Question

我们可以进一步优化此查询：

我需要得到：

SELECT * 
FROM table1 
WHERE (column1, column2) in (SELECT c1, c2 FROM table2);

由于不支持上述查询：

我有以下查询：

SELECT * 
FROM table1 join (SELECT c1, c2 from table2) as table3 
ON table1.column1 = c1 and table1.column2 = c2

编辑：为简单起见，我添加了table2。但它实际上是

select c1, min(c2) from table2 group by c1;

Answer 1

SELECT * from table1 t1
WHERE EXISTS (select *
   from table2  ex
   WHERE t1.column1 = ex.c1
     and t1.column2 = ex.c2
   );

更新：此处为MIN（c2）案例：

SELECT * from table1 t1
WHERE EXISTS (select *
   from table2 ex
   WHERE ex.c1 = t1.column1
     and ex.c2 = t1.column2
   )
AND NOT EXISTS (select *
   from table2 nx
   WHERE nx.c1 = t1.column1
     and nx.c2 < t1.column2
   );

Answer 2

如果我理解正确，不仅仅是

select * from table1 inner join table2 on table1.column1 = table2.c2 and table1.column1 = table2.c2;

Answer 3

答案是否定的。
你已经获得了最好的查询。 JOIN通常是最快的选择。但是，缺少的列别名会导致语法错误使用别名和更简单的USING子句修复和简化：

SELECT * 
FROM   table1 t1
JOIN  (
   SELECT c1 AS column1, min(c2) AS column2
   FROM   table2
   GROUP  BY 1
   ) t2 USING (column1, column2)

性能的关键是两个multi-column indices（唯一或主键约束也起作用）：

CREATE INDEX t1_mult_idx ON table1 (column1, column2);
CREATE INDEX t2_mult_idx ON table2 (c1, c2);

您可能对this related question on dba.SE感兴趣的是在此类查询中使用索引时出现的问题。

选择具有“IN”表达式的多个字段

3 个答案: