使用dsbulk将json数据加载到Cassandra

我觉得dsbulk文档中缺少将json文件加载到cassandra中的文档

以下是我尝试加载的json文件的一部分：

[
  {
    "tags": [
      "r"
    ],
    "owner": {
      "reputation": 23,
      "user_id": 12235281,
      "user_type": "registered",
      "profile_image": "https://www.gravatar.com/avatar/60e28f52215bff12adb9758fc2cf86dd?s=128&d=identicon&r=PG&f=1",
      "display_name": "Me28",
      "link": "https://stackoverflow.com/users/12235281/me28"
    },
    "is_answered": false,
    "view_count": 3,
    "answer_count": 0,
    "score": 0,
    "last_activity_date": 1589053659,
    "creation_date": 1589053659,
    "question_id": 61702762,
    "link": "https://stackoverflow.com/questions/61702762/merge-dataframes-in-r-with-different-size-and-condition",
    "title": "Merge dataframes in R with different size and condition"
  },
  {
    "tags": [
      "python",
      "location",
      "pyautogui"
    ],
    "owner": {
      "reputation": 1,
      "user_id": 13507535,
      "user_type": "registered",
      "profile_image": "https://lh3.googleusercontent.com/a-/AOh14GgtdM9KrbH3X5Z33RCtz6xm_TJUSQS_S31deNYUcA=k-s128",
      "display_name": "lowhatex",
      "link": "https://stackoverflow.com/users/13507535/lowhatex"
    },
    "is_answered": false,
    "view_count": 2,
    "answer_count": 0,
    "score": 0,
    "last_activity_date": 1589053657,
    "creation_date": 1589053657,
    "question_id": 61702761,
    "link": "https://stackoverflow.com/questions/61702761/want-to-get-a-grip-of-this-pyautogui-command",
    "title": "Want to get a grip of this pyautogui command"
  }
]

我尝试加载此文件的方式如下：

dsbulk load-url./data\u so1.json-k stackoverflow\u t-t staging\u t-h'182.14.0.1'-头false-u username-p password

这是我得到的最接近的值，它将值逐行推入Cassandra，如下所示：

 data
-------------------------------------------------------------------------------------------------------------------------------
      "title": "'Microsoft.ACE.OLEDB.12.0' provider is not registered on the local machine giving exception on client"
               "profile_image": "https://www.gravatar.com/avatar/05085ede54486bdaebefcf8363e081e2?s=128&d=identicon&r=PG&f=1",
                                                                                                            "view_count": 422,
                                                                                                      "question_id": 61702768,
                                                                                                          "user_id": 12235281,

这只是按原样接收行（包括逗号）。我尝试过使用-m键进行映射，但没有真正实现

---

如何正确地将这些值添加到各自的列中？

命令行中有几个不一致之处：1）需要使用-c JSON指定JSON连接器；2） -标题false为CSV，您可以将其删除；3） 根据文件的结构，您需要添加--connector.json.mode SINGLE_DOCUMENT@adutra我查看了Brian的博客文章系列，没有找到任何JSON示例。“也许我们需要多发一条帖子？”阿杜特拉谢谢，这很有效。现在我已经介绍了您，您知道在dsbulk映射部分的“owner”部分中获取这些嵌套值的好技巧吗？您不能将嵌套的json结构映射到单个列。您需要将整个“owner”部分映射到具有相同结构的UDT（用户定义类型）。