Add external scaffold to/from json and hdf5

[jaclark5]

Description

This module will allow users to write and read a json file that can be reformed into an initial dataset before submission. This means that records are not present.

Status

• Code base linted
• Ready to go

[bennybp]

Kind of surprised this works. I guess it's just from keeping names consistent :)

It would be safer (and faster?) to build entry object(s) and use ds.add_entries(obj_list).

Every dataset has it's own entry type, but you can get it with ds._entry_type. So this might work (going from memory)

entry_type = ds._entry_type
....
new_entry = entry_type(**entry)

I think specifications will be ok to do your current way, though. Those are consistent across dataset types.