BUG/TST: PeriodArray.__setitem__ with slice and list-like value fails

[TomAugspurger]

In [1]: import pandas as pd; import numpy as np

In [2]: arr = pd.core.arrays.period_array([2000, 2001], 'D')

In [3]: arr[:1] = ['2001']
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-3-907919e36fd9> in <module>()
----> 1 arr[:1] = ['2001']

~/sandbox/pandas/pandas/core/arrays/period.py in __setitem__(self, key, value)
    358         # work, since the freq can't be inferred.
    359         if is_list_like(value):
--> 360             if len(key) != len(value) and not com.is_bool_indexer(key):
    361                 msg = ("shape mismatch: value array of length '{}' does not "
    362                        "match indexing result of length '{}'.")

TypeError: object of type 'slice' has no len()

---

We should be checking that the if key is a slice

            is_slice = isinstance(key, slice)
            if (not is_slice
                    and len(key) != len(value)
                    and not com.is_bool_indexer(key)):
                msg = ("shape mismatch: value array of length '{}' does not "
                       "match indexing result of length '{}'.")
                raise ValueError(msg.format(len(key), len(value)))
            if not is_slice and len(key) == 0:
                return

Then, if they slice length doesn't match with value, NumPy will complain at the end when we do self._data[key] = value.

I've fixed this on my all-in-one branch. To close this issue, we should also add a base extension tests to BaseSetitemTests like

def test_setitem_slice_mismatch_length_raises(self, data):
    arr = data[:5]
    with pytest.raises(ValueError):
        arr[:1] = arr[:2]

and a successful one like

def test_setitem_slice_array(self, data):
    arr = data[:5].copy()
    arr[:5] = data[-5:]
    self.assert_extension_array_equal(arr, data[-5:])

(those are untested)

[charlesdong1991]

Hello, has your all-in-one branch been merged into master? i fetched all, and looks like this change is not in master yet? maybe i did something wrong...

[TomAugspurger]

No, it's not quite ready for a PR yet.