Categorical.__setitem__ should allow setting a Categorical with same unordered categories

[TomAugspurger]

We'd need to recode the source categorical's codes to the target's.

In [6]: a = pd.Categorical(['a', 'b'], categories=['b', 'a'])

In [7]: b = pd.Categorical(['a', 'b'], categories=['a', 'b'])

In [8]: a[[True, False]] = b[[True, False]]

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-8-9693ad7c05cc> in <module>
----> 1 a[[True, False]] = b[[True, False]]

~/sandbox/pandas/pandas/core/arrays/categorical.py in __setitem__(self, key, value)
   2082         if isinstance(value, Categorical):
   2083             if not value.categories.equals(self.categories):
-> 2084                 raise ValueError("Cannot set a Categorical with another, "
   2085                                  "without identical categories")
   2086

ValueError: Cannot set a Categorical with another, without identical categories

[TomAugspurger]

I may handle this in
#24114, since it's blocking a setitem-based where.

[jorisvandenbossche]

And does the where operation need to work with unequal categorical dtypes?