很难用语言解释这一点，所以我将尝试一个小的例子来向您解释我想要的东西：
例如，我有此SQL-Server表
+-------+-------+------+
| TestID| Start | End |
+-------+-------+------+
| 1 | DateA | DateB|
| 2 | DateA | DateB|
| 3 | DateA | DateB|
| 4 | DateA | |
+-------+-------+------+
我想要的是这张桌子：
$ Rest = 1000
+-------+-------+------+----------+-----------+----------+
| TestID| Start | End | Testtime | Totaltime | Resttime |
+-------+-------+------+----------+-----------|----------|
| 1 | DateA | DateB| 214 | 214 | 786 |
| 2 | DateA | DateB| 100 | 314 | 686 |
| 3 | DateA | DateB| 200 | 514 | 486 |
| 4 | DateA | | | | |
+-------+-------+------+----------+-----------+----------+
我有一个问题要理解我必须结合的东西
这是获取第一个SQL-Server表的代码
SELECT TestID, Start, End, DATEDIFF(hour, Start, End) AS Testtime
FROM Testresults
WHERE TesttableID = 1
每行都有TesttableID = 1 thx来寻求帮助。
编辑：SQL Server版本：9.0.5057
编辑：我得到的是结果，但没有得到正确的结果，它们在Totaltime和Resttime之间切换
SELECT t1.TestID,
t1.start,
t1.end,
t1.TesttableID,
DATEDIFF(hour,t1.start,t1.end) as Testtime,
(SELECT SUM(DATEDIFF(hour,t2.start,t2.end))
FROM Testresults t2
WHERE t2.TestID <= t1.TestsID AND t2.TesttableID = 1 ) AS Totaltime,
(SELECT 1000-SUM(DATEDIFF(hour,t2.start,t2.end))
FROM Testresults t2
WHERE t2.TestID <= t1.TestIDAND t2.TesttableID = 1 ) AS Resttime FROM Testresults t1 WHERE t1.TesttableID = 1
我得到的是这些结果，它们切换了..：
+-------+-------+------+----------+-----------+----------+
| TestID| Start | End | Testtime | Totaltime | Resttime |
+-------+-------+------+----------+-----------|----------|
| 1 | DateA | DateB| 214 | 514 | 486 |
| 2 | DateA | DateB| 100 | 300 | 700 |
| 3 | DateA | DateB| 200 | 200 | 800 |
| 4 | DateA | | | | |
+-------+-------+------+----------+-----------+----------+
答案 0 :(得分：2)
您可以尝试以下方法：
3'b100对于SQL Server 2005，您可以使用SELECT TestId,
Start,
End,
Testtime,
SUM(Testtime) OVER (ORDER BY TestId ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW),
1000 - SUM(Testtime) OVER (ORDER BY TestId ROWS BETWEEN AND UNBOUNDED PRECEDING AND CURRENT ROW)
FROM MyTable
完成此操作：
JOIN答案 1 :(得分：1)
基于样本数据，我们可以为您提供结果：
DECLARE @Rest int = 1000;
WITH VTE AS (
SELECT *
FROM (VALUES(1,'DateA','DateB',214),
(2,'DateA','DateB',100),
(3,'DateA','DateB',200),
(4,'DateA',NULL,NULL)) V(TestID,[Start],[End],Testtime))
SELECT VTE.TestID,
VTE.Start,
VTE.[End],
VTE.Testtime,
CASE WHEN [End] IS NOT NULL THEN SUM(VTE.Testtime) OVER (ORDER BY VTE.TestID ASC
ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) END AS TotalTime,
CASE WHEN [End] IS NOT NULL THEN @Rest - SUM(VTE.Testtime) OVER (ORDER BY VTE.TestID ASC
ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) END AS RestTime
FROM VTE;
请注意，您可能需要在PARTITION BY子句中实现某种OVER，但是我不知道基于我们的受限数据需要什么。
答案 2 :(得分：1)
您可以尝试：
CREATE TABLE #MyTable
(PrimaryKey int PRIMARY KEY,
DateValueBegin DATETIME,
DateValueEnd DATETIME,
NbValue int
);
GO
INSERT INTO #MyTable
SELECT 1, DATEADD(HOUR,-2,GETDATE()), GETDATE(), 214
UNION
SELECT 2, DATEADD(HOUR,-2,GETDATE()), DATEADD(HOUR,-1,GETDATE()), 100
UNION
SELECT 3, DATEADD(HOUR,-2,GETDATE()), GETDATE(), 200
UNION
SELECT 4, DATEADD(HOUR,-1,GETDATE()), NULL, 210
UNION
SELECT 5, DATEADD(HOUR,-1,GETDATE()), NULL, 0;
SELECT *, SUM(NbValue) OVER(ORDER BY PrimaryKey
ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) as 'Totaltime',
1000 - SUM(NbValue) OVER(ORDER BY PrimaryKey
ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) As 'RestTime' FROM #MyTable
DROP TABLE #MyTable
在此帖子中说明：Calculate a Running Total in SQL Server
结果：
PrimaryKey DateValueBegin DateValueEnd NbValue Totaltime RestTime
1 2019-01-17 09:48:05.123 2019-01-17 11:48:05.123 214 214 786
2 2019-01-17 09:48:05.123 2019-01-17 10:48:05.123 100 314 686
3 2019-01-17 09:48:05.123 2019-01-17 11:48:05.123 200 514 486
4 2019-01-17 10:48:05.123 NULL 210 724 276
5 2019-01-17 10:48:05.123 NULL 0 724 276
要使用以前的SQL Server版本的用户，请查看以下信息：http://geekswithblogs.net/Rhames/archive/2008/10/28/calculating-running-totals-in-sql-server-2005---the-optimal.aspx
版本对于先前的SQL版本：
CREATE TABLE #MyTable
(PrimaryKey int PRIMARY KEY,
DateValueBegin DATETIME,
DateValueEnd DATETIME
);
GO
INSERT INTO #MyTable
SELECT 1, DATEADD(HOUR,-214,GETDATE()), GETDATE()
UNION
SELECT 2, DATEADD(HOUR,-100,GETDATE()), DATEADD(HOUR,-1,GETDATE())
UNION
SELECT 3, DATEADD(HOUR,-200,GETDATE()), GETDATE()
UNION
SELECT 4, DATEADD(HOUR,-1,GETDATE()), NULL
UNION
SELECT 5, DATEADD(HOUR,-1,GETDATE()), NULL;
SELECT * FROM #MyTable
SELECT PrimaryKey,
DateValueBegin,
DateValueEnd,
DATEDIFF(hour,DateValueBegin,IIF(DateValueEnd IS NOT NULL ,DateValueEnd, DateValueBegin)) as Testtime,
(SELECT SUM(DATEDIFF(hour,DateValueBegin,IIF(DateValueEnd IS NOT NULL ,DateValueEnd, DateValueBegin)))
FROM #MyTable t2
WHERE t2.PrimaryKey <= t1.PrimaryKey ) AS Totaltime,
DATEDIFF(hour,DateValueBegin,IIF(DateValueEnd IS NOT NULL ,DateValueEnd, DateValueBegin)) as Testtime,
(SELECT 1000-SUM(DATEDIFF(hour,DateValueBegin,IIF(DateValueEnd IS NOT NULL ,DateValueEnd, DateValueBegin)))
FROM #MyTable t3
WHERE t3.PrimaryKey <= t1.PrimaryKey ) AS Resttime
FROM #MyTable t1
DROP TABLE #MyTable
结果：
PrimaryKey DateValueBegin DateValueEnd Testtime Totaltime Testtime Resttime
1 2019-01-08 18:17:35.430 2019-01-17 16:17:35.430 214 214 214 786
2 2019-01-13 12:17:35.430 2019-01-17 15:17:35.430 99 313 99 687
3 2019-01-09 08:17:35.430 2019-01-17 16:17:35.430 200 513 200 487
4 2019-01-17 15:17:35.430 NULL 0 513 0 487
5 2019-01-17 15:17:35.430 NULL 0 513 0 487
对不起，但我不知道您的所有配置 你可以试试吗：
SELECT t1.TestID,
t1.start,
t1.end,
t1.TesttableID,
DATEDIFF(hour,t1.start,t1.end) as Testtime,
(SELECT SUM(DATEDIFF(hour,t2.start,t2.end))
FROM Testresults t2
WHERE t2.TestID <= t1.TestsID AND t2.TesttableID = 1 ) AS Totaltime,
(SELECT 1000-SUM(DATEDIFF(hour,t2.start,t2.end))
FROM Testresults t2
WHERE t2.TestID <= t1.TestID AND t2.TesttableID = 1 ) AS Resttime FROM Testresults t1 WHERE t1.TesttableID = 1