WLM control by task handler #27
Conversation
| self._message_queue: MessageQueue = settings.MESSAGE_QUEUE | ||
| self._measure_queue: MeasureQueue = MeasureQueue() | ||
| self._channel_to_period: dict[int, timedelta] = {} | ||
| self._open_connection() | ||
|
|
||
| def _open_connection(self): |
There was a problem hiding this comment.
This method may cause several errors, e.g., there is no Config, failed to open WLM.
In fact, regarding WLM, it is always possible to fail the API call as the USB connection goes wrong sometimes.
I think we should carefully handle such error cases.
There was a problem hiding this comment.
Maybe modeling this handler as an FSM would be helpful..?
There was a problem hiding this comment.
Umm.. Indeed, I didn't consider any error on purpose.
To send the occurred error info to the request handler, we need to implement more APIs, hence I'd like to handle this after implementing main features.
What do you think about this? @kangz12345
There was a problem hiding this comment.
Alright then, for now the structure might vary in the future, so let's handle the errors later.
| self._wlm.close() | ||
|
|
||
| def _start_channel_measurement(self, channel: int): | ||
| setting = Setting.objects.filter(channel__name=channel).order_by('-created_at').first() |
There was a problem hiding this comment.
Let's keep this in mind: the filtering might cost quite a lot in the future.
For now I think it would be fine.
Co-authored-by: Jiyong Kang <[email protected]> See also: #6
This resolves #6.
I couldn't test these features since it needs a WLM device, so please check the codes carefully.