Refactor

[GDYendell]

@marcelldls I have been playing around. What do you think of this?

Seems to be failing on a tango test. Should check if this is broken on main as well.

[marcelldls]

@marcelldls I have been playing around. What do you think of this?

Seems to be failing on a tango test. Should check if this is broken on main as well.

Error is: ResourceWarning: unclosed <socket.socket fd=28, family=1, type=1, proto=0>

As far as the testing goes, your tests work on my machine in a devcontainer... Im going to see if I can reproduce this

[marcelldls]

I have a branch graphql-backend-g-m which also fails. However, on a whim if I suppress all the epics tests, then all the tango tests pass https://github.com/DiamondLightSource/FastCS/actions/runs/12256441074

This would make me think that we have a tear down problem? It doesnt make sense that the test which fails, fails but the others pass. I feel like pytest is really annoying with its side effects as a first class citizen...

[marcelldls]

I reran the graphql-backend tests which were previously passing - they now also fail https://github.com/DiamondLightSource/FastCS/actions/runs/12139830254/job/34201957331?pr=71

[marcelldls]

Looks great - thanks for the docstrings