A little heads up about daikin_madoka with latest HA versions due to bleak changes

[maraujop]

Hello there, I was trying to setup daikin madoka in my HA and found your fork, nice work

Core 2025.8.0
Supervisor 2025.07.3
Operating System 16.0

When trying to add integration of your fork after downloading it with HACs (very nice btw). I got error {“message”:”Invalid handler specified”}”

Logs of my system were showing:

Error occurred loading flow for integration daikin_madoka: cannot import name 'discover' from 'bleak' (/usr/local/lib/python3.13/site-packages/bleak/__init__.py)

Turns out latest pymadoka version uses here from bleak import discover I have opened a patch here, FYI mduran80/pymadoka#24

I fixed it by entering in the homeassistant docker container docker exec -ti homeassistant bash editing file /usr/local/lib/python3.13/site-packages/pymadoka/connection.py and removing the ,discover from there

This is more of a note for future people that have issues. You could have a monkey-path of pymadoka.discovery so that it doesn't break meanwhile if you are interested.

Doing the manual change, allows daikin Madoka integration to start.

[maraujop]

One more thing. Finally after adding the device and connecting to it, it was creating a storm of connections

Registrador: pymadoka.connection
Fuente: runner.py:156
Ocurrió por primera vez: 17:51:09 (66 ocurrencias)
Última vez registrado: 17:52:14

[org.bluez.Error.Failed] Operation already in progress
[org.bluez.Error.Failed] br-connection-canceled

I read your troubleshooting guide, but none of the commands were working in my RPi 4 HA OS. So I looked into it, and changed in /config/.storage/core.config_entries, looked for daikin_madoka line and changed these 2 parameters:

"scan_interval": 20
"force_update": false

then restarted with ha core restart this creates a stable connection without constant retries.

[GioScatta]

Hi @maraujop, I have the same problem, but I'm unable to edit myself the file connection.py, can you kindly explain to me how to solve the issue? My config is HA OS on a NUC, right now my only option is to restore an old backup. Thank you in advance.

[gganagno]

I managed to overcome the bleak error by patching the bleak import before the pymadoka import in the init.py file

# daikin_madoka\__init__.py
# Patch bleak import for pymadoka
try:
    from bleak import BleakScanner
    async def discover():
        return await BleakScanner.discover()

    # Create a fake bleak module with a discover symbol
    import bleak
    bleak.discover = discover
except Exception as e:
    print(f"Failed to patch bleak: {e}")
    
from pymadoka import Controller, discover_devices, force_device_disconnect