我想使用 CASE 语句来更新 sql server 2005 中的一些记录答案

【问题标题】：I want to use CASE statement to update some records in sql server 2005我想使用 CASE 语句来更新 sql server 2005 中的一些记录
【发布时间】：2013-03-23 20:45:28
【问题描述】：

UPDATE dbo.TestStudents  
SET LASTNAME = 
( CASE  
WHEN (LASTNAME = 'AAA') THEN 'BBB' 
WHEN (LASTNAME = 'CCC') THEN 'DDD' 
WHEN (LASTNAME = 'EEE') THEN 'FFF' 
ELSE  (LASTNAME)
END )

该语句用于此目的，但 else 条件扫描表中的每条记录。有什么办法可以让未受影响的行保持原样？

【问题讨论】：

标签： sql sql-server tsql sql-update

【解决方案1】：

添加WHERE 子句

UPDATE dbo.TestStudents  
SET     LASTNAME =  CASE  
                        WHEN LASTNAME = 'AAA' THEN 'BBB' 
                        WHEN LASTNAME = 'CCC' THEN 'DDD' 
                        WHEN LASTNAME = 'EEE' THEN 'FFF' 
                        ELSE LASTNAME
                    END 
WHERE   LASTNAME IN ('AAA', 'CCC', 'EEE')

【讨论】：

ELSE 语句永远不会运行，您可以删除它并像这样简单UPDATE dbo.TestStudents SET LASTNAME = CASE LASTNAME WHEN 'AAA' THEN 'BBB' WHEN 'CCC' THEN 'DDD' WHEN 'EEE' THEN 'FFF' END WHERE LASTNAME IN ('AAA', 'CCC', 'EEE')

【解决方案2】：

这也是 case-when...的另一种用法

UPDATE [dbo].[JobTemplates]
SET [CycleId] = 
    CASE [Id]
        WHEN 1376 THEN 44   --ACE1 FX1
        WHEN 1385 THEN 44   --ACE1 FX2
        WHEN 1574 THEN 43   --ACE1 ELEM1
        WHEN 1576 THEN 43   --ACE1 ELEM2
        WHEN 1581 THEN 41   --ACE1 FS1
        WHEN 1585 THEN 42   --ACE1 HS1
        WHEN 1588 THEN 43   --ACE1 RS1
        WHEN 1589 THEN 44   --ACE1 RM1
        WHEN 1590 THEN 43   --ACE1 ELEM3
        WHEN 1591 THEN 43   --ACE1 ELEM4
        WHEN 1595 THEN 44   --ACE1 SSTn     
        ELSE 0  
     END
WHERE
    [Id] IN (1376,1385,1574,1576,1581,1585,1588,1589,1590,1591,1595)

我喜欢在不允许重复值并且您的更新可能会创建它们的情况下使用临时表。例如：

SELECT
     [Id]
    ,[QueueId]
    ,[BaseDimensionId]
    ,[ElastomerTypeId]
    ,CASE [CycleId]
        WHEN  29 THEN 44
        WHEN  30 THEN 43
        WHEN  31 THEN 43
        WHEN 101 THEN 41
        WHEN 102 THEN 43
        WHEN 116 THEN 42
        WHEN 120 THEN 44
        WHEN 127 THEN 44
        WHEN 129 THEN 44
        ELSE    0
     END                AS [CycleId]
INTO
    ##ACE1_PQPANominals_1
FROM 
    [dbo].[ProductionQueueProcessAutoclaveNominals]
WHERE
    [QueueId] = 3
ORDER BY 
    [BaseDimensionId], [ElastomerTypeId], [Id];
---- (403 row(s) affected)

UPDATE [dbo].[ProductionQueueProcessAutoclaveNominals]
SET 
    [CycleId] = X.[CycleId]
FROM
    [dbo].[ProductionQueueProcessAutoclaveNominals]
INNER JOIN
(
    SELECT  
        MIN([Id]) AS [Id],[QueueId],[BaseDimensionId],[ElastomerTypeId],[CycleId] 
    FROM 
        ##ACE1_PQPANominals_1
    GROUP BY    
        [QueueId],[BaseDimensionId],[ElastomerTypeId],[CycleId] 
) AS X
ON
    [dbo].[ProductionQueueProcessAutoclaveNominals].[Id] = X.[Id];
----(375 row(s) affected)

【讨论】：

【解决方案3】：

如果您不想重复列表两次（根据@J W 的回答），则将更新放入表变量中并在UPDATE 中使用JOIN：

declare @ToDo table (FromName varchar(10), ToName varchar(10))
insert into @ToDo(FromName,ToName) values
 ('AAA','BBB'),
 ('CCC','DDD'),
 ('EEE','FFF')

update ts set LastName = ToName
from dbo.TestStudents ts
       inner join
     @ToDo t
       on
         ts.LastName = t.FromName

【讨论】：