feat: basic support for hello fairy #56
Conversation
src/led_ble/led_ble.py
Outdated
| model_num = 0 | ||
| if self._is_hello_fairy(): | ||
| if data[0] == 0xAA: | ||
| if data[1] == 0x00: # hw info | ||
| if len(data) > 7: | ||
| version_string = data[3:8].decode("ascii") | ||
| _LOGGER.debug("version %s", version_string) | ||
| self._state = replace( | ||
| self._state, | ||
| version_num=(data[3] - 48) * 100 | ||
| + (data[5] - 48) * 10 | ||
| + (data[7] - 48), | ||
| ) | ||
| if len(data) > 12: | ||
| model = data[8:13].decode("ascii") | ||
| _LOGGER.debug("model %s", model) | ||
| if len(data) > 24: | ||
| lights = data[24] # guessing | ||
| _LOGGER.debug("lights %d", lights) | ||
| if len(data) > 33: | ||
| effects = data[33] # guessing | ||
| _LOGGER.debug("effects %d", effects) | ||
|
|
||
| if data[1] == 0x01: # state info | ||
| if len(data) > 6: | ||
| self._state = replace(self._state, power=data[6] > 0) | ||
| else: | ||
| if len(data) == 4 and data[0] == 0xCC: | ||
| on = data[1] == 0x23 | ||
| self._state = replace(self._state, power=on) | ||
| return | ||
| if len(data) < 11: | ||
| return | ||
| model_num = data[1] | ||
| on = data[2] == 0x23 | ||
| preset_pattern = data[3] | ||
| mode = data[4] | ||
| speed = data[5] | ||
| r = data[6] | ||
| g = data[7] | ||
| b = data[8] | ||
| w = data[9] | ||
| version = data[10] | ||
| self._state = LEDBLEState( | ||
| on, (r, g, b), w, model_num, preset_pattern, mode, speed, version | ||
| ) | ||
|
|
There was a problem hiding this comment.
It would be better to abstract this into a new method in the protocol so each protocol can handle this different
There was a problem hiding this comment.
Agreed. We don't control the ProtocolBase class though, it comes from flux_led. One option would be to make a device class to encapsulate the differences between the flux_led devices and the hello fairy devices. This would include the (send) protocol, this notification_handler, and (to your other comments), the status command, and protocol_for_version_num.
There was a problem hiding this comment.
Its no problem to merge a PR to flux_led
There was a problem hiding this comment.
I cleaned up flux_led for Python 3.9+ and released 1.1.0
There was a problem hiding this comment.
do you want the new protocol method to return LEDBLEState (so move that class into flux-led)? or a generic dict?
| await self._send_command_while_connected([STATE_COMMAND]) | ||
| if self._is_hello_fairy(): | ||
| await self._send_command_while_connected( | ||
| [b"\xaa\x00\x00\xaa"] | ||
| ) # get version and capabilities | ||
| else: | ||
| await self._send_command_while_connected([STATE_COMMAND]) |
There was a problem hiding this comment.
Please add a new method to the protocol to get the state command instead
There was a problem hiding this comment.
that method is implemented but the protocol is not resolved yet at this point
bdraco
changed the title
|
I'm about to have family in town for the week so I'll likely miss GitHub notifications. If I don't respond for a day or so, feel free to ping me on discord (same handle) |
Support for addressable LED strings that use the 'hello fairy' app. The protocol is different but pretty analogous to the existing protocols.
I've tested this with HA. Very minimal changes are needed there to use these changes, as I'm sure you know.
I don't know if you're interested in expanding your library in this way. If not, feel free to close this and I'll just work with the fork (and I'll have to make a new HA integration essentially the same as led-ble, in that case).
This code would be cleaner if some refactoring were done to generalize some things and encapsulate device differences, e.g. response protocol, status request command, parameter scaling. I didn't do any of that because I didn't want to change more than necessary without coordination, but I could look at that.