求SQL server里不同记录日期时间差语句

[iefxftjeg]

求图中10多万条记录里单据号相同情况下第一条编码的日期与最后一条编码的日期时间差（分别到天、小时）

[唐诗三百首]

请提供表结构, 测试数据(非图片,几行即可), 处理逻辑和希望结果.

[iefxftjeg]

表数据如下 181165/03 18-04-00129-17 2018-12-02 13:54:28.263 181165/03 18-04-00129-21 2018-12-02 15:46:03.367 181165/03 18-04-00129-22 2018-12-02 15:35:52.127 181165/03 18-04-00129-23 2018-12-02 15:28:43.933 181165/03 18-04-00129-24 2018-12-02 13:04:01.687 181165/03 18-04-00129-25 2018-12-02 13:47:50.433 181165/03 18-04-00129-26 2018-12-02 13:55:29.420 181165/03 18-04-00129-27 2018-12-02 12:58:25.160 181165/03 18-04-00129-28 2018-12-02 12:53:49.583 181165/03 18-04-00129-29 2018-12-02 12:51:47.277 181165/02 00-04-00110-29 2018-11-30 01:27:02.953 181165/02 18-04-00110-02 2018-11-30 05:48:02.693 181165/02 18-04-00110-04 2018-11-30 05:45:29.860 181165/02 18-04-00110-06 2018-11-30 05:40:24.373 181165/02 18-04-00110-08 2018-11-30 04:19:58.320 181165/02 18-04-00110-21 2018-11-30 04:16:24.277 181165/02 18-04-00110-22 2018-11-30 04:07:45.080 181165/02 18-04-00110-23 2018-11-30 04:00:37.487 181165/02 18-04-00110-24 2018-11-30 03:47:53.587 181165/02 18-04-00110-25 2018-11-30 03:27:31.053 181165/02 18-04-00110-26 2018-11-30 03:17:50.397 181165/02 18-04-00110-27 2018-11-30 04:17:25.520 181165/02 18-04-00110-28 2018-11-30 02:49:48.680 181165/02 18-04-00110-29 2018-11-30 03:02:33.753 想实现结果 181165/03 消耗时间1小时 181165/02 消耗时间6小时

[吉普赛的歌]

USE tempdb
GO
IF OBJECT_ID('dbo.[t]') IS NOT NULL 
	DROP TABLE dbo.[t]
GO
CREATE TABLE dbo.[t](
[单据号] VARCHAR(30)
,[编码] VARCHAR(30)
,[日期] DATETIME	
)
GO
SET NOCOUNT ON
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-17',N'2018-12-02 13:54:28.263')
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-21',N'2018-12-02 15:46:03.367')
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-22',N'2018-12-02 15:35:52.127')
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-23',N'2018-12-02 15:28:43.933')
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-24',N'2018-12-02 13:04:01.687')
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-25',N'2018-12-02 13:47:50.433')
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-26',N'2018-12-02 13:55:29.420')
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-27',N'2018-12-02 12:58:25.160')
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-28',N'2018-12-02 12:53:49.583')
INSERT INTO dbo.[t] VALUES(N'181165/03',N'18-04-00129-29',N'2018-12-02 12:51:47.277')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'00-04-00110-29',N'2018-11-30 01:27:02.953')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-02',N'2018-11-30 05:48:02.693')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-04',N'2018-11-30 05:45:29.860')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-06',N'2018-11-30 05:40:24.373')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-08',N'2018-11-30 04:19:58.320')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-21',N'2018-11-30 04:16:24.277')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-22',N'2018-11-30 04:07:45.080')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-23',N'2018-11-30 04:00:37.487')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-24',N'2018-11-30 03:47:53.587')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-25',N'2018-11-30 03:27:31.053')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-26',N'2018-11-30 03:17:50.397')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-27',N'2018-11-30 04:17:25.520')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-28',N'2018-11-30 02:49:48.680')
INSERT INTO dbo.[t] VALUES(N'181165/02',N'18-04-00110-29',N'2018-11-30 03:02:33.753')

SELECT 
t.单据号
,datediff(day,Min(t.日期),max(t.日期)) AS [消耗时间(天)]
,datediff(hour,Min(t.日期),max(t.日期)) AS [消耗时间(小时)]  
FROM t
GROUP BY t.单据号

[iefxftjeg]

版主这方法是要所有单据号的情况下，我想对系统里所有单据号做同样操作的话有更简单的语句吗？

[iefxftjeg]

是我理解错了，不好意思