如何展平具有大量子值的 Json 数据

Posted 2023-02-23

【中文标题】如何展平具有大量子值的 Json 数据

【英文标题】：How to flatten Json data which having high number of child values

【发布时间】：2021-02-26 09:02:20

【问题描述】：

我需要将 json 格式的数据扁平化为结构化的表格格式。这种格式的json数据。它有100多个国家。我尝试了几种方法，但找不到解决方案。

   "Country":
      "0":"USA",
      "1":"Japan",
      "2":"Canada",
      "3":"Mexico"
   ,
   "Capital":
      "0":"Washington",
      "1":"Tokyo",
      "2":"Ottawa",
      "3":"Mexico City"
   ,
   "Currency":
      "0":"USD",
      "1":"Yen",
      "2":"Canadian Dollar",
      "3":"Peso"
   

到，

 Id| country | Capital    | Currency|
 -----------------------------------
  0| USA     | Washington | USD     |
  1| Japan   | Tokyo      | Yen     |

等等

【参考方案1】：

要在同时迭代 3 个数组时展平，请确保在取消嵌套时匹配每个 lateral flatten() 键：

with data as (select parse_json('
   "Country":
      "0":"USA",
      "1":"Japan",
      "2":"Canada",
      "3":"Mexico"
   ,
   "Capital":
      "0":"Washington",
      "1":"Tokyo",
      "2":"Ottawa",
      "3":"Mexico City"
   ,
   "Currency":
      "0":"USD",
      "1":"Yen",
      "2":"Canadian Dollar",
      "3":"Peso"
   
') v)

select country.value::varchar country
  , capital.value::varchar capital
  , currency.value::varchar currency
from data
  , lateral flatten(input => v:Country) country
  , lateral flatten(input => v:Capital) capital
  , lateral flatten(input => v:Currency) currency
where country.key=capital.key
and country.key=currency.key