Question

如何在以下T-SQL语句中避免重复的子查询：

update r set
  column1=(select top 1 max(column1) from region r2 where (overlapping test conditions)),
  column2=(select top 1 max(column2) from region r2 where (overlapping test conditions))
from region r

基本上，我有一个包含矩形区域的表格。对于每个重叠区域，我需要将一些列设置为该重叠区域的聚合值。

由于

编辑：我在这里添加了一个过度简化的重叠条件：

(r.left >= r2.left and r.left < r2.right) or (r.right <= r2.right and r.right > r.left)

重点是：r和r2都将在子查询中引用。对于常见的表表达式来说，这似乎是一个完美的情况，但我无法弄清楚如何将它用于每个记录。

Answer 1

UPDATE Region
SET
  Region.Column1 = r.Column1
  Region.Column2 = r.Column2
FROM
  Region
INNER JOIN 
 (
   SELECT
        Column1,
        Column2
   FROM Region
   WHERE (your condition here)
  ) r ON r.ID = Region.ID

我刚注意到MAX（）aggs。对不起第一次看到了。你可能必须有两个内连接才能使它工作......

所以拿我所拥有的并为column1

做

INNER JOIN (SELECT MAX(Column1) FROM ... WHERE ... ) r ON r.ID = Region.ID

然后是第二次INNER JOIN

INNER JOIN (SELECT MAX(Column2) FROM ... WHERE ...)r2 ON r2.ID = Region.ID

您的更新将变为

UPDATE Region
SET Region.Column1 = r1.Column1,
    Region.Column2 = r2.Column2
FROM Region
INNER JOIN
(SELECT MAX(Column1) as Column1 FROM ... WHERE ... ) r ON r.ID = Region.ID
INNER JOIN
(SELECT MAX(Column2) As Column2 FROM ... WHERE ... ) r2 ON r2.ID = Region.ID

从子查询更新多个列？

1 个答案: