Question

我有这三个表：

表1：

    InvoiceId         InvoiceNumber        CompanyId
    -------------------------------------------------
       3                   1                  2
       4                   2                  2
       6                   1                  1
       7                   3                  2
       8                   4                  2
       9                   2                  1

表2：

    CompanyId          CompanyName
    -------------------------------
       1                  Name1
       2                  Name2

表3：

    InvoiceId        AcceptDate       AcceptType
    ---------------------------------------------
       3                AAAA               O
       3                BBBB               P
       6                CCCC               P
       4                DDDD               O
       7                EEEE               O
       9                FFFF               P
       7                GGGG               P

表1和表3具有一对多的关系，表3具有多个链接的行。表3每个InvoiceId最多有2行。

任务

我想将表3行分组以得到这个结果：

    InvoiceId       AcceptDate1      AcceptDate2
    --------------------------------------------
        3              AAAA             BBBB
        6              CCCC             null
        4              DDDD             null
      and ....

然后将这些结果与表1和表2结合起来给出：

    InvoiceId     InvoiceNumber        AcceptDate1    AcceptDate2      Name
    --------------------------------------------------------------------------
       3               1                  AAAA           BBBB          Name2
       4               2                  DDDD           null          Name2
       7               3                  EEEE           GGGG          Name2
       8               4                  null           null          Name2
       6               1                  CCCC           null          Name1
       9               2                  FFFF           null          Name1

AcceptDate1和AcceptDate2是静态的。

Answer 1

你想要的是Pivot the data。

BEGIN TRAN

create table #table1 (InvoiceId int, InvoiceNumber int, CompanyId int)
create table #table2 (CompanyId int, CompanyName varchar(50))
create table #table3 (InvoiceId int, AcceptDate datetime, AcceptType char(1))
insert into #table1 values (3, 1, 2)
insert into #table1 values (4, 2, 2)
insert into #table1 values (6, 1, 1)
insert into #table1 values (7, 3, 2)
insert into #table1 values (8, 4, 2)
insert into #table1 values (9, 2, 1)
insert into #table2 values (1, 'Name1')
insert into #table2 values (2, 'Name2')
insert into #table3 values (3, '2012-01-01', 'o')
insert into #table3 values (3, '2012-01-02', 'p')
insert into #table3 values (6, '2012-01-03', 'p')
insert into #table3 values (4, '2012-01-04', 'o')
insert into #table3 values (7, '2012-01-05', 'o')
insert into #table3 values (9, '2012-01-06', 'p')
insert into #table3 values (7, '2012-01-07', 'p')

SELECT #table1.InvoiceId, #table1.InvoiceNumber, t.o as [AcceptDate1], t.p as [AcceptDate2], #table2.CompanyName
FROM #table3 pivot (max(AcceptDate) for AcceptType IN ("o", "p")) as t
JOIN #table1 on #table1.InvoiceId = t.InvoiceId
JOIN #table2 on #table2.CompanyId = #table1.CompanyId

ROLLBACK

这给出了：

InvoiceId   InvoiceNumber AcceptDate1             AcceptDate2             CompanyName
----------- ------------- ----------------------- ----------------------- --------------------------------------------------
3           1             2012-01-01 00:00:00.000 2012-01-02 00:00:00.000 Name2
4           2             2012-01-04 00:00:00.000 NULL                    Name2
6           1             NULL                    2012-01-03 00:00:00.000 Name1
7           3             2012-01-05 00:00:00.000 2012-01-07 00:00:00.000 Name2
9           2             NULL                    2012-01-06 00:00:00.000 Name1

(5 row(s) affected)

对表行进行分组并将结果连接到另一个表

1 个答案: