Series construction with dtype=JSONDtype() and index with collection as 'scalar'

[simonjayhawkins]

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

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

• (optional) I have confirmed this bug exists on the master branch of pandas.

---

Code Sample, a copy-pastable example

>>> import pandas as pd
>>>
>>> pd.__version__
'1.1.0.dev0+1436.g6016b9841'
>>>
>>> from pandas.tests.extension.json.array import JSONDtype
>>>
>>> pd.Series({"g": 63}, index=[1, 2, 3], dtype=JSONDtype())
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "C:\Users\simon\pandas\pandas\core\series.py", line 283, in __init__
    data, index = self._init_dict(data, index, dtype)
  File "C:\Users\simon\pandas\pandas\core\series.py", line 372, in _init_dict
    s = create_series_with_explicit_dtype(
  File "C:\Users\simon\pandas\pandas\core\construction.py", line 624, in create_series_with_explicit_dtype
    return Series(
  File "C:\Users\simon\pandas\pandas\core\series.py", line 329, in __init__
    data = sanitize_array(data, index, dtype, copy, raise_cast_failure=True)
  File "C:\Users\simon\pandas\pandas\core\construction.py", line 441, in sanitize_array
    subarr = _try_cast(data, dtype, copy, raise_cast_failure)
  File "C:\Users\simon\pandas\pandas\core\construction.py", line 537, in _try_cast
    subarr = array_type(arr, dtype=dtype, copy=copy)
  File "C:\Users\simon\pandas\pandas\tests\extension\json\array.py", line 64, in _from_sequence
    return cls(scalars)
  File "C:\Users\simon\pandas\pandas\tests\extension\json\array.py", line 52, in __init__
    raise TypeError("All values must be of type " + str(self.dtype.type))
TypeError: All values must be of type <class 'collections.abc.Mapping'>

Problem description

Extension arrays may be able to hold a collection as a scalar value. should this be allowed?

Expected Output

same as

>>> pd.Series([{"g": 63}, {"g": 63}, {"g": 63}], index=[1, 2, 3], dtype=JSONDtype())
1    {'g': 63}
2    {'g': 63}
3    {'g': 63}
dtype: json

Output of pd.show_versions()

[paste the output of pd.show_versions() here leaving a blank line after the details tag]

[TomAugspurger]

I don't think we can support this in general. In this case the "scalar" type already has a meaning in the Series constructor: the keys are the index values.

It would be up to the caller to call this as

   ...: >>> pd.Series({1: {"g": 63}}, index=[1, 2, 3], dtype=JSONDtype())
Out[4]:
1    {'g': 63}
2           {}
3           {}
dtype: json