Ronnie Dutta
Ronnie Dutta
E.g. for `TimePoint`: ```python-console >>> a = TimePoint(year=2022, hour_of_day=6) >>> b = a.copy(hour_of_day=0) ```
When adding a truncated `TimePoint` to a regular `TimePoint`, if the truncated one specifies a date, all units smaller than the ones provided are not taken into account, whereas if...
Instead of `CCYY`, it uses the year itself ```python >>> timepoint = TimePoint(year=2010, month_of_year=3, hour_of_day=10, time_zone_hour=-3) >>> timepoint._get_dump_format() '2010-MM-DDThh:mm:ss+hh:mm' ``` This might not be currently affecting anything if `_get_dump_format()` is...
Follow on from #162 Now that the data classes have been made hashable, the next step is to implement some form of `lru_cache` for some of the methods. I think...
Possibly related to #80 ```python >>> A = TimePoint(day_of_month=30, hour_of_day=9) ---30T09Z >>> A + Duration(hours=2) metomi/isodatetime/data.py in __add__(self, other, no_copy) 1323 if duration.hours: 1324 new.hour_of_day += duration.hours -> 1325 new.tick_over()...
... and a function for the reverse. See #196
Consider re-ordering `TimePoint` constructor kwargs so that `year`, `month_of_year`, `day_of_month`, `hour_of_day`, `minute_of_hour` and `second_of_minute` come first in that order, and making the others [keyword-only arguments](https://peps.python.org/pep-3102/). ``` >>> TimePoint(2020, 5, 5)...
We ought to make `isodatetime` a dead-end package which tells users to use `metomi-isodatetime` instead
From https://github.com/cylc/cylc-flow/issues/4908 When you add a conjugate Duration such as `P1M1D` to a TimePoint, it seems isodatetime does `+ P1D + P1M` internally, which I think is confusing. This can...
Different severity levels are all logged with the same colour in the logs, would be good to enable colour so that warnings and errors stand out.