用户访问跟踪数据集记录每次访问的第一个和最后一个操作时间：
idvisit user_id visit_date visit_first_action_time visit_last_action_time
5732 abdifa01 2017-11-13 2017-11-13 07:23:41 2017-11-13 12:54:10
5140 abdifa01 2017-11-09 2017-11-09 07:14:22 2017-11-09 18:13:04
4906 abdifa01 2017-11-08 2017-11-08 07:58:05 2017-11-08 19:52:09
4663 abdifa01 2017-11-07 2017-11-07 10:11:52 2017-11-07 12:09:01
4741 abdifa01 2017-11-07 2017-11-07 12:54:17 2017-11-07 20:11:51
5760 abirio00 2017-11-13 2017-11-13 08:14:17 2017-11-13 12:59:31
5681 abirio00 2017-11-12 2017-11-12 10:31:00 2017-11-12 16:12:01
5620 abirio00 2017-11-11 2017-11-11 09:27:21 2017-11-11 12:45:07
5643 abirio00 2017-11-11 2017-11-11 14:06:32 2017-11-11 17:01:38
5553 abirio00 2017-11-10 2017-11-10 15:21:07 2017-11-10 16:07:02
5554 abirio00 2017-11-10 2017-11-10 15:21:07 2017-11-10 15:21:07
5436 abirio00 2017-11-10 2017-11-10 08:24:53 2017-11-10 14:41:17
5183 abirio00 2017-11-09 2017-11-09 08:08:23 2017-11-09 08:08:23
5204 abirio00 2017-11-09 2017-11-09 08:42:23 2017-11-09 09:19:50
5236 abirio00 2017-11-09 2017-11-09 09:51:40 2017-11-09 10:20:54
5252 abirio00 2017-11-09 2017-11-09 10:57:29 2017-11-09 15:48:59
如何使用SQL查询生成类似的数据集，该查询在同一user_id上的同一visit_date上显示上述记录的之间的空闲时间？
user_id visit_date start_idle_time finish_idle_time
abdifa01 2017-11-07 2017-11-07 12:09:01 2017-11-07 12:54:17
abirio00 2017-11-11 2017-11-11 12:45:07 2017-11-11 14:06:32
abirio00 2017-11-10 2017-11-11 14:41:17 2017-11-11 15:21:07
abirio00 2017-11-09 2017-11-09 08:08:23 2017-11-09 08:42:23
abirio00 2017-11-09 2017-11-09 09:19:50 2017-11-09 09:51:40
abirio00 2017-11-09 2017-11-09 10:20:54 2017-11-09 10:57:29
答案 0 :(得分：1)
对于SQL Server 2012及更高版本，请使用LEAD()。对于SQL 2008，您可以使用ROW_NUMBER()和self join使用生成的行号偏移1来对齐＆＃34;下一个值＆＃34;到当前行。两种技术如下所示进行比较。 common table expression（CTE）用于促进自我加入。
MS SQL Server 2014架构设置：
CREATE TABLE Table1
([idvisit] int, [user_id] varchar(8), [visit_date] date, [visit_first_action_time] datetime, [visit_last_action_time] datetime)
;
INSERT INTO Table1
([idvisit], [user_id], [visit_date], [visit_first_action_time], [visit_last_action_time])
VALUES
(5732, 'abdifa01', '2017-11-13 00:00:00', '2017-11-13 07:23:41', '2017-11-13 12:54:10'),
(5140, 'abdifa01', '2017-11-09 00:00:00', '2017-11-09 07:14:22', '2017-11-09 18:13:04'),
(4906, 'abdifa01', '2017-11-08 00:00:00', '2017-11-08 07:58:05', '2017-11-08 19:52:09'),
(4663, 'abdifa01', '2017-11-07 00:00:00', '2017-11-07 10:11:52', '2017-11-07 12:09:01'),
(4741, 'abdifa01', '2017-11-07 00:00:00', '2017-11-07 12:54:17', '2017-11-07 20:11:51'),
(5760, 'abirio00', '2017-11-13 00:00:00', '2017-11-13 08:14:17', '2017-11-13 12:59:31'),
(5681, 'abirio00', '2017-11-12 00:00:00', '2017-11-12 10:31:00', '2017-11-12 16:12:01'),
(5620, 'abirio00', '2017-11-11 00:00:00', '2017-11-11 09:27:21', '2017-11-11 12:45:07'),
(5643, 'abirio00', '2017-11-11 00:00:00', '2017-11-11 14:06:32', '2017-11-11 17:01:38'),
(5553, 'abirio00', '2017-11-10 00:00:00', '2017-11-10 15:21:07', '2017-11-10 16:07:02'),
(5554, 'abirio00', '2017-11-10 00:00:00', '2017-11-10 15:21:07', '2017-11-10 15:21:07'),
(5436, 'abirio00', '2017-11-10 00:00:00', '2017-11-10 08:24:53', '2017-11-10 14:41:17'),
(5183, 'abirio00', '2017-11-09 00:00:00', '2017-11-09 08:08:23', '2017-11-09 08:08:23'),
(5204, 'abirio00', '2017-11-09 00:00:00', '2017-11-09 08:42:23', '2017-11-09 09:19:50'),
(5236, 'abirio00', '2017-11-09 00:00:00', '2017-11-09 09:51:40', '2017-11-09 10:20:54'),
(5252, 'abirio00', '2017-11-09 00:00:00', '2017-11-09 10:57:29', '2017-11-09 15:48:59')
;
使用LEAD（）：
查询select
idvisit
, user_id
, visit_date
, visit_first_action_time
, visit_last_action_time
, lead (visit_first_action_time) over (partition by user_id, visit_date
order by visit_first_action_time ASC, visit_last_action_time ASC) as start_idle_time
, datediff(ss,visit_last_action_time
, lead (visit_first_action_time) over (partition by user_id, visit_date
order by visit_first_action_time ASC, visit_last_action_time ASC)
) as secs_between
from Table1
order by user_id, visit_date, visit_first_action_time, visit_last_action_time
;
<强> Results ：
| idvisit | user_id | visit_date | visit_first_action_time | visit_last_action_time | start_idle_time | secs_between |
|---------|----------|------------|-------------------------|------------------------|----------------------|--------------|
| 4663 | abdifa01 | 2017-11-07 | 2017-11-07T10:11:52Z | 2017-11-07T12:09:01Z | 2017-11-07T12:54:17Z | 2716 |
| 4741 | abdifa01 | 2017-11-07 | 2017-11-07T12:54:17Z | 2017-11-07T20:11:51Z | (null) | (null) |
| 4906 | abdifa01 | 2017-11-08 | 2017-11-08T07:58:05Z | 2017-11-08T19:52:09Z | (null) | (null) |
| 5140 | abdifa01 | 2017-11-09 | 2017-11-09T07:14:22Z | 2017-11-09T18:13:04Z | (null) | (null) |
| 5732 | abdifa01 | 2017-11-13 | 2017-11-13T07:23:41Z | 2017-11-13T12:54:10Z | (null) | (null) |
| 5183 | abirio00 | 2017-11-09 | 2017-11-09T08:08:23Z | 2017-11-09T08:08:23Z | 2017-11-09T08:42:23Z | 2040 |
| 5204 | abirio00 | 2017-11-09 | 2017-11-09T08:42:23Z | 2017-11-09T09:19:50Z | 2017-11-09T09:51:40Z | 1910 |
| 5236 | abirio00 | 2017-11-09 | 2017-11-09T09:51:40Z | 2017-11-09T10:20:54Z | 2017-11-09T10:57:29Z | 2195 |
| 5252 | abirio00 | 2017-11-09 | 2017-11-09T10:57:29Z | 2017-11-09T15:48:59Z | (null) | (null) |
| 5436 | abirio00 | 2017-11-10 | 2017-11-10T08:24:53Z | 2017-11-10T14:41:17Z | 2017-11-10T15:21:07Z | 2390 |
| 5554 | abirio00 | 2017-11-10 | 2017-11-10T15:21:07Z | 2017-11-10T15:21:07Z | 2017-11-10T15:21:07Z | 0 |
| 5553 | abirio00 | 2017-11-10 | 2017-11-10T15:21:07Z | 2017-11-10T16:07:02Z | (null) | (null) |
| 5620 | abirio00 | 2017-11-11 | 2017-11-11T09:27:21Z | 2017-11-11T12:45:07Z | 2017-11-11T14:06:32Z | 4885 |
| 5643 | abirio00 | 2017-11-11 | 2017-11-11T14:06:32Z | 2017-11-11T17:01:38Z | (null) | (null) |
| 5681 | abirio00 | 2017-11-12 | 2017-11-12T10:31:00Z | 2017-11-12T16:12:01Z | (null) | (null) |
| 5760 | abirio00 | 2017-11-13 | 2017-11-13T08:14:17Z | 2017-11-13T12:59:31Z | (null) | (null) |
使用ROW_NUMBER（）和CTE 查询：
with myCTE as (
select
table1.*
, row_number() over (partition by user_id, visit_date
order by visit_first_action_time ASC, visit_last_action_time ASC) as rn
from table1
)
select
t1.idvisit
, t1.user_id
, t1.visit_date
, t1.visit_first_action_time
, t1.visit_last_action_time
, t2.visit_first_action_time as start_idle_time
, datediff(ss,t1.visit_last_action_time, t2.visit_first_action_time) as secs_between
from myCTE t1
left join myCTE t2 on t1.user_id = t2.user_id and t1.visit_date = t2.visit_date
and t1.rn = t2.rn-1
order by t1.user_id, t1.visit_date, t1.visit_first_action_time, t1.visit_last_action_time
;
<强> Results ：
| idvisit | user_id | visit_date | visit_first_action_time | visit_last_action_time | start_idle_time | secs_between |
|---------|----------|------------|-------------------------|------------------------|----------------------|--------------|
| 4663 | abdifa01 | 2017-11-07 | 2017-11-07T10:11:52Z | 2017-11-07T12:09:01Z | 2017-11-07T12:54:17Z | 2716 |
| 4741 | abdifa01 | 2017-11-07 | 2017-11-07T12:54:17Z | 2017-11-07T20:11:51Z | (null) | (null) |
| 4906 | abdifa01 | 2017-11-08 | 2017-11-08T07:58:05Z | 2017-11-08T19:52:09Z | (null) | (null) |
| 5140 | abdifa01 | 2017-11-09 | 2017-11-09T07:14:22Z | 2017-11-09T18:13:04Z | (null) | (null) |
| 5732 | abdifa01 | 2017-11-13 | 2017-11-13T07:23:41Z | 2017-11-13T12:54:10Z | (null) | (null) |
| 5183 | abirio00 | 2017-11-09 | 2017-11-09T08:08:23Z | 2017-11-09T08:08:23Z | 2017-11-09T08:42:23Z | 2040 |
| 5204 | abirio00 | 2017-11-09 | 2017-11-09T08:42:23Z | 2017-11-09T09:19:50Z | 2017-11-09T09:51:40Z | 1910 |
| 5236 | abirio00 | 2017-11-09 | 2017-11-09T09:51:40Z | 2017-11-09T10:20:54Z | 2017-11-09T10:57:29Z | 2195 |
| 5252 | abirio00 | 2017-11-09 | 2017-11-09T10:57:29Z | 2017-11-09T15:48:59Z | (null) | (null) |
| 5436 | abirio00 | 2017-11-10 | 2017-11-10T08:24:53Z | 2017-11-10T14:41:17Z | 2017-11-10T15:21:07Z | 2390 |
| 5554 | abirio00 | 2017-11-10 | 2017-11-10T15:21:07Z | 2017-11-10T15:21:07Z | 2017-11-10T15:21:07Z | 0 |
| 5553 | abirio00 | 2017-11-10 | 2017-11-10T15:21:07Z | 2017-11-10T16:07:02Z | (null) | (null) |
| 5620 | abirio00 | 2017-11-11 | 2017-11-11T09:27:21Z | 2017-11-11T12:45:07Z | 2017-11-11T14:06:32Z | 4885 |
| 5643 | abirio00 | 2017-11-11 | 2017-11-11T14:06:32Z | 2017-11-11T17:01:38Z | (null) | (null) |
| 5681 | abirio00 | 2017-11-12 | 2017-11-12T10:31:00Z | 2017-11-12T16:12:01Z | (null) | (null) |
| 5760 | abirio00 | 2017-11-13 | 2017-11-13T08:14:17Z | 2017-11-13T12:59:31Z | (null) | (null) |