Detect offline hub children and fail fast on writes to them (#1830)
Some checks failed
CI / Perform Lint Checks (3.15) (push) Has been cancelled
CodeQL Checks / Analyze (python) (push) Has been cancelled
CI / Python 3.11 on macos-latest (push) Has been cancelled
CI / Python 3.12 on macos-latest (push) Has been cancelled
CI / Python 3.13 on macos-latest (push) Has been cancelled
CI / Python 3.14 on macos-latest (push) Has been cancelled
CI / Python 3.15 on macos-latest (push) Has been cancelled
CI / Python 3.11 on ubuntu-latest (push) Has been cancelled
CI / Python 3.12 on ubuntu-latest (push) Has been cancelled
CI / Python 3.13 on ubuntu-latest (push) Has been cancelled
CI / Python 3.14 on ubuntu-latest (push) Has been cancelled
CI / Python 3.15 on ubuntu-latest (push) Has been cancelled
CI / Python 3.11 on windows-latest (push) Has been cancelled
CI / Python 3.12 on windows-latest (push) Has been cancelled
CI / Python 3.13 on windows-latest (push) Has been cancelled
CI / Python 3.14 on windows-latest (push) Has been cancelled
CI / Python 3.15 on windows-latest (push) Has been cancelled
Stale / stale (push) Has been cancelled

Hubs keep reporting for children they have lost contact with, so callers
see stale values as if they were live. Trying to write to such devices
fail after timeouts with a generic error.
This PR exposes the offline/online information as reported by the hub,
and refuses writes to unreachable child devices without waiting for
device errors.

In details:
- Add new API: `Device.is_available`; hub children report unavailable
when the hub marks them offline or no longer lists them.
- New feature `available` for child devices to make it visible to the
CLI.
- Refuse writes to unavailable children with the new
`DeviceUnavailableError` instead of waiting for the hub to time out.
This commit is contained in:
Teemu R.
2026-11-01 18:18:03 +01:00
committed by GitHub
parent 81243ed87f
commit a68f2b3304
13 changed files with 229 additions and 16 deletions

View File

@@ -122,6 +122,12 @@
:undoc-members:
```
```{eval-rst}
.. autoclass:: kasa.exceptions.DeviceUnavailableError
:members:
:undoc-members:
```
```{eval-rst}
.. autoclass:: kasa.exceptions.UnsupportedDeviceError
:members:

View File

@@ -137,6 +137,8 @@ The base exception for all library errors is {class}`KasaException <kasa.excepti
- If the device returns an error the library raises a {class}`DeviceError <kasa.exceptions.DeviceError>` which will usually contain an ``error_code`` with the detail.
- If the device fails to authenticate the library raises an {class}`AuthenticationError <kasa.exceptions.AuthenticationError>` which is derived
from {class}`DeviceError <kasa.exceptions.DeviceError>` and could contain an ``error_code`` depending on the type of failure.
- If a command is sent to a hub child that is not available (see {attr}`Device.is_available <kasa.Device.is_available>`) the library raises a {class}`DeviceUnavailableError <kasa.exceptions.DeviceUnavailableError>`,
derived from {class}`DeviceError <kasa.exceptions.DeviceError>`, without contacting the device.
- If the library encounters and unsupported deviceit raises an {class}`UnsupportedDeviceError <kasa.exceptions.UnsupportedDeviceError>`.
- If the device fails to respond within a timeout the library raises a {class}`TimeoutError <kasa.exceptions.TimeoutError>`.
- All other failures will raise the base {class}`KasaException <kasa.exceptions.KasaException>` class.