* Adjust font text such that it won't be drawn inside the bow line in image_processing.draw_box
* Adjust font_height after actually counting the pixels
* Thinned out line_width and adjusted font size
* Fix for occupancy sensor unique_id
* Add check for sensor attributes before adding entity
* Fixes temperature key issue
* Clean up code with better use of keys
* Code clean up
* Support Huawei LTE SSDP discovery
* Avoid KeyError on simultaneous user initiated flow
Co-Authored-By: Paulus Schoutsen <paulus@home-assistant.io>
* Format code
* Add already configured check
* Initialize context in test flows
* Move deviceType match to manifest
* Update generated.ssdp
* Add SSDP config flow test case
* Remove stale debug print from tests
Set auto mode should be exposed to all dyson fans (e.g. *Pure Cool Link* and *Pure Hot+Cool Link*) instead of only *Pure Cool*, as it is support in all of the models (i.e. similar to the set night mode).
* docs(ifttt): add information to services.yaml
* docs(ifttt): start examples lowercase
* docs(ifttt): start examples with capital letter
* docs(ifttt): end description on period
* Update Vivotek component
Fix building stream URL
* Update Vivotek component
Make stream path optionally configurable.
* Update Vivotek camera integration
Use f-string to build stream source URL.
This improve readability and I hear it runs faster.
* SSDP matching improvements
- support multiple match groups per domain
- require matches on all, not any item in a group
- support matching on all UPnP device description data
* Manifest structure fixes
This change addresses a bug where one is unable to change directly between the away and temperature hold presets, as temperature hold cannot be enabled on a Venstar thermostat if away mode is active.
Furthermore, this change removes redundant state checks as the set_away and set_schedule calls are idempotent in the venstarcolortouch library. See https://github.com/hpeyerl/venstar_colortouch/blob/master/src/venstarcolortouch/venstarcolortouch.py#L275.
* Prevent TypeError when KNX RGB(W) light value contains None.
* Pylint doesn't like 'w' as variable name, therefore using 'white' instead.
* Simplified code as suggested by pvizeli.
* Imported homeassistant.util and slugified address if no name is specified
* Added a custom validator function in case name is not set in config
* Removed logger.debug line only used for testing
* Clean up Xiaomi Air Quality Monitor support (cgllc.airmonitor.b1)
* Remove unused variable
* Provide a proper unique_id
* Incorporate review
* Wrap the method that cause the exception
* Undo mistakenly changed file. Fixed in the separate PR.
* convert keyboard_remote to async and add possibility to trigger on multiple event types, as well as emulate key hold events
* update requirements
* cleanup shutdown handling and config handling as well as address other minor comments
* cleanup unused return values and debug message formatting
* move start and stop event listen to separate coroutine plus minor cleanup
* make setup coroutine a function
* fix import order and attribute defined outside of init
* add to codeowners
* update codeowners
* Add and improve core and config_entries type hints
* Complete and improve config_entries type hints
* More entity registry type hints
* Complete helpers.event type hints
* Reorg and test attribute for HomematicIP Cloud
* Add dutyCycle check to security_group
* Edit test to improve coverage
* Add missing flow test
* apply suggestion
Co-Authored-By: Martin Hjelmare <marhje52@kth.se>
* fix assert condition
* Added Unifi Led
* fixed manifest
* fixed style issue
* removed unused setting
* added sugested changes.
* fixed order
* fixed settings that are required
* Fix review issues
* fix variable name that was too short
* Testing something
* Reverted to a previous version for testing
* Reverted testing changes.
* Add available status and increase version of unifiled package version.
* No io in init function.
* Add above and below to sensor condition extra_fields
* Change unit_of_measurement to suffix in extra_fields
* Check if sensor has unit when getting capabilities
* Improve tests
* Update ThermostatController to map directives to supported modes and add support for CUSTOM mode.
* Removed erroneous config value from test.
* Removed unnecessary use of a comprehension by dumbing down comment so pylint could comprehend.
* Removed erroneous import variable caused by removing erroneous config value from test.
* Removed unnecessary use of a comprehension.
* Reverted Removal or erroneous import variable and erroneous config value from test.
Apparently need for additional tests outside this component. Whoops.
* Update pycsspeechtts
From 1.0.2 to 1.0.3 as the old one is using an api that doesn't work
* Give a option to choose region
Api is now region dependent, so gave it a config
* Save client identifier from auth for future use
* Use websocket events to update Plex
* Handle websocket disconnections
* Use aiohttp, shut down socket cleanly
* Bad rebase fix
* Don't connect websocket during config_flow validation, fix tests
* Move websocket handling to external library
* Close websocket session on HA stop
* Use external library, revert unnecessary test change
* Async & lint fixes
* Clean up websocket stopper on entry unload
* Setup websocket in component, pass actual needed object to library
* added automatic protocol detection
* fix logger tests
* fix async tests
* add missin const.py
* fix log formatting
* wait for first update call
* migrate first tests
* migrated all test functions
* started to use state machine
* updated all tests to use async_setup_component
* slove hints
* update tests
* get state at correct position
* remove impossible tests
* fix autodetect tests
* use caplog fixture
* add test for duplicate
* catch concrete exceptions
* don't mock samsungctl exceptions
* add test for discovery
* get state when possible
* add test for autodetect without connection
* Using latest winthings_api module.
Drastically reduced complexity of tests.
* Removing import source.
* Fixing test requirements.
* Using requests_mock instead of responses module.
* Updating file formatting.
* Removing unused method.
* Adding support for new OAuth2 config flow.
* Addressing PR feedback.
Removing unecessary base_url from config, this is a potential breaking change.
* Addressing PR feedback.
* Modernization rework
- config entry support, with override support from huawei_lte platform in YAML
- device tracker entity registry support
- refactor for easier addition of more features
- internal code cleanups
* Remove log level dependent subscription/data debug hack
No longer needed, because pretty much all keys from supported
categories are exposed as sensors.
Closes https://github.com/home-assistant/home-assistant/issues/23819
* Upgrade huawei-lte-api to 1.4.1
https://github.com/Salamek/huawei-lte-api/releases
* Add support for access without username and password
* Use subclass init instead of config_entries.HANDLERS
* Update huawei-lte-api to 1.4.3 (#27269)
* Convert device state attributes to snake_case
* Simplify scanner entity initialization
* Remove not needed hass reference from Router
* Return explicit None from unsupported old device tracker setup
* Mark unknown connection errors during config as such
* Drop some dead config flow code
* Run config flow sync I/O in executor
* Parametrize config flow login error tests
* Forward entry unload to platforms
* Async/sync fixups
* Improve data subscription debug logging
* Implement on the fly add of new and tracking of seen device tracker entities
* Handle device tracker entry unload cleanup in component
* Remove unnecessary _async_setup_lte, just have code in async_setup_entry
* Remove time tracker on unload
* Fix to not use same mutable default subscription set for all routers
* Pylint fixes
* Remove some redundant defensive device tracker code
* Add back explicit get_scanner None return, hush pylint
* Adjust approach to set system_options on entry create
* Enable some sensors on first add instead of disabling everything
* Fix SMS notification recipients default value
* Add option to skip new device tracker entities
* Fix SMS notification recipient option default
* Work around https://github.com/PyCQA/pylint/issues/3202
* Remove unrelated type hint additions
* Change async_add_new_entities to a regular function
* Remove option to disable polling for new device tracker entries
* Added Unifi Led
* fixed manifest
* fixed style issue
* removed unused setting
* added sugested changes.
* fixed order
* fixed settings that are required
* Fix review issues
* fix variable name that was too short
* Testing something
* Reverted to a previous version for testing
* Reverted testing changes.
* cover all possible values for operation_mode
* Update climate.py
* Update climate.py
* Update climate.py
mapped homeassistant constants to client ones so we don't have to check for both
* black + pylint
* move mode_map to __init__()
* Update climate.py
* Update climate.py
The new release fixes a single regression from requests to aiohttp conversion.
Some devices do not respond with the correct mimetype which was not enforced
by requests but is enforced by aiohttp.
Related PR https://github.com/rytilahti/python-songpal/pull/59
* Config entry and device for Coolmaster integration
* Lint/isort/flake/etc...
* Black formatting
* Code review fixes
* Config flow tests for coolmaster
* Add pycoolmaster requirement to test
* Remove port selection from Coolmaster config flow
* Update config_flow.py
* More idoimatic hash concat
* Correctly compute the supported_features in cover.mqtt
* Update homeassistant/components/mqtt/cover.py
Co-Authored-By: Paulus Schoutsen <paulus@home-assistant.io>
* Correctly compute the supported_features in cover.mqtt
* Format
* Support for additional timeline events
* Update __init__.py
* Expose details on user
These lines expose apptype and event_by, which can be used to give information on events initiated by keypad users (vs. users on the mobile app, web app, or those initiated from HA through abodepy).
* removed a nesting level
* Lutron Pico fix
* Reverted logging change
Was unaware that f-strings aren't used in logging commands, reverted the usage
* Reverted logging change
Was unaware that f-strings aren't used in logging commands, reverted the usage
* fixed logic
* Move imports in nuheat component
* Fix tox tests
* Fix tox tests
* Update tests/components/nuheat/test_init.py
@Balloob suggested the change because direct replacement, the mock would never be reverted and impact the other tests.
Co-Authored-By: Paulus Schoutsen <paulus@home-assistant.io>
* Added missing Alexa.ChannelController functions. Specifically ChangeChannel
and SkipChannel commands. These functions will call the play_media function
in a media_player app if it has the capability published and pass on the
channel# or channel name. The selected media player can then use this to
select the channel on the device it is associated to.
Modified the existing Alexa.StepSpeaker Setvolume function to actually do
a stepped volume change using the steps sent by Alexa. The Alexa default
step of 10 for a simple volume up/down can be changed via an exposed
media_player attribute called volume_step_default.
The default is set to 1. Any other value then default will be sent
as sequential volume up /down to the media_player.
* The test code has some weird behaviour with passed boolean values. Had to surround them in quotes for the tests to pass properly.
* Reverted test_smart_home.py change. Issue was not the boolean value but the behavior in the handler. The test suite does not like multiple await calls in a loop. Will investigate further. The handler code works though.
* Added ChannelController/SkipChannels test in test_smart_home.py
Added test for callSign payload attribute.
* Modified smart home test to allow more than one call to services
* Added more tests for ChannelChange functions for various payload options.
Removed name options from metadata payload section. not needed.
* Reverted assert call change in alexa test __init__.py back to ==1. Not sure if it was the cause of the pytest's failing on github
* Corrected a comment. First commit after a rebase.
* Comment line change. Also wanted to force a code check on github.
* Added a loop delay in StepSpeaker and SkipChannel functions for safety
* Removed uneeded sleep from for loops. Let remote handle delays
Moved service type decision out of for loops in ChannelController and StepSpeaker
Used constants instead of numeric values for support options in test module
* Change media_player const import to be more specific in source
* Modifed test_smart_home to use media_play constants instead of hardcode valu
* Removed unecessary test volume_step_default attribute from test_smart_home
* Removed uneeded comment in StepSpeaker function.
Re-ordered constants in test_smart_home.py
* Modified call to media_player play_media service to use media_player constant instead of hard coded value.
* Changed constant use to be consistant with rest of function.
* Correct merge conflicts in handlers.py and capablities.py
* Fix#28104 - CalDav support for floating datetimes
Timzones are optional in CalDav
It is possible that an entry contains neither a TZID, nor is an UTC time.
When this is the case, it should be treated as a floating date-time value,
which represent the same hour, minute, and second value regardless of which
time zone is currently being observed.
For Home-Assistant the correct timezone therefore is whatever is configured
as local time in the settings.
See https://www.kanzaki.com/docs/ical/dateTime.html
* Revert "Fix #28104 - CalDav support for floating datetimes"
This reverts commit cf32a6e390.
* add test case: floating events fail with error without patch
* Fix#28104 - CalDav support for floating datetimes
Timzones are optional in CalDav
It is possible that an entry contains neither a TZID, nor is an UTC time.
When this is the case, it should be treated as a floating date-time value,
which represent the same hour, minute, and second value regardless of which
time zone is currently being observed.
For Home-Assistant the correct timezone therefore is whatever is configured
as local time in the settings.
See https://www.kanzaki.com/docs/ical/dateTime.html
* style fix
* Add support for SmartStrip type devices (HS300, HS107) to tplink component
* Incorporate feedback from @MartinHjelmare using changes suggested by @shbatm
- Setting `_state` now uses a list comprehension
- `_alias` will use aliases from the Kasa app
- `_device_id` will be set to `_mac` for single plugs to retain backwards compatibility
Zach
Jeff Irion
Grodesh
Paulus Schoutsen
HomeAssistant Azure
Aaron Bach
Pascal Vizeli
Santobert
Quentame
Jesse Hills
Bouwe Westerdijk
Marius Flage
Thom Troy
shred86
Bram Kragten
kennedyshead
Martin
chriscla
Ville Skyttä
Charles Garwood
Jess
SukramJ
Luca Zimmermann
Oscar Tin Lai
Jonas Janz
Tobias Efinger
Florent Thoumie
Fabian Affolter
Tom Robinson
Kevin McCormack
Maciej Bieniek
Joakim Plate
Tim McCormick
András Rutkai
michaeldavie
Nash Kaminski
ktnrg45
Mister Wil
Robert Svensson
jjlawren
SukramJ
phispi
John Mihalic
Robin Pronk
Michael Schoonmaker
Daniyar Yeralin
Malte Franken
escoand
thoscut
gngj
Nikolay Vasilchuk
Tsvi Mostovicz
ZiroNL
Robert Van Gorkom
Alexei Chetroi
ochlocracy
Mark Coombes
fredericvl
Steve M
Erik Montnemery
springstan
David Bonnes
Sebastian Muszynski
Renaud Martinet
Jonas Janz
Franck Nijhof
Josh Bendavid
Yann Jajkiewicz
Eliseo Martelli
Sören
J.P. Hutchins
fwestenberg
Floris Van der krieken
Markus Pöschl
Hayley McIldoon
On Freund
guillempages
Joakim Sørensen
Diefferson Koderer Môro
Otto Winter
Floris Van der krieken
Tyler Page
Teemu R
Rami Mosleh
Villhellm
Adrien Foulon
libots
Jon Gilmore
Marius Flage
Alain Turbide
rolfberkenbosch
Per-Øyvind Bruun
Ernst Klamer
Matt Schmitt
Lukas
javicalle
Matt Kasa
Fredrik Erlandsson
Raman Gupta