[SPARK-52592][PS] Prevent error when creating ps.Series from ps.Series

[petern48]

What changes were proposed in this pull request?

I fixed the constructor to explicitly handle Series input similar to how it handles DataFrame input. It previously would fall to the else case that inputs it into a regular pandas Series constructor, leading to the error.

Why are the changes needed?

Previously, this code

obj = ps.Series([x for x in range(3)])
obj = ps.Series(obj)

would error:

...
pyspark.pandas.exceptions.PandasNotImplementedError: The method `pd.Series.__iter__()` is not implemented. If you want to collect your data as an NumPy array, use 'to_numpy()' instead.

Does this PR introduce any user-facing change?

Yes, previously, when you try to pass a ps.Series object into the ps.Series.__init__(), you can an error. Now, it will work successfully similar to how traditional pandas Series works.

How was this patch tested?

Yes, I added a test.

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

No

[petern48]

Once this is merged, I'd like to backport this as far back as 3.4.0 if possible, or at least back to 3.5.0

[allisonwang-db]

cc @HyukjinKwon