使用Amazon Athena创建表并查询json数据吗？

Question

我想使用Amazon Athena查询格式的JSON数据：

[{"id":"0581b7c92be",
  "key":"0581b7c92be",
  "value":{"rev":"1-ceeeecaa040"},
  "doc":{"_id":"0581b7c92be497d19e5ab51e577ada12","_rev":"1ceeeecaa04","node":"belt","DeviceId":"C001"}},
 {"id":"0581b7c92be49",
  "key":"0581b7c92be497d19e5",
  "value":{"rev":"1-ceeeecaa04031842d3ca"},
  "doc":{"_id":"0581b7c92be497","_rev":"1ceeeecaa040318","node":"belt","DeviceId":"C001"}
 }
]

Answer 1

Athena DDL基于Hive，因此您希望数组中的每个json对象都放在单独的行中：

{"id": "0581b7c92be", "key": "0581b7c92be", "value": {"rev": "1-ceeeecaa040"}, "doc": {"_id": "0581b7c92be497d19e5ab51e577ada12", "_rev": "1ceeeecaa04", "node": "belt", "DeviceId": "C001"} }
{"id": "0581b7c92be49", "key": "0581b7c92be497d19e5", "value": {"rev": "1-ceeeecaa04031842d3ca"}, "doc": {"_id": "0581b7c92be497", "_rev": "1ceeeecaa040318", "node": "belt", "DeviceId": "C001"} }

嵌套字段（“值”，“ doc”）可能会出现问题，因此，如果可以将json展平，则操作起来会更容易。 （例如：参见Hive了解复杂的嵌套Json ）