есть такая таблица
+----+--------------------------------+-----+----+
|id |topping |type |name|
+----+--------------------------------+-----+----+
|0001|{5001, None} |donut|Cake|
|0001|{5002, Glazed} |donut|Cake|
|0001|{5005, Sugar} |donut|Cake|
|0001|{5007, Powdered Sugar} |donut|Cake|
|0001|{5006, Chocolate with Sprinkles}|donut|Cake|
|0001|{5003, Chocolate} |donut|Cake|
|0001|{5004, Maple} |donut|Cake|
+----+--------------------------------+-----+----+
как можно разбить столбец topping на два столбца key и value при помощи spark.sql или spark.dataframe ?