Refactor discovery

[Darsstar]

First commit:

• test with 3.11 and 3.12
• require combined coverage from all versions to be 100% instead of individual Python versions, this allows sys.version_info < (3, 10) checks to be used in the code base.

Second commit:

• The main purpose is to speed up discovery.
• But is also provides similair functionality to Optional model parameter to bypass discovery #144.
  • The difference is this PR does not use the class name, but creates names entry points instead.
  • This PR allows for a list, instead of a single class/model. The Home Assistant integration could add a config entry to manage which inverter classes/models are enabled. During discovery it would then only pass those enabled inverter classes/models to solax.discover()
    • solax.discover() can also return a list of all inverter models/classes instead of returning the fist that succeeded.
      Which can be used by Home Assistant to have the user resolve such situations.

Third commit:

• Provides functionality to fix Fix SolaX solar energy production not registering energy in energy dashboard home-assistant/core#107615 properly
  • x3_hybrid_g4 has a monlty reseting energy, which isn't supported by Measurement, but that can easily be added if desired.

Fourth commit:

• Adds inverter SN to the named tuple. Home Assistant could then use that instead of the dongle SN. Which would have the benefit of having only one entity if a broken dongle needs to be replaced.

@squishykid It took me longer to finish than expected. Hope you enjoyed your holiday!
PS. I assume you are the one who started following me on Trakt :p

[brew-your-own]

Nice!
Do you also have patches for HA to apply your approach in the component?

[squishykid]

looks good to me! nice work mate.

[nazar-pc]

It is Data, not data in JSON. Just like Information below. Why did you change it?

UPD: Hm... it does seem to work regardless though.

[Darsstar]

Because I lowercased all keys in a previous PR.

[Darsstar]

I accedentilly hit "Comment" in the mobile app...

I meant to write commit not PR, but both are wrong, it was in the same commit.

Anyway, I assume #161 is why you decided to comment.

[nazar-pc]

Yes, I was going through recent changes to see if something broke, but then remembered about firmware upgrade and found the root cause