嵌套的json函数

Question

这些是我的桌子：

Products
+-------------+---------+------+-----+
| Field       | Type    | Null | Key |
+-------------+---------+------+-----+
| id          | int(11) | NO   | PRI |
| retailer_id | int(11) | NO   | MUL |
+-------------+---------+------+-----+

Clothings
+------------+-------------+------+-----+
| Field      | Type        | Null | Key |
+------------+-------------+------+-----+
| id         | int(11)     | NO   | PRI |
| product_id | int(11)     | NO   | MUL |
| material   | varchar(30) | YES  |     |
+------------+-------------+------+-----+

ClothingColors
+-------------+------------+------+-----+
| Field       | Type       | Null | Key |
+-------------+------------+------+-----+
| id          | int(11)    | NO   | PRI |
| clothing_id | int(11)    | NO   | MUL |
| color       | varchar(6) | NO   |     |
+-------------+------------+------+-----+

ClothingSizes
+------------------+-------------+------+-----+
| Field            | Type        | Null | Key |
+------------------+-------------+------+-----+
| id               | int(11)     | NO   | PRI |
| clothingcolor_id | int(11)     | NO   | MUL |
| symbol           | varchar(20) | NO   |     |
| availability     | tinyint(1)  | NO   |     |
| size_details     | json        | NO   |     |
+------------------+-------------+------+-----+

我需要一个查询结果：

[
        {
            "productId": 1,
            "material": "explanation",
            "variety": [
                {
                    "colorId": 12,
                    "color": "ffffff",
                    "sizes": [
                        { "sizeId": 33, "symbol": "L", "availability": true, "sizeDetails": {} },
                        { "sizeId": 34, "symbol": "XL", "availability": true, "sizeDetails": {} }
                    ],
                }
            ]
        },
        {
            "productId": 1,
            "material": "explanation",
            ...
        },
        ...
]

到目前为止，我可以提出以下查询：

SELECT Products.id AS productId,
            GROUP_CONCAT(DISTINCT Clothings.material) AS material,
             JSON_ARRAY(
                    GROUP_CONCAT(
                         JSON_OBJECT(
                                        "colorId", ClothingColors.id,
                                        "color", ClothingColors.color
                                        "sizes", -- WHAT SHOULD BE HERE?
                                    )
                    )
             ) AS variety

FROM Products

    INNER JOIN Clothings ON Clothings.product_id = Products.id
    INNER JOIN ClothingColors ON ClothingColors.clothing_id = Clothings.id
    INNER JOIN ClothingSizes ON ClothingSizes.clothingcolor_id = ClothingColors.id

    GROUP BY Products.id

问题：我该如何写一个查询，该查询会导致我的期望输出，并且由于我没有生产经验，您是否认为这将是一个缓慢的查询，而且最好在应用层？