Question

我在理解这个相关子查询时遇到了问题：

SELECT Country, CustomerID,
   (SELECT COUNT(*)
    FROM Customers AS CustomersInner
    WHERE CustomersInner.CustomerID < CustomersOuter.CustomerID
    AND CustomersInner.Country=CustomersOuter.Country) + 1
AS customer_seq_num
FROM Customers AS CustomersOuter ORDER BY Country;

此示例来自http://www.w3schools.com/sql/trysql.asp?filename=trysql_select_all

此查询如何正常工作？为什么在CustomersInner.CustomerID＆lt;之间进行比较？ CustomersOuter.CustomerID？如何解释+1？更深入的解释会很棒。

Answer 1

对主查询中的每一行执行相关子查询。在这种情况下，主查询是

SELECT Country, CustomerID
  FROM Customers As CustomersOuter
 ORDER BY Country

此查询按国家/地区返回所有客户ID。

Country   | Customer ID
-----------------------
Argentina | 12
Argentina | 54
Argentina | 64
Austria   | 20
Austria   | 59

对每个国家/客户ID对执行相关子查询。在这种情况下：

第一行：

SELECT COUNT(*)
  FROM Customers AS CustomersInner
 WHERE CustomersInner.CustomerID < 12
   AND CustomersInner.Country='Argentina') + 1

由于没有ID＆lt;的客户12在阿根廷count(*)返回零，+1得到1

第二行：

SELECT COUNT(*)
  FROM Customers AS CustomersInner
 WHERE CustomersInner.CustomerID < 54
   AND CustomersInner.Country='Argentina') + 1

结果将是1（customerID = 12）+ 1 = 2

依旧......

当您更改国家/地区时，count(*)会再次从零开始，因此您有一个正在运行的计数器：因为每个国家都会计算客户数量。

Country   | Customer ID | customer_seq_num
------------------------------------------
Argentina | 12          | 1
Argentina | 54          | 2
Argentina | 64          | 3
Austria   | 20          | 1    <--- change country, reset counter (no customerID < 20 in 'Austria')
Austria   | 59          | 2

SQL - 相关子查询

1 个答案: