如何通过组合来自 3 个不同表的数据来获得如下输出

Question

所以我的数据库中有 3 个 table，其中所有 3 个 table 都包含一个具有相似数据的列，但该列的名称在所有 3 tables，下面是一个例子。

禁止 Table

user_id	ban_user_id	ban_date	reason	end_date
1300	1	xyz	xyz	xyz
32	1	xyz	xyz	xyz
43	2	xyz	xyz	xyz

报告Table

user_id	last_modified_user_id	report_date	reason	end_date
1300	1	xyz	xyz	xyz
32	2	xyz	xyz	xyz
43	2	xyz	xyz	xyz

警告Table

user_id	warning_user_id	warning_date	reason	end_date
1300	1	xyz	xyz	xyz
32	2	xyz	xyz	xyz
43	3	xyz	xyz	xyz

现在我想把这3个table组合起来取数据，其中ban_user_id、last_modified_user_id、warning_user_id是拿的工作人员的数据操作，所以我想按员工 ID 对数据进行分组。

我要找的输出如下：

staff_id	total_reports	total_bans	total_warnings
1	1	2	1
2	2	1	1
3	0	0	1

它通过对第 2 列 ban_user_id、last_modified_user_id、warning_user_id 进行分组来计算每个 table 的数据。而不是合并数据。

我尝试了 UNION All 之类的东西，但没有成功。

提前感谢您的帮助

Answer 1

您可以使用 Table joins (Inner/outer/left/right) 来获取数据而不是并集。我假设 staff_id 等同于 user_id 列，因为你没有提到任何相关内容，所以你的脚本看起来像这样：

SELECT W.user_id AS staff_id, 
    B.ban_user_id, 
    R.last_modified_user_id, 
    W.warning_user_id
FROM Warning AS W
LEFT JOIN Reports AS R on R.user_id = W.user_id
LEFT JOIN Ban AS B on B.user_id = W.user_id
group by W.user_id

Answer 2

对所有 3 个表使用 UNION ALL，然后聚合：

SELECT staff_id,
       COUNT(report) AS total_reports,
       COUNT(ban) AS total_bans,
       COUNT(warning) AS total_warnings
FROM (
  SELECT last_modified_user_id AS staff_id, 1 AS report, null AS ban, null AS warning FROM Reports
  UNION ALL
  SELECT ban_user_id, null, 1, null FROM Ban
  UNION ALL
  SELECT warning_user_id, null, null, 1 FROM Warning
) t
GROUP BY staff_id;

或者：

SELECT staff_id,
       SUM(report) AS total_reports,
       SUM(ban) AS total_bans,
       SUM(warning) AS total_warnings
FROM (
  SELECT last_modified_user_id AS staff_id, 1 AS report, 0 AS ban, 0 AS warning FROM Reports
  UNION ALL
  SELECT ban_user_id, 0, 1, 0 FROM Ban
  UNION ALL
  SELECT warning_user_id, 0, 0, 1 FROM Warning
) t
GROUP BY staff_id;

参见demo。

如何通过组合来自 3 个不同表的数据来获得如下输出

How to get the output as following by combining data from 3 different tables

mysql

sql

union

group-by

count