SQL 基于公共行值合并

Question

我正在尝试从一个服务台软件迁移到另一个，其中一部分是迁移我们的工单历史数据库，我将其导出到 excel 电子表格，然后导入到 SQL服务器数据库。

目前table的格式如下：

第 headers 列：

Description: Initial ticket text  
ticketnum: Ticket number  
subject: Ticket subject header  
value: ticket comment

以防万一：

Description: nvarchar(max)  
ticketnum: float  
subject: nvarchar(max)  
value: nvarchar(max)

编辑：要求的示例数据：

Description            TicketNum     Subject                       Value
-------------------------------------------------------------------------
Our computer exploded   10020        CPU Explosion        Our computer exploded
Our computer exploded   10020        CPU Explosion        The computer is a dell
Virus Found             10021        Virus                We need help with a virus

我希望它变成：

Description            TicketNum     Subject              Value
-------------------------------------------------------------------------
Our computer exploded   10020        CPU Explosion        Our computer exploded; The computer is a dell
Virus Found             10021        Virus                We need help with a virus

每一行都是工单的新条目，因此值字段始终是唯一的，而每个工单号的描述和值保持不变。因此，我希望根据票号字段合并行，keep/merge 值字段的所有数据，并且只保留主题和描述字段的一个值。我已经看到了很多接近我的场景的问题，但是 none 匹配，遗憾的是我对 SQL 知之甚少，并且不足以根据我所看到的进行推断来解决这个问题独唱

在此先感谢您。

编辑 2：修改后的代码，运行但不合并，生成一个空 table:

declare @tmpTable table ([Description] nvarchar(max), niceid float, 
[Subject] nvarchar(max), [Value] nvarchar(max))

SELECT * INTO tmpTable FROM sheetnew$

declare @newTable table ([Description] nvarchar(max), niceid float, 
[Subject] nvarchar(max), [Value] nvarchar(max))


insert into @newTable
select distinct 
    x.[Description]
    ,x.niceID
    ,x.[Subject]
   ,[Value] = stuff((
          select '; ' + y.[Value]
          from @tmptable y
          where x.niceid = y.niceid
          for xml path(''), type).value('.', 'NVARCHAR(MAX)'), 1, 1, '')
from @tmptable x

Answer 1

这是使用 STUFF

的方法

SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
SET ANSI_WARNINGS ON
GO
SET ANSI_PADDING ON
GO

declare @table table ([Description] varchar(64), TicketNum int, [Subject] varchar(64), [Value] varchar(4000))
insert into @table
values
('Our computer exploded',10020,'CPU Explosion','Our computer exploded'),
('Our computer exploded',10020,'CPU Explosion','The computer is a dell'),
('Virus Found',10021,'Virus','We need help with a virus')





declare @newTable table ([Description] varchar(64), TicketNum int, [Subject] varchar(64), [Value] varchar(4000))
insert into @newTable
select distinct 
    x.[Description]
    ,x.TicketNum
    ,x.[Subject]
    ,[Value] = stuff((
          select '; ' + y.[Value]
          from @table y
          where x.TicketNum = y.TicketNum
          for xml path(''), type).value('.', 'NVARCHAR(MAX)'), 1, 1, '')
from @table x


select * from @newTable

SQL 基于公共行值合并

SQL combine based on common row value

sql-server

merge

key

sql-server-2017