Postgres编号取决于下一个ID

Question

我有一个场景：

这是我的查询：

select id, empid, dt,ROW_NUMBER() OVER(PARTITION BY id ORDER BY dt)  from
(select 21919 id,219 empid,'2018-06-01 08:41:59' dt
union all
select 21919,219,'2018-06-01 08:43:29'
union all
select 21919,219,'2018-06-01 08:50:02'
union all
select 21919,88,'2018-06-01 08:56:45'
union all
select 21919,20,'2018-06-01 11:25:55'
union all
select 21919,20,'2018-06-01 12:04:36'
union all
select 21919,88,'2018-06-01 12:05:17'
union all
select 21919,132,'2018-06-01 12:18:27'
union all
select 21919,132,'2018-06-04 13:34:53') as a

和我的结果：

但是我需要这样的结果：

在Postgres中可以做到吗？

任何建议都值得赞赏。

Answer 1

如果我正确地阅读了您的任务，它会像添加用于分组的属性一样简单？..

select id, empid, dt,ROW_NUMBER() OVER(PARTITION BY id ORDER BY dt,empid)  from
(select 21919 id,219 empid,'2018-06-01 08:41:59' dt
union all
select 21919,219,'2018-06-01 08:43:29'
union all
select 21919,219,'2018-06-01 08:50:02'
union all
select 21919,88,'2018-06-01 08:56:45'
union all
select 21919,20,'2018-06-01 11:25:55'
union all
select 21919,20,'2018-06-01 12:04:36'
union all
select 21919,88,'2018-06-01 12:05:17'
union all
select 21919,132,'2018-06-01 12:18:27'
union all
select 21919,132,'2018-06-04 13:34:53') as a