Closed tobixen closed 1 year ago
Some observations:
copy()
without arguments, there seems to be sufficient test code for this case in the testskeep_uid=True
. Actually, there seems to be no tests for this case in the test code, but there are tests for keep_uid=True
in combination with new_parent
.event.parent
is None in the breakage, but is required to be set.event.parent
should always be set to the calendar the event belongs to, so one question here is how it became unset ... hm.
One approach may be to try to throw in some asserts in the code, catching instances where event.parent
is None and rerun the tests
Closing this one. I threw in plenty of assert in a side-branch to pinpoint the problem. The problem was due to test code in the HomeAssistant creating an Event without a parent.
ref https://github.com/home-assistant/core/pull/85536#issuecomment-1383034763 and https://github.com/home-assistant/core/actions/runs/3920550422/jobs/6702347492#step:10:8387