Json jq 对特定键进行过滤并将值整理到单个 csv 单元格中

给定这个json:

{"data": [
  {
    "account_id": "123",
    "account_status": "active",
    "name": "john doe",
    "email": "johndoe@anywhere.com",
    "product_access": []
  },
  {
    "account_id": "345",
    "account_status": "active",
    "name": "jane doe",
    "email": "janedoe@anywhere.com",
    "last_active": "2023-08-23T13:03:27.811473590Z",
    "product_access": [
      {
        "name": "Product 1",
        "key": "product1",
        "url": "acme1.com"
      },
      {
        "name": "Product 1",
        "key": "product1",
        "url": "acme2.com"
      },
      {
        "name": "Product 1",
        "key": "product1",
        "url": "acme3.com"
      },
      {
        "name": "Product 2",
        "key": "product2",
        "url": "acme4.com",
        "last_active": "2023-08-23T13:03:27.811473590Z"
      },
      {
        "name": "Product 3",
        "key": "product3",
        "url": "acme5.com"
      },
      {
        "name": "Product 1",
        "key": "product1",
        "url": "acme4.com",
        "last_active": "2023-08-17T18:21:52.472085713Z"
      }
    ]
  }
]
}

我正在try 编写一个JQ表达式，以使CSV文件中的输出如下所示(为简单起见，此处不添加转义字符):

"account_id", "account_status", "name", "email", "Product 1", "Product 2", "Product 3"
"123", "active", "john doe", "johndoe@anywhere.com", ",,"
"345", "active", "jane doe", "janedoe@anywhere.com", "acme1.com, acme2.com, acme3.com, acme4.com", "acme4.com", "acme5.com"

到目前为止我已经试过了...

jq -r '["account_id", "account_status", "name", "email", "product 1", "product 2", "product 3"],
        (.data[] | [.account_id, .account_status, .name, .email,
                    (.product_access[] | select(.key == "product1").url // ""),
                    (.product_access[] | select(.key == "product2").url // ""),
                    (.product_access[] | select(.key == "product3").url // "")] | join(",")) | @csv'

它不会将产品n URL整理到单个单元格中.