Move Data Retrieval Code to New Repository

[WardLT]

As we modify the codes that publish data from certain repositories, should we move it to their own repositories? They provide great examples for how to use MDF Connect, and are kind of hidden here.

I'm willing to do this. Does anyone object?

[jgaff]

As discussed offline:
I think it's a good idea to update the harvesters that can still be useful to us (MDR, for example) and move them to the docs/examples dir in the connect_client repo. Having practical, functional examples of dataset submission is a great idea, and I think having those examples close to the client will aid discovery.

[WardLT]

I'm adding a list of databases that we should build indexers for, which is very incomplete at the moment:

• AiiDA workflow tools, which have a distributed database assoicated with them
• Materials Cloud
• Materials Project
• HTEM DB