2

我需要专家帮助。我认为我非常了解 SQL，至少到目前为止。我正在使用 SQL Server，需要构建一个查询来列出每个注册成员，并在单独的列而不是行中显示他们的来宾的每个名称。每个成员可以有 0:N 个客人与每个成员相关联。所以，这是我的表：

成员表：memberID (PK), FName, LName, ...

来宾表：GuestID (PK)、FName、LastName、...

事件表：GuestID (PK)、MemberID (PK)、...

我正在尝试构建这样的查询输出：

会员ID | Member_FName | Member_LName | 来宾1ID | 来宾1_Fname | 来宾2ID | 来宾2_FName ...

样本：

Member Table
MemberID | FName | LName
001        Frank   Smith   
002        Mary    Jane
003        John    Henry

Guest Table
GuestID | FName | LName
101       Steve   Smith
102       Peter   Smith
103       Mike    Jane

Event Table
MemberID | GuestID
001        101
001        102
002        103

输出：

MemberID | FName | LName| GuestID1 | FName1 | LName1 |GuestID2 | FName2 | LName2
001        Frank   Smith  101        Steve    Smith   102        Peter    Smith
002        Mary    Jane   103        Mike     Jane
003        John    Henry

如果我需要包含其他信息，请告诉我。

提前致谢！

4

2 回答 2

3

您可以同时实现 theUNPIVOT和 thenPIVOT函数来获得结果。获取您的UNPIVOT列并将数据转换为行，而数据透视表获取最终结果并将其转换回列：

select MemberID,
  memberfirst,
  memberlast,
  isNull(GuestId_1, '') GuestId_1, 
  isNull(fname_1, '') fname_1, 
  isNull(lname_1, '') lname_1,
  isNull(GuestId_2, '') GuestId_2, 
  isNull(fname_2, '') fname_2, 
  isNull(lname_2, '') lname_2
from
(
  select MemberID,
    memberfirst,
    memberlast,
    col+'_'+cast(rn as varchar(10)) col,
    value
  from
  (
    select m.MemberID,
      m.fname MemberFirst,
      m.lname MemberLast,
      isNull(cast(g.GuestID as varchar(5)), '') GuestId,
      isNull(g.fname, '') fname,
      isNull(g.lname, '') lname,
      row_number() over(partition by m.parentid order by g.guestid) rn
    from member m
    left join Event r
      on m.parentid = r.memberid
    left join guest g
      on r.guestid = g.guestid
  ) src
  unpivot
  (
    value
    for col in (GuestId, fname, lname)
  ) unpiv
) src1
pivot
(
  max(value)
  for col in (GuestId_1, fname_1, lname_1,
              GuestId_2, fname_2, lname_2)
) piv

请参阅带有演示的 SQL Fiddle

如果您提前知道记录的数量，上述方法非常有用，但如果不知道，那么您将需要使用动态 sql：

DECLARE  @query  AS NVARCHAR(MAX),
    @colsPivot as  NVARCHAR(MAX),
    @colsPivotNull as  NVARCHAR(MAX)

select @colsPivot = STUFF((SELECT  ',' 
                      + quotename(c.name +'_'+ cast(t.rn as varchar(10)))
                    from
                    (
                      select cast(row_number() over(partition by m.MemberID order by g.guestid) as varchar(50)) rn
                      from member m
                      left join Event r
                        on m.parentid = r.memberid
                      left join guest g
                        on r.guestid = g.guestid
                    ) t
                    cross apply sys.columns as C
                   where C.object_id = object_id('guest')
                   group by c.name, t.rn
                   order by t.rn
            FOR XML PATH(''), TYPE
            ).value('.', 'NVARCHAR(MAX)') 
        ,1,1,'')

select @colsPivotNull = STUFF((SELECT  ', IsNull(' 
                      + quotename(c.name +'_'+ cast(t.rn as varchar(10)))+', '''') as '+c.name +'_'+ cast(t.rn as varchar(10))
                    from
                    (
                      select cast(row_number() over(partition by m.MemberID order by g.guestid) as varchar(50)) rn
                      from member m
                      left join Event r
                        on m.parentid = r.memberid
                      left join guest g
                        on r.guestid = g.guestid
                    ) t
                    cross apply sys.columns as C
                   where C.object_id = object_id('guest')
                   group by c.name, t.rn
                   order by t.rn
            FOR XML PATH(''), TYPE
            ).value('.', 'NVARCHAR(MAX)') 
        ,1,1,'')

set @query 
  = 'select 
          MemberID,
          memberfirst,
          memberlast, '+@colsPivotNull+'
      from
      (
        select MemberID,
          memberfirst,
          memberlast,
          col+''_''+cast(rn as varchar(10)) col,
          value
        from 
        (
          select m.MemberID,
            m.fname MemberFirst,
            m.lname MemberLast,
            isNull(cast(g.GuestID as varchar(5)), '''') GuestId,
            isNull(g.fname, '''') fname,
            isNull(g.lname, '''') lname,
            row_number() over(partition by m.parentid order by g.guestid) rn
          from member m
          left join Event r
            on m.parentid = r.memberid
          left join guest g
            on r.guestid = g.guestid
        ) x
        unpivot
        (
         value
          for col in (GuestId, fname, lname)
        ) u
      ) x1
      pivot
      (
        max(value)
        for col in ('+ @colspivot +')
      ) p'

exec(@query)

请参阅带有演示的 SQL Fiddle

两个版本产生相同的结果：

| MemberID | MEMBERFIRST | MEMBERLAST | GUESTID_1 | FNAME_1 | LNAME_1 | GUESTID_2 | FNAME_2 | LNAME_2 |
-------------------------------------------------------------------------------------------------------
|        1 |       Frank |      Smith |       101 |   Steve |   Smith |       102 |   Peter |   Smith |
|        2 |        Mary |       Jane |       103 |    Mike |    Jane |           |         |         |
|        3 |        John |      Henry |           |         |         |           |         |         |
于 2013-01-09T20:46:22.813 回答
0

实际上，为客人制作一列并用逗号分隔列表更容易。

你可以这样做：

SELECT DISTINCT memberID, FName, LName, 
 STUFF((SELECT ', ' + FName + ' ' + LName + ' ('+GuestID+')'
  FROM Relationship R 
  JOIN Guest G ON R.GuestID = G.GuestID
  WHERE R.MemberID = Memeber.MemberID
  FOR XML PATH ('')),1,2,'') AS GuestList
FROM Member

看起来像这样：

MemberID | FName | LName| GuestList
001        Frank   Smith  Steve Smith (101), Peter Smith (102)
002        Mary    Jane   Mike Jane (103)
003        John    Henry
于 2013-01-09T20:35:41.137 回答