尝试使用分区而不是子查询

时间:2012-09-03 22:27:30

标签: sql data-partitioning

我目前的系统中有两个表,其中包含当前的电子邮件,还有一个我创建并导出全局通讯簿中的所有电子邮件地址的表,以及一个BIT值,即主电子邮件地址和用户ID值。 我使用子查询编写了一个查询来查找应该更改的电子邮件地址,但是我应该能够在不使用子查询的情况下执行此操作,但无法弄清楚如何操作?

有没有人知道如何在不使用子查询的情况下更好地完成这项工作?

我有什么工作,但我想学习如何更好地编写SQL。

我虽然本应该能够使用OVER(Partition BY Member_ID)和MAX(CAST([primary] AS int))来做到这一点,但花了几个小时试图找出它并且是不合适的。

User_Emails
{
email varchar(75)
}

eg:
User_Emails
bob@oldemail.com
sue@evenolderemail.com
jim@otheremail.com
bill@oldemail.com

...

Global_Address_Book
{
Member_ID int
primary BIT
email_address varchar(75)
}

eg:
Member_ID        | primary | email_address
1                | 1       | bob@newemail.com
1                | 0       | bob@reallyoldemail.org
1                | 0       | bob@oldemail.com
2                | 1       | sue@newemail.com
2                | 0       | sue@oldemail.com
2                | 0       | sue@reallyoldemail.org
2                | 0       | sue@evenolderemail.com
3                | 1       | bill@newemail.com
3                | 0       | bill@oldemail.com
4                | 1       | jan@alternativeemail.com

...

SELECT Changed_Email_list.email AS Old_Email, New_Email_List.email_address AS New_Email
FROM (SELECT Member_ID, email
      FROM [Database].[Global_Address_Book] INNER JOIN [Database].[User_Emails]
       ON [Global_Address_Book].[email_address] = [User_Emails].[email]) AS     
      Changed_Email_list LEFT JOIN 
      (SELECT Member_ID, email_address 
       FROM [Database].[Global_Address_Book]
       WHERE [primary] = 1) AS New_Email_List
       ON Changed_Email_list.Member_ID = New_Email_List.Member_ID

1 个答案:

答案 0 :(得分:2)

这应该是我猜的

select UserEmails.Email as OldEmail,GAB.Email as NewEamil
from   UserEmailsfromAddressBook  
inner join UserEmails on UserEmails.Email=UserEmailsfromAddressBook.Email
      and isprimary =0
inner join UserEmailsfromAddressBook  GAB on GAB.memberId=UserEmailsfromAddressBook.memberID
     and GAB.isprimary=1

sqlfiddle

中的测试脚本