是否可以为 select 中的每一行更新一个变量

Question

在MySQL中可以像

一样轻松完成

set @var= (select value from my_sequence);

insert into new_table (id, name)
select @var := @var + 1, name from old_table;

但我找不到在 SQL 服务器中做同样事情的方法。可能吗？有哪些替代方案？

Answer 1

只需使用ROW_NUMBER:

DECLARE @Var int = (SELECT [value] FROM dbo.my_sequence WITH (UPDLOCK)); 
--Ensure the row can't be used while we're doing it, as I assume that
--the value would be updated later with the new sequence value

INSERT INTO dbo.new_table (id, name)
SELECT ROW_NUMBER() OVER (ORDER BY {Column to Order By}) + @Var,
       [name]
FROM dbo.old_table;

你甚至可以在没有变量的情况下实现这一点：

INSERT INTO dbo.new_table (id, name)
SELECT ROW_NUMBER() OVER (ORDER BY {Column to Order By}) + ms.[value],
       ot.[name]
FROM dbo.old_table ot
     CROSS JOIN dbo.my_sequence ms WITH (UPDLOCK);

作为旁注，语法 SELECT @Variable = @Variable + ... FROM 是 SQL 服务器中的 documented antipattern，应避免使用。

是否可以为 select 中的每一行更新一个变量

Is it possible to update a variable for each row in select

sql

sql-server