* Added initial files for air-Q integration
* Allow FIXME comments in pylint (temporary)
Also reintroduce --ignore-missing-annotations=y
* Set up air-q entry to connect to aioairq's API (initial attempt)
Also add necessary constants
* Implement a class for sensors and its update logic
Very early stage, WIP
* Zeroconf and authentication are working
* Complete the bare-bone minimal working version
Specifically, make AirQSensor update its values.
* Handle invalid authentication gracefully
* Handle ClientConnectionError gracefully
* Add field hint for the login form
The key in the schema, which defines the form in
`ConfigFlow.async_show_form` is looked up in both `airq/strings/json`
and `airq/translations/en.json`. I am still not 100% sure how this
lookup is performed. WIP
* Minor cleanups
* Extend sensor list to all supported by SensorDeviceClass
Also manage warming up sensors
* aioairq is published to PyPI and mentioned in requirements
* Reordered constants and list content alphabetically
As required by style guides. Also turned SENSOR_TYPES to a list
* Updated file docstrings for a dev unfamiliar w/homeassistant like myself
* Adding a bit of logging for the integration setup process
* Expose scan interval & smoothing flag
Also streamline test_authentication in config_flow.
* Fix a type annotation mistake
* Use as many constants from homeassistant.const as possible
My only concern is using CONST_IP_ADDRESS = "ip_address" for smth which
stands for both IP address and mDNS...
* Temporarily rollback ConfigFlow.async_step_configure and use defaults
TODO: implement OptionFlowHandler instead
* Define custom Coordinator, w subset of airq config
The latter is then accessed from entity / sensor constructors to define
correct DeviceInfo
* Provide translations to de & fr + minor changes to en
* Provide translations to ru + a minor en changes
* Make translation a little more helpful and polite
* Fix devicename and entry title
* Remove stale commented out code
* Test config_flow
At this point two helper functions which interact with the external
library are not tested
* Clean up unrelated and meant as temporary changes
* Clean up unnecessary comments meant for internal use
* Move fetching config to a dedicated async coordinator method
As opposed to it being a potentially poorly justified step in
async_setup_entry
* Remove zeroconf support since it is not yet ready
* Remove translations other than en
* Remove unnecessary comments, manifest.json entries, and constants
* Improve exception handling
- `_LOGGER` uses `debug` and not `error` levels.
- Drop `ClientConnect` and catch `aiohttop.ClientConnectError` directly
- Drop `Exception` as it is not expected from `aioairq` (remove the
corresponding test too)
* Drop strings for obsolete errors and steps
Specifically, `unknown` error isn't caught any more. `configure` step
has also been removed.
* Refactor en.json to be consistent with strings.json
* Move target_route from a coordinator argument to a constant
At this point a user cannot configure the target_route route, thus it
does not make sense to expose it half-heartedly in
`AirQCoordinator.__init__`, since it cannot be accessed.
* Fix an async call in `AirQCoordinator.async_setup_entry`
* Refactor underlying aioairq API
- Use `homeassistant.helpers.aiohttp.async_get_clientsession` and pass a
single persistent session to `aioariq.AirQ.__init__`
- `aioairq.AirQ.fetch_device_info` now returns a `DeviceInfo` object heavily
inspired and almost compatible with `homeassistant.helpers.entity.DeviceInfo`.
Make heavier use of this object and define a single `DeviceInfo` in the
`AirQCoordinator` (instead of recreating the same object for each sensor
of the device in `sensor.AirQSensor`)
- Drop two helper functions in `config_flow.py` and operate on `aioariq.AirQ`
methods directly
* Fix the version of aioairq
* Add 15 more sensors + icons
* Remove cnt* & TypPS, change units of health & performance
* Add 12 more sensors
* Add a missing icon
* Gracefully handle device not being available on setup
If the device and the host are not on the same WiFi,
ServerTimeoutError is raised, which is caught by
ClientConnectionError.
If the device is powered off, ClientConnectionError is expected.
In both cases, ConfigEntryNotReady is raised, as prescribed by the docs.
Newer version of aioairq times-out far quicker than the default 5 mins.
* Rename two sensors
* Validate provided IP address / mDNS
aioairq now raises InvalidInput if provided IP / mDNS does not seem
valid. Handle this exception correctly
* Apply suggestions from code review
Clean up the comments and rename the logger
Co-authored-by: Erik Montnemery <erik@montnemery.com>
Co-authored-by: Artem Draft <Drafteed@users.noreply.github.com>
* Only fetch device info during the first refresh
- Fetched info is stored in AirQCoordinator.device_info.
- In `AirQSensor.native_value` only multiply by the factor if
the sensor reading is not None
- Fix the tests for ConfigFlow for aioairq==0.2.3. Specifically
make the dummy data pass the new validation step upstream
+ add a test which fails it
* Drop custom device classes for now
* Apply suggestions from code review
Co-authored-by: Artem Draft <Drafteed@users.noreply.github.com>
* Only fetch device info during ConfigFlow.async_step_user
Store the result obtained by `airq.fetch_device_info` it in
`config_entry.data`. Pass the entire config entry to `AirQCoordinator`
and build the entire `homeassistant.helpers.entity.DeviceInfo` in the
`AirQCoordinator.__init__`. This makes
`AirQCoordinator._async_fetch_device_info` and overloaded
`AirQCoordinator._async_config_entry_first_refresh` obsolete.
Bump aioairq version.
Turn update_interval from `AirQCoordinator.__init__` argument into a
contestant.
* Custom entity description exposing a hook to modify sensor value
Use a `AirQEntityDescription` with a callable `value_fn` which allows to
change the sensor value retrieved from the device. Note that the
callable does not handle data retrieval itself (even from
`coordinator.data`). Instead it is purely a hook to transform obtained
value.
* Avoid duplicated use of unique_id
Device info is fetched during the `ConfigFlow.async_user_step`.
`unique_id` is taken from the device info and is **not** stored
in `config_entry.data`. Subsequently `config_entry.unique_id` is
used instead.
* Drop unnecessary try-except
Co-authored-by: Artem Draft <Drafteed@users.noreply.github.com>
* Clarify the use of value_transform_fn
* Refactor the use of lambdas in AirQEntityDescription
Now it is the job of the callable under `value` to get the sensor
reading from the coordinator's data. Factoring this functionality into a
callback decouples the key of the description from the key of dict,
returned by the API, so `AirQEntityDescription` no longer requires its key
to be set to smth clearly internal (e.g. `nh3_MR100`).
* Use a callback to update native_value
Since all `native_value`s are updated synchronously, it can as well be
done in a callback for better state consistency (right?)
* Revert the description keys to match data keys
Must match given the current way of identifying available sensors. On a
broader scale, they must match to be able to relate the descriptions to
sensors, unless a separate lookup table is maintained.
* Reduce number of loops when adding sensors
Filtering warming up sensors and non-sensor keys can be combined with
adding entities.
* Remove obsolete imports
* Update integrations.json
* Add integration_type
Integration supports multiple devices => hub
Co-authored-by: dl2080 <daniel.lehmann@runbox.com>
Co-authored-by: Erik Montnemery <erik@montnemery.com>
Co-authored-by: Artem Draft <Drafteed@users.noreply.github.com>
Co-authored-by: Daniel Lehmann <43613560+dl2080@users.noreply.github.com>
Co-authored-by: Martin Selbmann <job@martin-selbmann.de>
* Bump bleak-retry-connector to 2.8.2
Tweaks for the esp32 proxies now that we have better error
reporting. This change improves the retry cases a bit with
the new https://github.com/esphome/esphome/pull/3971
* empty
* Restore homekit_controller BLE broadcast_key from disk
Some accessories will sleep for a long time and only send broadcasted
events which makes them have very long connection intervals to save
battery. Since we need to connect to get a new broadcast key we now
save the broadcast key between restarts to ensure we can decrypt
the advertisments coming in even though we cannot make a connection
to the device during startup. When we get a disconnected event later
we will try again to connect and the device will be awake which will
trigger a full sync
* bump bump
* mapping
* update
* working
* tests
* fixes
* temp add binary sensor platform to make sure it works
* fixes
* 100% cover
* adjust
* disable chatty sensors by default
* remove binary sensor_platform for next PR
* time is much nicer than counter
* Add Freebox button to mark calls as read
* Bump Freebox to 1.0.1
* Fix black/flake8
* Add entity_category + fix reboot button name
* Add has_entity_name to reboot
* Remove 'missed' as it put all calls as read
* base unique_id on key and not name
* unique_id base on key later with migration step
* Keep the same name
Co-authored-by: epenet <6771947+epenet@users.noreply.github.com>
Co-authored-by: epenet <6771947+epenet@users.noreply.github.com>
* Expose Nobø Switch as temperatur sensor.
* - Review
- Hub may report current temperature as None
* Avoid update during entity addition, and fix race condition
* Update pynobo to 1.6.0
Use new method to fix potential race condition.
* Use generator expressions
* Bump aioshelly to 4.0.0
* Remove leftover
* Fix number platform
* Set last_update_success to false upon failure in number and climate
* Set last_update_success upon failurie in entity
* Update bsblan integration
Update the integration to current standards
* removed unused code
update coverage
* some cleanup
* fix conflicts due upstream changes
* fix prettier json files
* fix remove comment code
* use dataclass instead of tuple
* fix spelling
* Set as class attribute
main entity doesn't need to give own name
* fix requirements
* Bump Wallbox package
* remove debug message
* Force update of auth token by emptying it first
* Force token refresh by emptying token
Improve exception handling
* include tests
* Update __init__.py
* Removed the clearing ot jwt token, issue is fixed by upstream fix in wallbox package.
* Catch connectionerror
* Update homeassistant/components/wallbox/__init__.py
Co-authored-by: Martin Hjelmare <marhje52@gmail.com>
* Run black
Co-authored-by: Martin Hjelmare <marhje52@gmail.com>
Using the ha- fork until upstream can pickup and merge
pending PRs. The plan is to revert back to upstream
HAP-python when its back in sync
Fixes#79305Fixes#79304
J. Nick Koston
Renat Sibgatulin
mkmer
Dennis Schroer
Austin Brunkhorst
Bram Kragten
Allen Porter
Poltorak Serguei
Franck Nijhof
Daniel Hjelseth Høyer
puddly
Shay Levy
Willem-Jan van Rootselaar
Tobias Sauerwein
Chris Talkington
aschmitz
Jafar Atili
Raj Laud
mezz64
Vincent Knoop Pathuis
René Klomp
Maciej Bieniek
muppet3000
Robert Hillis
Thibaut
Michał Huryn
Sean Vig
Avi Miller
Paulus Schoutsen
Daniel Gangl
David F. Mulcahey
Joakim Sørensen
Jesse Hills
Steven Looman
shbatm
Quentame
Oliver
Øyvind Matheson Wergeland
Kevin Addeman
Nathan Spencer
Robert Svensson
Fredrik Erlandsson
Bouwe Westerdijk
jjlawren
uvjustin
Timm Schäuble
luar123
niobos
Christopher Bailey
Marcus R. Brown
Florent Thoumie
Jo De Boeck
mvn23
Joakim Plate
Jeef
Austin Mroczek
mnorrsken
Marvin Wichmann
hesselonline
Dave T
Garrett
kingy444
Khole
Michael Molisani
majuss