数据库按照符号分割多行

原始数据

例如：下面的sql语句

select
address_new as address,
'new_address' as source
from
th_rw.dwd_th3_new_source_address_all dtaa
lateral view explode(split(address,',')) t as address_new
where
`source`  = 'dwd_tw_huji_aft_8'
and
address like '%,%'

经过sql语句转换之后

按照数据量对数据进行分区展示

select 
address,
case when rank<=1200000 then 'ods_dzs_person_1'
     when rank<=2400000 then 'ods_dzs_person_2'
     when rank<=3600000 then 'ods_dzs_person_3'
     when rank<=4800000 then 'ods_dzs_person_4'
     when rank<=6000000 then 'ods_dzs_person_5'
     when rank<=7200000 then 'ods_dzs_person_6'
     when rank<=8400000 then 'ods_dzs_person_7'
     when rank<=9600000 then 'ods_dzs_person_8'
     else 'ods_dzs_person_9' end  as source
from 
(
select address, row_number() over (partition by source order by address) as rank
from dwd_th3_new_source_address_all88
where source = 'ods_dzs_person'
)