Update unit tests to pass in different timezones #323
Conversation
|
That _Utc class was totalled not needed. It is much simpler now. I understand that this doesn't interact well with #89, but I'm looking for a smaller issue to chew on for my first contribution. :) Tested in the following timezones: |
| utc_time = local_time - datetime.timedelta(seconds=time.timezone) | ||
| expected_timestamp = time.mktime(utc_time.timetuple()) | ||
| utc_time = datetime.datetime(2012, 1, 14) | ||
| expected_timestamp = calendar.timegm(utc_time.timetuple()) |
There was a problem hiding this comment.
This seems like a sorta clunky way to get a timestamp for UTC, how about this:
utc_time = datetime.datetime(2012, 1, 14, tzinfo=datetime.timezone.utc)
expected_timestamp = utc_time.timestamp()
| datetime_from_time = datetime.datetime.fromtimestamp(time.time()) | ||
| timezone_adjusted_datetime = datetime_from_time + datetime.timedelta(seconds=time.timezone) | ||
| assert timezone_adjusted_datetime == expected_datetime | ||
| datetime_from_time = datetime.datetime.utcfromtimestamp(time.time()) |
There was a problem hiding this comment.
I'm not convinced that this is the right thing to do, but I also don't think the assertion this replaces makes much sense.
It seems to be asserting that datetime.datetime.now() should not be equal to datetime.fromtimestamp(time.time()) if the test is run in a locale with an offset, which is not the same behavior as what datetime.datetime does. Fixing this assumption in the tests (and code?) is how this should be fixed, I think.
|
I think it would probably be a good idea to add a CI job that runs the tests in some non-UTC locale (or possibly in a few locales). Presumably it's OK to just use whatever the latest Python is and run the test suite in a |
Fixes #322