Concat_rows bug

[Pballen]

dataset.concat_rows breaks when a column appears in one dataset but not the other and destroy is set to True (it's fine if destroy is set to False). From the docs:

ds1 = rt.Dataset({'col_'+str(i):np.random.rand(5) for i in range(3)})
ds2 = rt.Dataset({'col_'+str(i):np.random.rand(5) for i in range(2)})
rt.Dataset.concat_rows([ds1, ds2], destroy=True)