Fix roborock timers' next_schedule on repeated requests

[phil9909]

Every call to get_next of a croniter instance returns a different timestamp (because it's a cron iterator).

from croniter import croniter
from datetime import datetime

cron = croniter("* * * * *", datetime(2022, 1, 1))

cron.get_next(ret_type=datetime) # returns datetime.datetime(2022, 1, 1, 0, 1)
cron.get_next(ret_type=datetime) # returns datetime.datetime(2022, 1, 1, 0, 2)
cron.get_next(ret_type=datetime) # returns datetime.datetime(2022, 1, 1, 0, 3)

Because of this Timer.next_schedule will return different values each time. I don't think this is intended.

This PR will fix this behavior.

[codecov-commenter]

Codecov Report

Merging #1520 (a85e42f) into master (207a053) will increase coverage by 0.05%.
The diff coverage is 90.90%.

@@            Coverage Diff             @@
##           master    #1520      +/-   ##
==========================================
+ Coverage   82.15%   82.21%   +0.05%     
==========================================
  Files         145      145              
  Lines       14161    14168       +7     
  Branches     3416     3418       +2     
==========================================
+ Hits        11634    11648      +14     
+ Misses       2302     2295       -7     
  Partials      225      225              

Impacted Files	Coverage Δ
...o/integrations/vacuum/roborock/vacuumcontainers.py	80.71% <90.90%> (+3.05%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[phil9909]

Had to add a tests to make codecov happy.

While doing this I fixed a typing mistake in the init function of the Timer class (use pytz.BaseTzInfo instead of datetime.tzinfo which seems more accurate).

[rytilahti]

LGTM, the 3.11-dev is sometimes flaky so let's ignore that for now. Thanks @phil9909! 👍

[phil9909]

Any plans to create a release soon?

[rytilahti]

Hi @phil9909, the current master has diverged so much that it's not a simple task to create a new release from it, but I might consider creating a new point release based on the latest release and some cherry-picked fixes if there is an urgent need that cannot wait?