Question

我有这种查询类型，

select StateName ,countryName
from country_master left join state_master on country_master.countryID = state_master.countryID

产生以下结果，

古吉拉特邦印度
德里印度
  印度人力资源印度委员会   NY USA
LA USA
  芝加哥美国
WDC USA
伦敦英国

我想将此结果放在单行中，其关系如

印度 - 古吉拉特邦德里HR MP
美国 - 纽约洛杉矶芝加哥 - 伦敦

有任何构建此查询的想法
建议我，谢谢，

Answer 1

此问题基本上与Concatenate many rows into a single text string?重复。

以下是一种适用于Sql Server 2008的方法：

WITH Ranked ( countryID, rnk, stateName )  
         AS ( SELECT countryID,
                     ROW_NUMBER() OVER( PARTITION BY countryID ORDER BY countryID ),
                     CAST( stateName AS VARCHAR(8000) ) 
                FROM state_master),
AnchorRanked ( countryID, rnk, stateName )  
         AS ( SELECT countryID, rnk, stateName 
                FROM Ranked
               WHERE rnk = 1 ),
RecurRanked ( countryID, rnk, stateName ) 
         AS ( SELECT countryID, rnk, stateName 
                FROM AnchorRanked
               UNION ALL 
              SELECT Ranked.countryID, Ranked.rnk,
                     RecurRanked.stateName + ', ' + Ranked.stateName
                FROM Ranked
               INNER JOIN RecurRanked 
                  ON Ranked.countryID = RecurRanked.countryID 
                 AND Ranked.rnk = RecurRanked.rnk + 1 )
SELECT countryName, MAX( stateName ) 
  FROM RecurRanked
  INNER JOIN country_master on RecurRanked.countryID = country_master.countryID
 GROUP BY countryName;

您可以在此处看到几种提取此类技术的方法：https://www.simple-talk.com/sql/t-sql-programming/concatenating-row-values-in-transact-sql/

如果你的连续城市列表长度超过8000个字符，你需要小心，尽管如果这是一个问题，你可能能够做出类似varchar（max）之类的事情（除了事实之外）它可能不是查看数据的好方法。

Answer 2

使用递归合并函数可以非常轻松地完成此任务。

您创建一个连接递归查询结果的函数，然后在查询中调用该函数。

我对这个问题的回答中的完整代码和示例： How to "Implode" (de-normalize/concat) multiple columns into a single column?

Answer 3

Select t.countryName , isnull(Stuff((SELECT ', ' + StateName
          FROM state_master where countryID = t.countryID
          ORDER BY StateName
          For XML PATH ('')),1,1,''),'') as c
from country_master  t

Answer 4

select c.countryName,
       stuff((select ' '+s.StateName
              from state_master as s
              where s.countryID = c.countryID
              for xml path(''), type).value('.', 'nvarchar(max)'), 1, 1, '')
from country_master as c

显示结果单行数据，无论是否。 Sql server中的列

4 个答案: