* [mqtt.homeassistant] handle null component name
channels from such components will not have a group. this is
now done by zigbee2mqtt for the "default" component of a device,
such as the light. HASS encourages this as of release 2023.8
Signed-off-by: Cody Cutrer <cody@cutrer.us>
* [mqtt.homeassistant] support color temp on JSON schema lights
also adds a color_mode channel if color temp is possible, so you can
know how the bulb is behaving
* put color mode channel construction into buildChannels()
---------
Signed-off-by: Cody Cutrer <cody@cutrer.us>
Depending on device configuration and used central [1], HM devices may
indicate long press repetition either by a single PRESS_CONT event or by
a PRESS_CONT + PRESS_LONG combination. In the latter case, make sure to
not generate a LONG_REPEATED trigger channel event for both PRESS_CONT
and PRESS_LONG, but instead keep LONG_REPEATED generation to the
PRESS_CONT handling.
[1] I'm not sure what a real CCU is doing, but for Homegear, a
configured long press timeout of less than 1s generates only
PRESS_CONT, while a timeout of more than 1s generates
PRESS_CONT + PRESS_LONG ... see [2].
[2] https://github.com/Homegear/Homegear-HomeMaticBidCoS/blob/master/src/BidCoSPeer.cpp#L1711-L1716
Signed-off-by: Danny Baumann <dannybaumann@web.de>
* Fix resource leak when discovery handling failed and NPE when a IPv6
address is reported.
* remove callApi(request/innerRequest from Shelly1HttpApi, which changes
to callApi/httpRequest in ShellyHttpCLient.
Signed-off-by: Markus Michels <markus7017@gmail.com>
* New translations sonos.properties (French)
* New translations influxdb.properties (French)
* New translations netatmo.properties (French)
* New translations jpa.properties (French)
* New translations chatgpt.properties (French)
It seems that Bountysource has been shut down as the website isn't reachable anymore and they don't respond to support mails.
Signed-off-by: mueller-ma <mueller-ma@users.noreply.github.com>
* [influxdb] Improve connection handling
Especially for InfluxDB2 the connection check was not properly implemented. It only checked if a connections was ever successfully established. Since we removed the full crash when a write error occured, this lead to a situation where a broken connection was not detected. A ping is now implemented and also a failed write results in a disconnect.
---------
Signed-off-by: Jan N. Klug <github@klug.nrw>
* [fineoffsetweatherstation] Improve tracing
* [fineoffsetweatherstation] Fix wrong handling temperature reading for of WH34
Signed-off-by: Andreas Berger <andreas@berger-freelancer.com>
* Added a few new channels:
- cleaning the mop
- check if mop drying is enabled
- remaining time for mop drying
Signed-off-by: David Kumar <github@truidix.de>
Some devices report "Window covering controller" instead of "Window covering device". They are controlled in the same way.
Signed-off-by: Jan N. Klug <github@klug.nrw>
