SQL Server 从两个表创建视图

Question

我在一个跟踪社交网络的 SQL Server 数据库中有两个表。首先是friends表，看起来像

userId | friendId | acceptDate
==============================
   1   |     8    | 2018-01-20
   8   |     12   | 2017-11-20
   12  |     1    | 2017-12-18

所以每一对都是唯一的组合，顺序只是由哪个用户发起连接来决定的。

第二张表是有用户信息的表

userId | name | email | ...
   1   | John  | john@example.com  | ...
   2   | Bill  | bill@example.com  | ...
   3   | Cathy | cathy@example.com | ...

我想创建一个视图，在该视图中，我可以通过最少的服务器端处理轻松获取特定用户的朋友所需的数据。所以本质上我想创建一个数据看起来像的视图：

userId | friendID | friendName | friendEmail | acceptDate | ....
   1   |     8    |  8's Name  |  8's Email  | 2018-01-20 |...
   1   |    12    |  12's Name |  12's Email | 2017-12-18 |...
   8   |    12    |  12's Name |  12's Email | 2017-11-20 |...
   8   |     1    |  1's Name  |  1's Email  | 2018-01-20 |...
  12   |     1    |  1's Name  |  1's Email  | 2017-12-18 |...
  12   |     8    |  8's Name  |  8's Email  | 2017-11-20 |...

基本上，它将从朋友表中存在的每个友谊对创建两行（一个按现有顺序，一个翻转），并将加入用户表中的朋友信息。这样，我可以简单地查询视图以找到我所有特定用户的朋友，它只会返回该用户的朋友的行及其信息，我无需进行任何处理来整理。

任何帮助将不胜感激。

谢谢！

Answer 1

这是一种可以通过 JOIN 完成的方法：

SELECT u.UserID, fi.UserID AS FriendID, fi.Name, fi.email, f.acceptdate
FROM UserInfo u
INNER JOIN Friends f
  ON u.UserID=f.UserID
  OR u.UserID=f.FriendID
INNER JOIN UserInfo fi
  ON (fi.UserID=f.UserID AND u.UserID=f.FriendID)
  OR (fi.UserID=f.FriendID AND u.UserID=f.UserID)
ORDER BY u.UserID, fi.UserID

Answer 2

因为您必须切换前两列中的所有数据，所以您需要将数据量翻倍。这可能是使用 UNION ALL 最简单的方法。这样的事情应该很接近了。

select *
from
(
    select UserID
        , FriendID
        , FriendName = sot.name
    from SomeTable st
    join SomeOtherTable sot on sot.UserID = st.UserID

    UNION ALL

    select FriendID
        , UserID
        , FriendName = sot.name
    from SomeTable st
    join SomeOtherTable sot on sot.UserID = st.FriendID
) x
order by x.UserID
    , x.FriendID

Answer 3

-- Friends I accepted 
select userid, f.userId friendID, f.name friendName, f.email friendEmail, friends.acceptDate
from users inner join friends on users.userId = friends.userId inner join users f on friends.friendId = f.Userid

union all
-- friends who accepted me. 
select userid, f.userId friendID, f.name friendName, f.email friendEmail, friends.acceptDate
from users inner join friends on users.userId = friends.friendID inner join users f on friends.userid = f.Userid

Answer 4

这是一个工作示例：

编辑，我已经添加了来自用户 ID 的电子邮件，并添加了 union all 以支持 ops 之心，快乐 v-day。

declare @user as table (
    userID       int identity(1,1) not null primary key clustered
,   Name_        nvarchar(255) not null
,   EmailAddress nvarchar(255) not null
);

declare @duo as table (
    duoID   int identity(1,1) not null primary key clustered 
,   userID  int not null
,   loserID int not null
,   LoveDate date not null
);

insert into @user (name_, EmailAddress)

select 'bob', 'bob@example.com' union all
select 'tom', 'tom@example.com' union all
select 'nick', 'nick@example.com' union all
select 'Random', 'awesomesauce@example.com';

insert into @duo (userid, loserid, LoveDate)
select 1, 2, '2018-01-14' union all
select 1, 3, '2018-02-14' union all
select 1, 4, '2018-01-15' union all
select 2, 3, '2018-01-14' union all
select 2, 4, '2018-01-18' union all
select 3, 4, '2018-01-13';


select a.UserID
     , a.Name_ Person
     , a.EmailAddress
     , c.userID LoserID
     , c.Name_ Loser
     , c.EmailAddress 'Loser''s Email'
     , b.LoveDate
  from @user a
  join @duo b
    on a.userid = b.userid
  join @user c
    on b.loserid = c.userid

union all

select c.userID 
     , c.Name_ Person
     , c.EmailAddress
     , a.UserID LoserID
     , a.Name_ Loser
     , a.EmailAddress 'Loser''s Email'
     , b.LoveDate
  from @user a
  join @duo b
    on a.userid = b.userid
  join @user c
    on b.loserid = c.userid