formalize and implement bulk loading

[ksonda]

GraphDB seems to be able to ingest both jsonld graphs and what I would call jsonld "feature collections" that repeat the context over and over without much problem (see /create-bulk-files for examples).

This requires rules and implementation both for Gleaner @fils and for pids.geoconnex.us @webb-ben

Question is how to formalize submissions of these type over at https://github.com/internetofwater/geoconnex.us as part of the PID registration process.

Suggestion:

• Come up with threshold count to ask for a regex
• make regex require a "bulk link" column for a single .jsonld file with no pagination

Considerations:

• Even bulk graphs may have some practical size limitations. This is why my workflow in /create-bulk-files splits large collections up into multiple .jsonld files