在SQL Server 2012中查找日期间隔？

如果员工有开始日期和结束日期以及历史记录。我们怎么能说有差距呢

样本数据：

+------+-------------+-------------+
|EmpID | StartDate   |  EndDate    |
+------+-------------+-------------+
| 555  |  7/8/2015   |  15/12/2015 |
| 555  |  16/12/2015 |  25/06/2016 |
| 555  |  28/06/2016 |  20/12/2016 | --Here 2 days gaps.  
| 555  |  21/12/2016 |  31/12/9999 |
+------+-------------+-------------+

---

假设没有重叠的日期，您可以使用

lag

获取上一个结束日期，并获取与当前行开始日期的差异，并检查差异是否大于1（根据所示的示例数据，这意味着差距）

---

假设没有重叠的日期，您可以使用

lag

获取上一个结束日期，并获取与当前行开始日期的差异，并检查差异是否大于1（根据所示的示例数据，这意味着差距）

---

如果是SQL Server>=2012，则可以使用lag，如下所示

Select *, GapDays = Coalesce(DateDiff(Day,lag(EndDate) over(partition by empid order by Startdate), StartDate)-1, 0)
    from #emp

输出如下：

+-------+------------+------------+---------+
| EmpId | StartDate  |  EndDate   | GapDays |
+-------+------------+------------+---------+
|   555 | 2015-08-07 | 2015-12-15 |       0 |
|   555 | 2015-12-16 | 2016-06-25 |       0 |
|   555 | 2016-06-28 | 2016-12-20 |       2 |
|   555 | 2016-12-21 | 9999-12-31 |       0 |
+-------+------------+------------+---------+

---

如果是SQL Server>=2012，则可以使用lag，如下所示

Select *, GapDays = Coalesce(DateDiff(Day,lag(EndDate) over(partition by empid order by Startdate), StartDate)-1, 0)
    from #emp

输出如下：

+-------+------------+------------+---------+
| EmpId | StartDate  |  EndDate   | GapDays |
+-------+------------+------------+---------+
|   555 | 2015-08-07 | 2015-12-15 |       0 |
|   555 | 2015-12-16 | 2016-06-25 |       0 |
|   555 | 2016-06-28 | 2016-12-20 |       2 |
|   555 | 2016-12-21 | 9999-12-31 |       0 |
+-------+------------+------------+---------+

我的做法如下：

create table #emp(EmpID int,StartDate date,EndDate date)

insert into #emp values
(555,'7/8/2015',  '15/12/2015'),
(555,'16/12/2015','25/06/2016'),
(555,'28/06/2016','20/12/2016'),    --Here 2 days gaps.
(555,'21/12/2016','31/12/9999')

查询：

select iq.EmpID,iq.StartDate,iq.EndDate, (DATEDIFF(day,iq.prev_date,iq.startdate)-1) as 'gap'
from
 (select *, lag(enddate) over (partition by EmpID order by empid, startdate) as prev_date
  from #emp
 )iq
where (DATEDIFF(day,iq.prev_date,iq.startdate)-1) > 0

输出：

EmpID       StartDate  EndDate     gap
----------- ---------- ----------  -----
555         2016-06-28 2016-12-20  2

我的做法如下：

create table #emp(EmpID int,StartDate date,EndDate date)

insert into #emp values
(555,'7/8/2015',  '15/12/2015'),
(555,'16/12/2015','25/06/2016'),
(555,'28/06/2016','20/12/2016'),    --Here 2 days gaps.
(555,'21/12/2016','31/12/9999')

查询：

select iq.EmpID,iq.StartDate,iq.EndDate, (DATEDIFF(day,iq.prev_date,iq.startdate)-1) as 'gap'
from
 (select *, lag(enddate) over (partition by EmpID order by empid, startdate) as prev_date
  from #emp
 )iq
where (DATEDIFF(day,iq.prev_date,iq.startdate)-1) > 0

输出：

EmpID       StartDate  EndDate     gap
----------- ---------- ----------  -----
555         2016-06-28 2016-12-20  2