Fix fan modes

[gentoo-root]

Fix the ELECTRA_AC workaround that became a no-op in de5a190 (not tested with the real hardware).

Prettify the way how fan modes are displayed in the UI for ACs that use min/medium/max instead of low/medium/high.

[mateuszdrab]

Hey @gentoo-root

This is a pretty nice PR as I did notice my min and max options are not localized. We're investigating some issues with ARGO decoding in here.

I was wondering if you could also add handling for a mid-high state that Tasmota reports for my ARGO AC as no existing supported_fan_speeds options work with it.

This unit's fan speeds are auto, min, med-high, high.

If not, I'll create a PR after this one merges so that I can adapt handling of med-high so it shows correctly in the UI along with low and max which you corrected.

[hristo-atanasov]

The proposed changes are breaking another tweak for ELECTRA_AC, listed below, which is tested on a real device. This is why you think there is a redundant checks in the code. It is not redundant, it is for limiting the fix only for those devices, that needs it. Please, rewrite your PR code, without breaking the existing fix.

  supported_modes:
  # Some devices have "auto" and "fan_only" switched
  # If the following two lines are uncommented, "auto" and "fan" shoud be commented out
  #- "auto_fan_only" #if remote shows fan but tasmota says auto
  #- "fan_only_auto" #if remote shows auto but tasmota says fan

  supported_fan_speeds:
  # Some devices say max,but it is high, and auto which is max
  # If you uncomment the following two, you have to comment high and max
  # - "auto_max" #woud become max
  # - "max_high" #would become high

[gentoo-root]

The proposed changes are breaking another tweak for ELECTRA_AC, listed below, which is tested on a real device.

If it's tested and works on a real device, I'm curious to know how. One of my commits is actually supposed to fix ELECTRA_AC after a breaking change made in commit de5a190. As I explained in the commit message, the relevant ELECTRA_AC code flows became unreachable after de5a190, because HVAC_FAN_MAX_HIGH and HVAC_FAN_AUTO_MAX can't be in self._attr_fan_modes after __init__(), therefore all if HVAC_FAN_MAX_HIGH in (self._attr_fan_modes or []) and HVAC_FAN_AUTO_MAX in (self._attr_fan_modes or []) checks throughout the code are always false. Are you sure ELECTRA_AC was tested after commit de5a190? If yes, could you please explain where I'm wrong?

Assuming my understanding is correct, the code I removed in 8e4a1b7 can be removed safely:

1. Removal of dead code doesn't change the behavior, so it can't be breaking anything.
2. It's not needed after my next commit (which reenables ELECTRA_AC quirks) either, because the first check if fan_mode not in (self._attr_fan_modes or []) would cover if fan_mode != FAN_HIGH and fan_mode != HVAC_FAN_MAX, as both FAN_HIGH and HVAC_FAN_MAX will be in self._attr_fan_modes.
3. (Not really relevant with the current code, given the above, but if the ELECTRA_AC-specific error message was reachable, it would print a wrong list of fan modes, missing the explicitly enabled ones.)