You have a DataFrame df with a nested column address.city. How would you extract this nested field into a top-level column named city using Spark? Provide the code snippet.