BUG: Series.resample fails on NaT index

[rhshadrach]

• closes BUG: Series.resample fails on NaT PeriodIndex #39227
• tests added / passed
• Ensure all linting tests pass, see here for how to run them
• whatsnew entry

This will also resolve the core issue of #34998 (comment). cc @jorisvandenbossche

The added test:

    s = series.copy()
    s.index = PeriodIndex([NaT] * len(s), freq=freq)
    result = getattr(s.resample(freq), resample_method)()

currently fails for freq="M", where upsample is used instead of downsample leading to failure since the index is not unique. I can't tell if this is a bug or an operation that should raise. It also fails for DataFrame, so I think is unrelated to the issue here.

cc @jbrockmendel

[jreback]

wait why do we have NaT in an index? e.g. why don't we have 0 bins?

[rhshadrach]

Below this is the logic for when there are NaT along with non-NaT (lines 1712-1718 here):

        if nat_count > 0:
            # NaT handling as in pandas._lib.lib.generate_bins_dt64()
            # shift bins by the number of NaT
            bins += nat_count
            bins = np.insert(bins, 0, nat_count)
            binner = binner.insert(0, NaT)
            labels = labels.insert(0, NaT)

The added logic agrees with this in the case there are all NaT.

[jreback]

i c. can you unify this logic (maybe to a function)?

[jreback]

thanks @rhshadrach