`period_range` and `Period` should accept `date`, `datetime` and `pd.Timestamp`

[Daverball]

pd.period_range and pd.Period currently only accept str and pd.Period as arguments, but the implementation also accepts datetime, date and pd.Timestamp as can be seen here in the cython implementation: https://github.com/pandas-dev/pandas/blob/v2.0.1/pandas/_libs/tslibs/period.pyx#L2626-L2643

It technically also allows integers/floats. period_range internally uses Period to convert the start and end arguments, so the type annotations can be used interchangeably between those two.

To Reproduce

import pandas as pd

pd.Period(pd.Timestamp('2020-01-01'), freq='Q')  # mypy: arg-type — Argument 1 to "Period" has incompatible type "Timestamp"
pd.period_range(pd.Timestamp('2020-01-01), '2020-06-01', freq='Q')  # mypy: arg-type — Argument 1 to "period_range" has incompatible type "Timestamp"

• OS: Linux (OpenSUSE Tumbleweed)
• Python 3.10
• mypy 1.0.1
• pandas-stubs 2.0.1.230501

[Dr-Irv]

For the stubs, we generally follow what is documented, and while the implementation does accept datetime, date, and pd.Timestamp, and maybe even int or float, the docs just say Period or str, so we'd have to find out if that is an error in the implementation or a documentation issue.

Can you open up an issue in the pandas repo about the docs, and once that is resolved, we can address the stubs?

Thanks.

[Dr-Irv]

Can you open up an issue in the pandas repo about the docs, and once that is resolved, we can address the stubs?

Also, when you open that issue, please cross-reference this one.