Refactor discovery

[Expl0dingBanana]

The goal of this PR is to make discovery more unit-testable outside of HASS. There will be a follow-up PR the begins to merge manual discovery and auto discovery into a single pipeline to reduce the amount of duplicate code.

After the code has been merged, a separate PR will be created for linting and will be added to a git-ignore list. The github action will be updated to ensure linting and tests passes.

[jdeath]

Do you plan on expanding the auto discovery? I do not support, so would be good if you can take requests to add support as well. All the external actions are way beyond my skill, so I will not be able to support them in the future

[Expl0dingBanana]

I plan on reducing the amount of code between the two discoveries. I am working on another branch locally where manual and auto more or less work the same way. The logic is as follows:

• Gather all devices from the API and index them based on id
• If CONF_FRIENDLYNAMES is populated, lookup all ids where the friendlyName matches
• If CONF_ROOMNAMES is populated, get children from room_data.children
• If neither CONF_FRIENDLYNAMES or CONF_ROOMNAMES is specified, use all hashed IDs

• Utilize a dataclass that contains all relevant information for HA. Post initialization will perform the device lookup code (such as converting to the DriskolFan).

  • Add all specified devices in light.py::setup_platform

  I am probably going to get rid of some of this code when finished since we do not need to breakout manual / auto. This is my first time working with HA so that may be a future thing supporting more services. Once this is all setup, I am going to look how to create devices from entities and then can probably look into services.

  PS - This fan was not worth the hassle 😂

[jdeath]

PS - This fan was not worth the hassle 😂

I know. I created this whole project for a single light I bought!

I merged your stuff. But I will wait on a new release till everything is done