Question

我有一个从4 UNIONS收集用户名的查询。如何仅显示该结果中的重复用户？（理想情况下，只显示每个重复一次）。

谢谢！

以下是查询：

(SELECT user1 AS user FROM   friendships WHERE  user2 = 'johnsmith') 
UNION ALL 
(SELECT user2 AS user FROM   friendships WHERE  user1 = 'johnsmith') 
UNION ALL 
(SELECT user1 AS user FROM   friendships WHERE  user2 = 'johndoe') 
UNION ALL 
(SELECT user2 AS user FROM   friendships WHERE  user1 = 'johndoe')

Answer 1

您可以使用聚合：

SELECT user
FROM ((SELECT user1 AS user 
       FROM friendships 
       WHERE user2 = 'johnsmith' 
      ) UNION ALL 
      (SELECT user2 AS user 
       FROM friendships 
       WHERE user1 = 'johnsmith' 
      ) UNION ALL 
      (SELECT user1 AS user 
       FROM friendships 
       WHERE user2 = 'johndoe')
      ) UNION ALL 
      (SELECT user2 AS user 
       FROM friendships 
       WHERE user1 = 'johndoe'
      ) 
     ) u
GROUP BY user
HAVING COUNT(*) > 1;

您可以表达这一点，因此它只会传递一次数据：

SELECT (CASE WHEN user1 = 'johnsmith' THEN user2
             WHEN user2 = 'johnsmith' THEN user1
             WHEN user1 = 'johndoe' THEN user2
             WHEN user2 = 'johndoe' THEN user1
        END) as user
FROM friendships f
WHERE 'johnsmith' in (user1, user2) or 'johndoe' in (user1, user2)
GROUP BY user
HAVING COUNT(*) > 1;

请注意， 完全相同。它只计算'johnsmith'和'johndoe'之间的友谊。

Answer 2

SELECT user, count(*) FROM (
(SELECT user1 AS user FROM   friendships WHERE  user2 = 'johnsmith') 
UNION ALL 
(SELECT user2 AS user FROM   friendships WHERE  user1 = 'johnsmith') 
UNION ALL 
(SELECT user1 AS user FROM   friendships WHERE  user2 = 'johndoe') 
UNION ALL 
(SELECT user2 AS user FROM   friendships WHERE  user1 = 'johndoe') ) x
GROUP BY user
HAVING COUNT(*) >= 2

USE GROUP BY和HAVING

仅显示MySQL中UNION ALL查询的重复项

2 个答案: