Question

基本上我需要从表中取3列，然后使用第一列的值将它们组合成字符串行。

使用以下方式启动基本输出：

    select ItemID, Month, Year
      from tableA 
        where ID = @id

输出：

ItemID Month Year

 1. 4 1 2012
 2. 4 2 2012
 3. 4 3 2012
 4. 4 6 2012
 5. 4 8 2012
 6. 12 1 2012
 7. 12 2 2012
 8. 12 4 2012
 9. 12 5 2012
 10. 12 6 2012
 11. 12 7 2012
 12. 53 8 2012
 13. 53 9 2012

我想要获得的输出应该看起来与此类似：

 1. 41201222012320126201282012
 2. 12120122201242012520126201272012
 3. 538201292012

基本上，将所有相同数量的ItemID组合在一起，并将月份和年份添加到该行的末尾，直到不再有，然后在有新的ItemID时开始换行。

我尝试了各种临时表和数据透视表但无法设法正确输出。任何帮助将不胜感激。

由于

所以，使用这个：

    declare @table table (idmy varchar(8))

insert into @table
    select cast(ItemID as varchar(2)) + 
        cast(Month as varchar(2)) + 
        cast(Year as varchar(4)) as idmy
        from TableA
            where ID = @id

select idmy from @table

我可以获得输出：

idmy
 1. 412012
 2. 422012
 3. 432012
 4. 462012
 5. 482012
 6. 1212012
 7. 1222012
 8. 1242012
 9. 1252012
 10. 1262012
 11. 1272012
 12. 5382012
 13. 5392012

现在在上面做的时候将所有类似的初始行组合起来......

Answer 1

select 
coalesce(cast(itemid as nvarchar(4)),'') +
coalesce(cast([1] as nvarchar(500)),'') +
coalesce(cast([2] as nvarchar(500)),'') +
coalesce(cast([3] as nvarchar(500)),'') +
coalesce(cast([4] as nvarchar(500)),'') +
coalesce(cast([5] as nvarchar(500)),'') +
coalesce(cast([6] as nvarchar(500)),'') +
coalesce(cast([7] as nvarchar(500)),'') data
from 
(select itemid, 
    cast(Month as varchar(2)) + 
    cast(Year as varchar(4)) as idmy,
    row_number() over (partition by itemid order by cast(ItemID as varchar(2)) + cast(Month as varchar(2)) + cast(Year as varchar(4))) rownum
from test
) f
pivot (max(idmy) for rownum in ([1], [2], [3], [4], [5], [6], [7])) p

我使用了row_number窗口函数来获取每一行的顺序，然后将每一行转移到它自己的列，然后转换为varchar，然后按itemid分组后连接到varchar

Answer 2

create table #tableA
(ItemID int, Month int, Year int)

insert into #tableA
values (4, 1, 2012)
insert into #tableA
values (4, 2, 2012)
insert into #tableA
values (4, 2, 2012)
insert into #tableA
values (4, 3, 2012)
insert into #tableA
values (4, 4, 2012)
insert into #tableA
values (4, 6, 2012)
insert into #tableA
values (4, 8, 2012)
insert into #tableA
values (12, 1, 2012)
insert into #tableA
values (12, 2, 2012)
insert into #tableA
values (12, 4, 2012)
insert into #tableA
values (12, 6, 2012)
insert into #tableA
values (12, 7, 2012)
insert into #tableA
values (53, 8, 2012)
insert into #tableA
values (53, 9, 2012)



declare @results as table
(line varchar(max))

declare @currentItemId int
set @currentItemId = -1

declare @str varchar(max)

while(@currentItemId is not null)
begin

    set @str = ''

    select @currentItemId=min(ItemID)
    from #tableA
    where @currentItemId < ItemID

    if(@currentItemId is null)
        break

    select @str=@str+cast(Month as varchar)+cast(Year as varchar)
    from #tableA
    where ItemID = @currentItemId
    order by Year, Month

    insert into @results
    select cast(@currentItemId as varchar)+@str

end

select *
from @results

创建：

412012220122201232012420126201282012
121201222012420126201272012
538201292012

将表的多个列组合为单个列，其中第一列值确定新行

2 个答案: