Add getting dataset status by compute tag

[bennybp]

Description

Adds a function (client & server) to get the status of a dataset grouped by tag.

Tested with some larger datasets on a test server and it seems fast enough (typically less than 3 seconds or so)

(if there's any more, I imagine we should make one general function)

Requested by @jaclark5

Status

• Code base linted
• Ready to go

[jaclark5]

@bennybp This looks great! It seems I can't test it live until it's merged and on the servers?

So it would return something like:

response = [
    ("tag1", "running", 5),
    ("tag1", "waiting", 400),
    ("tag2", "error", 300),
]

Would it be possible for it to return the record ids in each tuple? Either that or could ds.iterate_records() have a keyword for compute_tag?

[bennybp]

The former might be better. But do you want record ids or entry/specification strings?

[jaclark5]

Probably the entry/specification since it's faster to search by entries, it just thought record ids would be easier since the former would result in a tuple like (tag, status, count, [(entry1, spec1), (entry2, spec2),....)]).

[bennybp]

Record ids are definitely easier on my end. I guess whats missing is a function to get a record by ID from the dataset? Although you can just use the top-level client for that

[jaclark5]

The next step I would want to take with this information is get Molecule properties from the different categories. Having entries would be easiest since it's fast to call those from the dataset. If I had record ids I'd have to call them through the client to get entry names and then call those from the dataset, so entry names and specs are probably better for what I'm looking for.