BUG: Assigning datetime to pa.date32() type array should raise

[WillAyd]

Pandas version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

import pandas as pd
import pyarrow as pa

pa_arr = pa.array([
    datetime.date(2024, 1, 1),
    datetime.date(2024, 1, 2),
    datetime.date(2024, 1, 3),
])
ser = pd.Series(pa_arr, dtype=pd.ArrowDtype(pa.date32()))
ser.iloc[0] =  datetime.datetime(2024, 12, 31, 12, 20, 0)

Issue Description

Assigning a datetime value to a pyarrow date array type seems to implicitly drop the time components

Expected Behavior

Should raise TypeError

Installed Versions

In [24]: pd.version
Out[24]: '3.0.0.dev0+681.g434fda08cf'

In [25]: pa.version
Out[25]: '15.0.0'

[udit5656]

take

[PedroVerardo]

take

[PedroVerardo]

Hi @WillAyd, I'm new to the pandas repository, and I think I've found the problem with this issue. However, I don't know exactly how to change this transformation in the code.

This happens in the ArrowExtensionArray, on the function _box_pa_scalar

pa_scalar = pa.scalar(value, type=pa_type, from_pandas=True)

The only possibility I see is to remove this and raise an error

[PedroVerardo]

I made a change that solves the issue specifically in the "Reproducible Example," but I think there are some similar problems that need to be solved. It is necessary to explore this problem more.

For example, if you do the opposite, as you do in this example, the same thing happens, but this time, 'ns' will be inferred.

import pandas as pd
import pyarrow as pa
import datetime

pa_arr = pa.array([
    datetime.datetime(2024, 1, 1, 1, 1, 1),
    datetime.datetime(2024, 4, 5, 1, 1, 1),
    datetime.datetime(2024, 6, 7, 1, 1, 1),
])
ser = pd.Series(pa_arr, dtype=pd.ArrowDtype(pa.timestamp('ns')))
ser.iloc[0] = datetime.datetime(2024, 12, 31)
print(ser)

If needed, I would love to help @WillAyd.

[WillAyd]

Looks like this is an issue upstream in arrow - see apache/arrow#41896