Ingest - handling "custom" scripts

Follow up to #833 to limit scope a little bit

We'll need to figure out how we want to handle these. I don't know that there will be a one-size-fits-all solution.

Specific script sources

• hra_centers (scrape/soup -> probably custom script source for now) this is defunct
• moeo_socialservicesitelocations (joins 4 datasets. could/should be split into 5 recipe datasets) has its own issue #1265
  • would be four simple recipes, and then one novel recipe to combine them. OR we just make four recipes, and save the unioning for facdb
• nycdoc_corrections (beautiful soup source. but just one webpage, nothing dynamic)
  • maybe we could get by with read_html? This one would be good to explore first I think
• nycoc_checkbook (special script. Connector!)
  • likely custom connector
• usfws_nyc_wetlands (eventually could break into multiple that get merged, for now just keep as custom script source)
  • break into multiple, union in gft
• uscourts_courts - funky one. Hits two endpoints, appends, deduplicates. Not could break into two, but not sure
  • break into multiple, union in pipelines

Start by reviewing the follow-up context in #833 and the listed sources, especially nycdoc_corrections, nycoc_checkbook, usfws_nyc_wetlands, and uscourts_courts. Determine which sources can become standard recipes, connectors, or split-and-union pipelines; done means the handling and scope for each remaining source are agreed.