Fix map_filter and map_zip_with output order

File "/Users/jiaan.geng/git-local/github-forks/spark/python/pyspark/sql/connect/functions.py", line 1423, in pyspark.sql.connect.functions.map_filter
Failed example:
    df.select(map_filter(
        "data", lambda _, v: v > 30.0).alias("data_filtered")
    ).show(truncate=False)
Expected:
    +--------------------------+
    |data_filtered             |
    +--------------------------+
    |{baz -> 32.0, foo -> 42.0}|
    +--------------------------+
Got:
    +--------------------------+
    |data_filtered             |
    +--------------------------+
    |{foo -> 42.0, baz -> 32.0}|
    +--------------------------+
    <BLANKLINE>
**********************************************************************
File "/Users/jiaan.geng/git-local/github-forks/spark/python/pyspark/sql/connect/functions.py", line 1465, in pyspark.sql.connect.functions.map_zip_with
Failed example:
    df.select(map_zip_with(
        "base", "ratio", lambda k, v1, v2: round(v1 * v2, 2)).alias("updated_data")
    ).show(truncate=False)
Expected:
    +---------------------------+
    |updated_data               |
    +---------------------------+
    |{SALES -> 16.8, IT -> 48.0}|
    +---------------------------+
Got:
    +---------------------------+
    |updated_data               |
    +---------------------------+
    |{IT -> 48.0, SALES -> 16.8}|
    +---------------------------+
    <BLANKLINE>