背景：在做项目时，经常会遇到这样的表结构在主表的中有一列保存的是用逗号隔开ID。如，当一个员工从属多个部门时、当一个项目从属多个城市时、当一个设备从属多个项目时，很多人都会在员工表中加入一个deptIds VARCHAR(1000)列（本文以员工从属多个部门为例），用以保存部门编号列表（很明显这不符合第一范式，但很多人这样设计了，在这篇文章中我们暂不讨论在这种应用场景下，如此设计的对与错，有兴趣的可以在回复中聊聊），然后我们在查询列表中需要看到这个员工从属哪些部门。
初始化数据：
部门表、员工表数据：
复制代码 代码如下:

IF EXISTS (SELECT * FROM sys.objects WHERE object_id = OBJECT_ID(N'[dbo].[Department]'))
DROP TABLE [dbo].Department
GO
--部门表
CREATE TABLE Department
(
id int,
name nvarchar(50)
)
INSERT INTO Department(id,name)
SELECT 1,'人事部'
UNION
SELECT 2,'工程部'
UNION
SELECT 3,'管理部'
SELECT * FROM Department

IF EXISTS (SELECT * FROM sys.objects WHERE object_id = OBJECT_ID(N'[dbo].[Employee]'))
DROP TABLE [dbo].Employee
GO
--员工表
CREATE TABLE Employee
(
id int,
name nvarchar(20),
deptIds varchar(1000)
)
INSERT INTO Employee(id,name,deptIds)
SELECT 1,'蒋大华','1,2,3'
UNION
SELECT 2,'小明','1'
UNION
SELECT 3,'小华',''
SELECT * FROM Employee

希望得到的结果：

解决方法：

第一步，是得到如下的数据。即将员工表集合与相关的部门集合做交叉连接，其中使用了fun_SplitIds函数（作用是将ids分割成id列表），然后员工集合与这个得到的集合做交叉连接
复制代码 代码如下:
SELECT E.*,ISNULL(D.name,'') AS deptName
FROM Employee AS E
OUTER APPLY dbo.fun_SplitIds(E.deptIds) AS DID
LEFT JOIN Department AS D ON DID.ID=D.id;

第二步，已经得到了如上的数据，然后要做的就是根据ID分组，并对deptName列做聚合操作，但可惜的是SQL SERVER还没有提供对字符串做聚合的操作。但想到，我们处理树形结构数据时，用CTE来做关系数据，做成有树形格式的数据，如此我们也可以将这个问题转换成做树形格式的问题，代码如下：
复制代码 代码如下:
;WITH EmployeT AS(
--员工的基本信息(使用OUTER APPLY将多个ID拆分开来,然后与部门表相关联)
--此时已将员工表所存的IDS分别与部门相关联,下面需要将此集合中的deptName聚合成一个记录
SELECT E.*,ISNULL(D.name,'') AS deptName
FROM Employee AS E
OUTER APPLY dbo.fun_SplitIds(E.deptIds) AS DID
LEFT JOIN Department AS D ON DID.ID=D.id
),mike AS(
SELECT id,name,deptIds,deptName
,ROW_NUMBER()OVER(PARTITION BY id ORDER BY id) AS level_num
FROM EmployeT
),mike2 AS(
SELECT id,name,deptIds,CAST(deptName AS NVARCHAR(100)) AS deptName,level_num
FROM mike
WHERE level_num=1
UNION ALL
SELECT m.id,m.name,m.deptIds,CAST(m2.deptName+','+m.deptName AS NVARCHAR(100)) AS deptName,m.level_num
FROM mike AS m
INNER JOIN mike2 AS m2 ON m.ID=m2.id AND m.level_num=m2.level_num+1
),maxMikeByIDT AS(
SELECT id,MAX(level_num) AS level_num
FROM mike2
GROUP BY ID
)

SELECT A.id,A.name,A.deptIds,A.deptName
FROM mike2 AS A
INNER JOIN maxMikeByIDT AS B ON A.id=B.ID AND A.level_num=B.level_num
ORDER BY A.id OPTION (MAXRECURSION 0)

结果如下：

全部SQL：
复制代码 代码如下:
IF EXISTS (SELECT * FROM sys.objects WHERE object_id = OBJECT_ID(N'[dbo].[Department]'))
DROP TABLE [dbo].Department
GO
--部门表
CREATE TABLE Department
(
id int,
name nvarchar(50)
)
INSERT INTO Department(id,name)
SELECT 1,'人事部'
UNION
SELECT 2,'工程部'
UNION
SELECT 3,'管理部'

SELECT * FROM Department


IF EXISTS (SELECT * FROM sys.objects WHERE object_id = OBJECT_ID(N'[dbo].[Employee]'))
DROP TABLE [dbo].Employee
GO
--员工表
CREATE TABLE Employee
(
id int,
name nvarchar(20),
deptIds varchar(1000)
)
INSERT INTO Employee(id,name,deptIds)
SELECT 1,'蒋大华','1,2,3'
UNION
SELECT 2,'小明','1'
UNION
SELECT 3,'小华',''

SELECT * FROM Employee

--创建一个表值函数，用来拆分用逗号分割的数字串，返回只有一列数字的表
IF EXISTS (SELECT * FROM sys.objects WHERE object_id = OBJECT_ID(N'[dbo].[fun_SplitIds]'))
DROP FUNCTION [dbo].fun_SplitIds
GO
CREATE FUNCTION dbo.fun_SplitIds(
@Ids nvarchar(1000)
)
RETURNS @t_id TABLE (id VARCHAR(36))
AS
BEGIN
DECLARE @i INT,@j INT,@l INT,@v VARCHAR(36);
SET @i = 0;
SET @j = 0;
SET @l = len(@Ids);
while(@j < @l)
begin
SET @j = charindex(',',@Ids,@i+1);
IF(@j = 0) set @j = @l+1;
SET @v = cast(SUBSTRING(@Ids,@i+1,@j-@i-1) as VARCHAR(36));
INSERT INTO @t_id VALUES(@v)
SET @i = @j;
END
RETURN;
END
GO


;WITH EmployeT AS(
--员工的基本信息(使用OUTER APPLY将多个ID拆分开来,然后与部门表相关联)
--此时已将员工表所存的IDS分别与部门相关联,下面需要将此集合中的deptName聚合成一个记录
SELECT E.*,ISNULL(D.name,'') AS deptName
FROM Employee AS E
OUTER APPLY dbo.fun_SplitIds(E.deptIds) AS DID
LEFT JOIN Department AS D ON DID.ID=D.id
),mike AS(
SELECT id,name,deptIds,deptName
,ROW_NUMBER()OVER(PARTITION BY id ORDER BY id) AS level_num
FROM EmployeT
),mike2 AS(
SELECT id,name,deptIds,CAST(deptName AS NVARCHAR(100)) AS deptName,level_num
FROM mike
WHERE level_num=1
UNION ALL
SELECT m.id,m.name,m.deptIds,CAST(m2.deptName+','+m.deptName AS NVARCHAR(100)) AS deptName,m.level_num
FROM mike AS m
INNER JOIN mike2 AS m2 ON m.ID=m2.id AND m.level_num=m2.level_num+1
),maxMikeByIDT AS(
SELECT id,MAX(level_num) AS level_num
FROM mike2
GROUP BY ID
)

SELECT A.id,A.name,A.deptIds,A.deptName
FROM mike2 AS A
INNER JOIN maxMikeByIDT AS B ON A.id=B.ID AND A.level_num=B.level_num
ORDER BY A.id OPTION (MAXRECURSION 0)