我有两张桌子，一张卡片，上面附有一张CardLines列表
Card (Unique: Code, Key)
----
Id | Code | Key
CardLine (FK: CardId, Unique: CardId, SubKey)
----
ID | CardId | Value | SubKey
考虑一些数据
Card
----
ID | CODE | Key
1 | C1 | K1
2 | C2 | K1
3 | C3 | K2
CardLine
----
ID | CardId | Value | SubKey
1 | C1 | 10 | SK1
2 | C1 | 20 | SK2
3 | C2 | 30 | SK1
4 | C2 | 40 | SK3
5 | C3 | 50 | SK4
这里，在Card表中，C2现在需要转换为C1（因为C2是错误的条目，它应该是C1）。为此，我尝试编写MERGE查询，如果匹配，我需要运行子查询以合并每个匹配的CardLines。像（伪代码）
的东西MERGE into Card as target
USING (SELECT ID, Code, Key from Card where Code = 'C2')
ON (target.Code = 'C1' and target.Key = K1)
WHEN MATCH
THEN
-- Merge the Cardlines in a similar mananner for a matched card (subquery?) - basically
-- Delete Card if no more lines assigned to it after merge
WHEN NOT MATCH
-- just update the Code from C2 -> C1
所需的输出
Card
----
ID | CODE | Key
1 | C1 | K1
3 | C3 | K2
CardLine
----
ID | CardId | Value | SubKey
1 | C1 | 40 | SK1 <-- merge of CardLine Id=1 and ID = 3, then 3 deleted
2 | C1 | 20 | SK2
4 | C1 | 40 | SK3 <-- no merge required, only product code update to C1 from C2
5 | C3 | 50 | SK4
我不是SQL专家（.Net），所以如果这不是正确的方法，会欣赏任何正确的方向