SQL Server特有的缺勤期

Question

請你看看下面的任務嗎？

我有一個表DATA結構如下：

CREATE TABLE DATA 
(
    EMPLOYEE nvarchar(50) NOT NULL, 
    ABSENCE_START nvarchar(50) NOT NULL, 
    ABSENCE_END nvarchar(50) NOT NULL 
) 

它包含以下數據：

insert into DATA (EMPLOYEE, ABSENCE_START, ABSENCE_END) 
values ('emp01','21/06/2017','22/06/2017'), 
     ('emp01','24/06/2017','27/06/2017'), 
     ('emp02','22/06/2017','24/06/2017'), 
     ('emp02','22/06/2017','27/06/2017') 

正如你可以看到「emp02」員工，我們有沒有兩個時間段和其中一個（第4行）完全包含另一段缺席（第3行）。

我需要將這兩行合併爲一行（第4行）。

預期的結果將是：

EMPLOYEE ABSENCE_START ABSENSE 
--------------------------------------- 
emp01  21/06/2017  22/06/2017 
emp02  24/06/2017  27/06/2017 
emp01  22/06/2017  27/06/2017 

預先感謝您！

Answer 1

如果你正在尋找的唯一重疊校正是當一行完全包含另一個，你有沒有唯一的行標識符，您可以使用not exists()像這樣：

set dateformat dmy; 
;with cte as (
    select 
     employee 
    , absence_start= convert(date,absence_start) 
    , absence_end = convert(date,absence_end) 
    from data 
)  
select * 
from cte d 
where not exists (
    select 1 
    from cte i 
    where i.employee=d.employee 
     and i.absence_start <= d.absence_start 
     and i.absence_end >= d.absence_end 
     and (i.absence_start <> d.absence_start 
     or i.absence_end <> d.absence_end) 
    ) 

rextester演示：http://rextester.com/PHF88226

回報：

+----------+---------------+-------------+ 
| employee | absence_start | absence_end | 
+----------+---------------+-------------+ 
| emp01 | 2017-06-21 | 2017-06-22 | 
| emp01 | 2017-06-24 | 2017-06-27 | 
| emp02 | 2017-06-22 | 2017-06-27 | 
+----------+---------------+-------------+ 

Answer 2

查詢應該是這樣的：

select 
* 
from dbo.DATA t1 
where not exists (
      select 
      * 
      from dbo.DATA t2 
      where t1.ABSENCE_START >= t2.ABSENCE_START 
       and t1.ABSENCE_END < t2.ABSENCE_END) 

但是如果你有重疊的缺席。像這樣：

insert into DATA(EMPLOYEE,ABSENCE_START,ABSENCE_END) values 
('emp01','21/06/2017','22/06/2017'), 
('emp01','24/06/2017','27/06/2017'), 
('emp02','22/06/2017','29/06/2017'), 
('emp02','22/06/2017','27/06/2017')