使用 Bigquery 删除具有“重复”模式的嵌套列

Posted 2023-03-22

【中文标题】使用 Bigquery 删除具有“重复”模式的嵌套列

【英文标题】：Using Bigquery remove the nested columns which has "REPEATED" mode

【发布时间】：2021-12-30 05:00:06

【问题描述】：

使用 Bigquery，我试图删除如下所示的嵌套架构，但无法这样做。谁能告诉我。如何实现？

表格：

FiledName          Type       Mode
Person            RECORD   REPEATED
Person.Name       STRING   NULLABLE
Person.Add        RECORD   NULLABLE
Person.Add.line   STRING   NULLABLE

代码：

create table `project_id.dataset.new_table_name` as 
select * replace(
   (select as ARRAY(struct person.* except(add))) as person
)
from `project_id.dataset.table_name`;

预期输出：

FiledName          Type       Mode
Person            RECORD   REPEATED
Person.Name       STRING   NULLABLE

【参考方案1】：

考虑以下方法

create table `project_id.dataset.new_table_name` as 
select * replace(
    array(select as struct person.* except(add) from t.person) as person
  )
from `project_id.dataset.table_name` t;

【讨论】：

这是字符串函数的替换函数吗？这 -> cloud.google.com/bigquery/docs/reference/standard-sql/…

不。查看Modifiers for * operator

【参考方案2】：

试试这个

create table `project_id.dataset.new_table_name` as 
SELECT 
      [STRUCT (P.NAME AS NAME)] AS PERSON

from `project_id.dataset.table_name`,UNNEST(PERSON) AS P;