Question

如何在SQL中完成此连接？

TABLE1

+----+-----------+-----------+---------+
| ID | FILTER1   | FILTER2   | DATA1   |
| 1  | filter1-A | filter2-A | data1-A |
| 2  | filter1-B | filter2-B | data1-B |
+----+-----------+-----------+---------+

TABLE2

+----+-----------+-----------+---------+
| ID | FILTER1   | FILTER2   | DATA1   |
| 1  | filter1-B | filter2-B | data2-B |
| 2  | filter1-C | filter2-C | data2-C |
+----+-----------+-----------+---------+

结果

+------------+-----------+---------+---------+
| FILTER1    | FILTER2   | DATA1   | DATA2   | 
| filter1-A  | filter2-A | data1-A | NULL    | 
| filter1-B  | filter2-B | data1-B | data2-B | 
| filter1-C  | filter2-C | NULL    | data2-C | 
+------------+-----------+---------+---------+

Answer 1

SELECT
  COALESCE(TABLE1.filter_1, TABLE2.filter_1)   AS filter_1,
  COALESCE(TABLE1.filter_1, TABLE2.filter_2)   AS filter_2,
  TABLE1.data1                                 AS data_1,
  TABLE2.data2                                 AS data_2
FROM
  TABLE1
FULL OUTER JOIN
  TABLE2
    ON  TABLE1.filter_1 = TABLE2.filter_1
    AND TABLE1.filter_2 = TABLE2.filter_2

FULL OUTER JOIN保留每个表中的每条记录，无论另一个表中是否匹配。

COALESCE() （有些使用ISNULL()）然后可用于扫描缺失/ NULL值以查找第一个非NULL值。

Answer 2

这称为FULL OUTER JOIN。

SELECT 
  ISNULL(T1.FILTER1, T2.FILTER1) AS FILTER_1, 
  ISNULL(T1.FILTER2, T2.FILTER2) AS FILTER_2,
  T1.DATA1 AS DATA_1, 
  T2.DATA1 AS DATA_2
FROM TABLE1 T1
FULL OUTER JOIN TABLE2 T2
  ON T1.FILTER1 = T2.FILTER1
 AND T1.FILTER2 = T2.FILTER2

Answer 3

SELECT G.FILTER1, G.FILTER2, T1.DATA1, T2.DATA1 as DATA2 FROM
    (SELECT FILTER1, FILTER2
    FROM TABLE1
  UNION
    SELECT FILTER1, FILTER2
    FROM TABLE2
  GROUP BY FILTER1, FILTER2) as G
LEFT JOIN TABLE1 as T1 ON T1.FILTER1 = G.FILTER1 AND T1.FILTER2 = G.FILTER2
LEFT JOIN TABLE2 as T2 ON T2.FILTER1 = G.FILTER1 AND T2.FILTER2 = G.FILTER2

Answer 4

您也可以使用union all和group by执行此操作 - 假设表中的两个表都没有重复项：

select filter_1, filter_2, max(data_1) as data_1, max(data_2) as data_2
from ((select filter_1, filter_2, data_1, NULL as data_2
       from table1
      ) union all
      (select filter_1, filter_2, NULL, data_2
       from table2
      )
     ) t
group by filter_1, filter_2;

我提供此选择。首先，（对我而言）union all / group by的行为与full outer join相同是很有趣的。更重要的是，如果您开始添加更多表，那么full outer join方法会变得很麻烦。将此方法扩展到更多表格很容易。

如何在SQL中完成此连接？

4 个答案: