BUG: torch: sort arguments of result_type

This way, python scalars are always treated last. Otherwise,
"runs" of python scalars are problematic:
`result_type(float32, int64, 1, 2) -> ValueError`

Author: ev-br

Diff for: array_api_compat/torch/_aliases.py

@@ -131,10 +131,24 @@ def result_type(*arrays_and_dtypes: Union[array, Dtype, bool, int, float, comple
         if isinstance(x, torch.dtype):
             return x
         return x.dtype
+
     if len(arrays_and_dtypes) > 2:
+        # sort the scalars to the left so that they are treated last
+        scalars, others = [], []
+        for x in arrays_and_dtypes:
+            if isinstance(x, _py_scalars):
+                scalars.append(x)
+            else:
+                others.append(x)
+        if len(scalars) == len(arrays_and_dtypes):
+            raise ValueError("At least one array or dtype is required.")
+
+        arrays_and_dtypes = scalars + others
         return result_type(arrays_and_dtypes[0], result_type(*arrays_and_dtypes[1:]))
 
+    # the binary case
     x, y = arrays_and_dtypes
+
     if isinstance(x, _py_scalars):
         if isinstance(y, _py_scalars):
             raise ValueError("At least one array or dtype is required.")