* Switch back to using call_later for the slow entity update warning
I had originally changed this to create a task and wait
in #41184 but that does not make sense anymore with newer
cpython as the profile now shows the original method is cheaper
(or I did it wrong the first time)
* fix missing block till done since there is no longer a task being created which would run the event loop once
* Factor out _is_dmr_device function
* Use DMR device's MAC to match existing config entries
Some DMR devices change their every time they boot, against the DMR specs.
Try to match such devices to existing config entries by using their MAC
addresses.
* Add DMR device's MAC as a device_registry connection
* Use doc-only IPs (RFC5737) for dlna_dmr tests
* Fix incorrect types of test data structures
* Loosen MIME-type filtering for async_browse_media
* Add option to not filter results when browsing media
Some devices do not report all that they support, and in this case
filtering will hide media that's actually playable. Most devices are OK,
though, and it's better to hide what they can't play. Add an option, off by
default, to show all media.
* Fix linting issues
* Listen for config updates from DlnaDmrEntity.async_added_to_hass
Use `Entity.async_on_remove` for dealing with callback cancellation,
instead of re-inventing the wheel with `_remove_ssdp_callbacks`.
* Use async_write_ha_state within async methods
* Import YAML config from async_setup_platform
* Import flow prompts user when device is uncontactable during migration
When config flow is able to contact a device, or when it has information
from SSDP, it will create config entries without error. If the device is
uncontactable at this point then it will appear as unavailable in HA
until it is turned on again.
When import flow cannot migrate an entry because it needs to contact the
device and can't, it will notify the user with a config flow form.
* Don't del unused parameters, HA pylint doesn't care
* Remove unused imports from tests
* Abort config flow at earliest opportunity
* Return async_abort instead of raising AbortFlow
* Consolidate config entry test cleanup into a single function
* fixup! Consolidate config entry test cleanup into a single function
Revert "Consolidate config entry test cleanup into a single function"
This reverts commit 8220da7263.
* Check resource acquisition/release in specific tests
* fixup! Check resource acquisition/release in specific tests
* Remove unused network dependency from manifest
* _on_event runs in async context
* Call async_write_ha_state directly (not via shedule_update)
Co-authored-by: Martin Hjelmare <marhje52@gmail.com>
Co-authored-by: Martin Hjelmare <marhje52@gmail.com>
* Modernize dlna_dmr component: configflow, test, types
* Support config-flow with ssdp discovery
* Add unit tests
* Enforce strict typing
* Gracefully handle network devices (dis)appearing
* Fix Aiohttp mock response headers type to match actual response class
* Fixes from code review
* Fixes from code review
* Import device config in flow if unavailable at hass start
* Support SSDP advertisements
* Ignore bad BOOTID, fix ssdp:byebye handling
* Only listen for events on interface connected to device
* Release all listeners when entities are removed
* Warn about deprecated dlna_dmr configuration
* Use sublogger for dlna_dmr.config_flow for easier filtering
* Tests for dlna_dmr.data module
* Rewrite DMR tests for HA style
* Fix DMR strings: "Digital Media *Renderer*"
* Update DMR entity state and device info when changed
* Replace deprecated async_upnp_client State with TransportState
* supported_features are dynamic, based on current device state
* Cleanup fully when subscription fails
* Log warnings when device connection fails unexpectedly
* Set PARALLEL_UPDATES to unlimited
* Fix spelling
* Fixes from code review
* Simplify has & can checks to just can, which includes has
* Treat transitioning state as playing (not idle) to reduce UI jerking
* Test if device is usable
* Handle ssdp:update messages properly
* Fix _remove_ssdp_callbacks being shared by all DlnaDmrEntity instances
* Fix tests for transitioning state
* Mock DmrDevice.is_profile_device (added to support embedded devices)
* Use ST & NT SSDP headers to find DMR devices, not deviceType
The deviceType is extracted from the device's description XML, and will not
be what we want when dealing with embedded devices.
* Use UDN from SSDP headers, not device description, as unique_id
The SSDP headers have the UDN of the embedded device that we're interested
in, whereas the device description (`ATTR_UPNP_UDN`) field will always be
for the root device.
* Fix DMR string English localization
* Test config flow with UDN from SSDP headers
* Bump async-upnp-client==0.22.1, fix flake8 error
* fix test for remapping
* DMR HA Device connections based on root and embedded UDN
* DmrDevice's UpnpDevice is now named profile_device
* Use device type from SSDP headers, not device description
* Mark dlna_dmr constants as Final
* Use embedded device UDN and type for unique ID when connected via URL
* More informative connection error messages
* Also match SSDP messages on NT headers
The NT header is to ssdp:alive messages what ST is to M-SEARCH responses.
* Bump async-upnp-client==0.22.2
* fix merge
* Bump async-upnp-client==0.22.3
Co-authored-by: Steven Looman <steven.looman@gmail.com>
Co-authored-by: J. Nick Koston <nick@koston.org>
Erik Montnemery
Meow
J. Nick Koston
epenet
Franck Nijhof
Michael Chisholm
Steven Looman
Paulus Schoutsen
Robert Hillis
Diego Elio Pettenò
Michael Chisholm