如何在需要过滤的两个表上保持联接？

问题描述

我有两个表需要连接在一起，但是两个表都需要先过滤。现在，我在sql Server中打开了两个表，并且都在特定列上选择了两个表，以便仅给出特定结果。基本上，如果这是Excel，我将过滤两个电子表格，然后过滤VLOOKUP。

恐怕我的左联接将使我得到表2的所有行（50,000行）的结果，而我实际上必须使用15,000行。

• 表1-供应商名称（3,000条记录）
• 表2-供应商名称（15,000条记录）

我只需要加入表格，就可以看到有多少1：1比赛回到表1。我们只是在寻找100％1：1比赛。

解决方法

如果您知道首先要获取表1中的所有记录，则可以将过滤条件添加到表2的JOIN的ON语句中，或添加到派生表2的JOIN上。

INNER JOIN过滤器示例：

SELECT
    *
FROM Table1 AS t1
INNER JOIN Table2 AS t2
    ON t1.SupplierId = t2.SupplierId
    AND t2.Column1 = ...condition
    AND t2.Column2 = ...condition
WHERE
   ...Table1 filters

在派生表上的INNER JOIN：

SELECT
    *
FROM Table1 AS t1
INNER JOIN (
   SELECT * FROM Table2 WHERE ...Table2 filters
) AS t2
    ON t1.SupplierId = t2.SupplierId
WHERE
   ...Table1 filters

要返回Table1中的所有行，而不管Table2中是否存在匹配项，则可以执行LEFT JOIN。

LEFT JOIN过滤器示例：

SELECT
    *
FROM Table1 AS t1
LEFT JOIN Table2 AS t2
    ON t1.SupplierId = t2.SupplierId
    AND t2.Column1 = ...condition
    AND t2.Column2 = ...condition
WHERE
   ...Table1 filters

在派生表上左联接：

SELECT
    *
FROM Table1 AS t1
LEFT JOIN (
   SELECT * FROM Table2 WHERE ...Table2 filters
) AS t2
    ON t1.SupplierId = t2.SupplierId
WHERE
   ...Table1 filters

更新：

我只希望表1中的一列实质上显示匹配项或NULL。

/*
    Show all rows in [2020_supplier1] where SetID = 'ID' 
    and a count of matching records from [2020supplier2]
*/

SELECT DISTINCT
    t1.[Supplier Name],n.Table2Count -- will be NULL without a match.
FROM [2020_supplier1] AS t1 
OUTER APPLY (

    SELECT COUNT ( * ) AS Table2Count FROM [2020supplier2] AS t2
    WHERE
        t2.[Vendor Name] = t1.[Supplier Name]
        AND t2.[Engagement Status] = 'your condition'
        -- any additional filters...

) AS n
WHERE 
    t1.[SetID] = 'ID';

如果您不关心[2020supplier2]计数，可以使用以下方法进一步简化此操作并节省处理时间：

SELECT DISTINCT
    t1.[Supplier Name],n.VendorMatch
FROM [2020_supplier1] AS t1 
OUTER APPLY (

    SELECT CASE 
        WHEN EXISTS (

            SELECT * FROM [2020supplier2] AS t2 WHERE
                t2.[Vendor Name] = t1.[Supplier Name]
                AND t2.[Engagement Status] = 'your condition'
                -- any additional filters...
        
        ) THEN 1
        ELSE 0
    END AS VendorMatch

) AS n
WHERE 
    t1.[SetID] = 'ID';

,

如果您不关心第二张表中的数据，则可以使用以下方法：

with cte as 
 ( -- return unique [Vendor Name] only
   select distinct [Vendor Name]
   from [2020supplier2]
   where [Engagement Status] = 'whatever'
 )
Select t1.*
   -- if there's no matching row the Outer Join returns NULL,case when t2.[Vendor Name] is null then 0 else 1 end as suppplier_exists
FROM [2020_supplier1] as t1
LEFT JOIN cte as t2
  ON t1.[Supplier Name] = t2.[Vendor Name]
WHERE t1.[SetID] = 'ID'