[SPARK-55624][PS][TESTS] Ignore ArrowDtype in tests with pandas 3

[ueshin]

What changes were proposed in this pull request?

Ignores ArrowDtype in tests with pandas 3.

Why are the changes needed?

The ArrowDtype is used for the result of the predicates operations on some dtypes in pandas 3 by default, but the values should be same as BooleanDtype.

It should be ignored in tests.

>>> pser = pd.Series(["x", "y", "z", None], dtype="string")
>>> other_pser = pd.Series([None, "z", "y", "x"], dtype="string")
>>>
>>> pser == other_pser
0     <NA>
1    False
2    False
3     <NA>
dtype: bool[pyarrow]
>>> (pser == other_pser).astype("boolean")
0     <NA>
1    False
2    False
3     <NA>
dtype: boolean

Does this PR introduce any user-facing change?

No.

How was this patch tested?

Updated the test util.

Was this patch authored or co-authored using generative AI tooling?

No.

[ueshin]

cc @gaogaotiantian @HyukjinKwon @zhengruifeng