('Access Denied' page and server name resolution error) that appear to be anti-abuse measures.
• Because of the above, only single-threaded downloading was implemented here.

TODO

• Docket parser mapping to database structure: maintainer input needed
• Python 3.8 support if necessary (I used some 3.9+ syntax)
• Tests for docket parser
• Tests currently omitted because they would need additional mocking
• Better handling of the two host responses mentioned above
• Multithreaded downloading, if desired (implemented locally but not included in this draft)

[CLAassistant]

All committers have signed the CLA.