Google BigQuery SQL：将多级 JSON (list +json+list+json) 解析为列

Posted 2023-03-25

【中文标题】Google BigQuery SQL：将多级 JSON (list +json+list+json) 解析为列

【英文标题】：Google BigQuery SQL: Parse multi level JSON (list +json+list+json) to Columns

【发布时间】：2021-03-15 11:42:22

【问题描述】：

我有两个级别的 Json 字符串：list[jsonlist[json]]:

UserID               Json_List
100     ["application_charge":"id":13409353813,"name":"Starter","api_client_id":2485321,"usage_charges":["id":48216805,"description":"Extras","price":"60.70"]]
200     ["application_charge":"id":13409353814,"name":"Starter","api_client_id":2485322,"usage_charges":["id":48216890,"description":"Extras","price":"80.79"]]

需要像表格一样的输出：

UserID  application_charge.id   name    api_client_id   usage_charges.id    description   price
100      13409353813          Starter    2485321        48216805           Extras         60.7
200      13409353814          Starter    2485322        48216890           Extras         80.79

我设法退出了第一步“application_charge”，但不明白如何进入下一步“usage_charges”

select  
  UserID,
  json_extract_scalar(json, '$.application_charge.id')  as id,
  json_extract_scalar(json, '$.application_charge.name')  as name,
  json_extract_scalar(json, '$.application_charge.api_client_id')  as api_client_id
from `be-prod-data.retailers_billing_data_production.shopify_application_charges`,
unnest(json_extract_array( json_list, '$')) json

如何从第二阶段提取数据？？

【参考方案1】：

再多一个unnest(json_extract_array(...

with mytable as (
  select 100 as userid, '["application_charge":"id":13409353813,"name":"Starter","api_client_id":2485321,"usage_charges":["id":48216805,"description":"Extras","price":"60.70"]]' as json_list union all
  select 200, '["application_charge":"id":13409353814,"name":"Starter","api_client_id":2485322,"usage_charges":["id":48216890,"description":"Extras","price":"80.79"]]'
)
select  
  UserID,
  json_extract_scalar(json, '$.application_charge.id')  as id,
  json_extract_scalar(json, '$.application_charge.name')  as name,
  json_extract_scalar(json, '$.application_charge.api_client_id')  as api_client_id,
  json_extract_scalar(nested, '$.id')  as usage_charges_id,
from mytable,
unnest(json_extract_array( json_list, '$')) json, unnest(json_extract_array(json, '$.usage_charges')) as nested