Spark SQL JSON数据集查询嵌套数据结构

Question

我有一个简单的JSON数据集,如下所示.如何查询所有parts.lock的id= 1.

JSON:

{
    "id": 1,
    "name": "A green door",
    "price": 12.50,
    "tags": ["home", "green"],
    "parts" : [
        {
            "lock" : "One lock",
            "key" : "single key"
        },
        {
            "lock" : "2 lock",
            "key" : "2 key"
        }
    ]
}

查询:

select id,name,price,parts.lockfrom product where id=1

关键是如果我使用parts[0].lock它将返回如下一行:

{u'price': 12.5, u'id': 1, u'.lock': {u'lock': u'One lock', u'key': u'single key'}, u'name': u'A green door'}

但我想返回所有locks的parts结构.它将返回多行,但这是我正在寻找的那一行.这种我想要完成的关系连接.

请在这件事上给予我帮助

Answer 1

df.select($"id", $"name", $"price", explode($"parts").alias("elem"))
  .where("id = 1")
  .select("id", "name", "price", "elem.lock", "elem.key").show

+---+------------+-----+--------+----------+
| id|        name|price|    lock|       key|
+---+------------+-----+--------+----------+
|  1|A green door| 12.5|One lock|single key|
|  1|A green door| 12.5|  2 lock|     2 key|
+---+------------+-----+--------+----------+