Amend PLEP 7 to include plasmapy.io

[namurphy]

It's likely that in the future we'll need a subpackage that contains input/output (I/O) functionality for open data formats commonly used in plasma science. This possible amendment to PLEP 7 would define plasmapy.io, which would be akin to astropy.io and sunpy.io. This would be the place to put the functionality for reading in and outputting files in the OpenPMD format, for example, which are currently in plasmapy.plasma.

[StanczakDominik]

The hdf5 case is a pretty good example!

[StanczakDominik]

I can't help but notice how we've been reverting (and with good reason!) package structure cleanups I did a while back. Non ironically, really happy to see that!

[namurphy]

Converting it to a draft again for the moment to give it a bit more time to think about, in particular since I don't think we're going to implement this for a while.

@rocco8773 — did you want to mention your comments here?

[rocco8773]

My comment: io feels like it should belong in plasmapy.utils instead of plasmapy. I view the top-level packages should be focused on plasma topics, except for utils and tests. Thus, io fits better with the plasmapy.utils package. Additionally, a goal of PLEP-7 is to manage the plasmapy namespace and adding io feels like a pollutant to that namespace.

With that said, I can still see some arguments for having io in the top-level namespace, especially if it's robust and defines metadata standards. However, that does feel like a soft argument to me. I'd like to see the scope and framework of io defined better and that could convince be to put io into the top-level namespace.