我有以下数据集:
{"date":{"year":2017,"month":7,"day":2},"data":{"param1":[10,20,30,440],"param2":[55,65,75,85],"param3":[43,55,71,72]}}
我从名为rawData的表中检索数据。
现在我想要删除数组中的值,并将它们放在另一个表(seperateData)中的4个单独的行和列中,因此对于此示例,它看起来像:
year | month | day | param1 | param2 | param3
2017 7 2 10 55 43
2017 7 2 20 65 55
2017 7 2 30 75 71
2017 7 2 440 85 72
我猜我必须使用不正确和正常来实现这一点,但我仍然坚持我应该如何继续。数组param1,param2,param3的长度始终相同。
我已经对这些数据做了一个小提琴: https://www.db-fiddle.com/f/jZhYLEvdSERzLCDyoAJz65/0
答案 0 :(得分:1)
正确的做法不是最好的方法我的PostgreSQL体验并不是那么好但它有效。
<强>查询强>
SELECT
((rawData.values)::json->'date')::json->'year' AS year
, ((rawData.values)::json->'date')::json->'month' AS month
, ((rawData.values)::json->'date')::json->'day' AS day
, JSON_ARRAY_ELEMENTS(((rawData.values)::json->'data')::json->'param1') AS param1
, JSON_ARRAY_ELEMENTS(((rawData.values)::json->'data')::json->'param2') AS param2
, JSON_ARRAY_ELEMENTS(((rawData.values)::json->'data')::json->'param3') AS param3
FROM
rawData
demo https://www.db-fiddle.com/f/jZhYLEvdSERzLCDyoAJz65/1
当数组param1,param2,param3的长度不相等时,查询也会起作用。
demo https://www.db-fiddle.com/f/jZhYLEvdSERzLCDyoAJz65/4
感谢您的回答,它迄今为止有效。我唯一忘记的是,我 实际上也需要索引作为单独的列。所以索引了 数组元素。任何想法
这可能与GENERATE_SERIES结合使用JSON_ARRAY_LENGTH
<强>查询强>
SELECT
((rawData.values)::json->'date')::json->'year' AS year
, ((rawData.values)::json->'date')::json->'month' AS month
, ((rawData.values)::json->'date')::json->'day' AS day
, JSON_ARRAY_ELEMENTS(((rawData.values)::json->'data')::json->'param1') AS param1
, GENERATE_SERIES(1, JSON_ARRAY_LENGTH(((rawData.values)::json->'data')::json->'param1')) AS param1_array_index
, JSON_ARRAY_ELEMENTS(((rawData.values)::json->'data')::json->'param2') AS param2
, GENERATE_SERIES(1, JSON_ARRAY_LENGTH(((rawData.values)::json->'data')::json->'param2')) AS param2_array_index
, JSON_ARRAY_ELEMENTS(((rawData.values)::json->'data')::json->'param3') AS param3
, GENERATE_SERIES(1, JSON_ARRAY_LENGTH(((rawData.values)::json->'data')::json->'param3')) AS param3_array_index
FROM
rawData