hivesql解析json数组并拆分成多行

原始数据：

	[{"name":"woods","app_id":"abc123"},{"name":"tiger","app_id":"def456"}]

数据存在表dev.woods_test中

需求与方法：

1. 解析json，一行拆分成两行

select a_json
from
(
select split(regexp_replace(regexp_extract (json_col ,'(\\[)(.*?)(\\])',2),'\\},\\{','\\}|\\{'),'\\|') as a_list
from dev.woods_test
) a
lateral view explode(a_list) a_list_tab as a_json

结果：

{"name":"woods","app_id":"abc123"}
{"name":"tiger","app_id":"def456"}

2. 解析上述结果json,取出name和app_id值

select 
get_json_object(a_json,'$.name') as name,
get_json_object(a_json,'$.app_id') as app_id
from
(
select split(regexp_replace(regexp_extract (json_col ,'(\\[)(.*?)(\\])',2),'\\},\\{','\\}|\\{'),'\\|') as a_list
from dev.woods_test
) a
lateral view explode(a_list) a_list_tab as a_json

结果：

woods,abc123
tiger,def456

---

欢迎访问个人博客:Woods Blog