问题
I have a table like the following
Node_Id Node_Name Parent_Node_Id
------- --------- --------------
1 Root 0
2 Node1 1
3 Node2 1
4 Node3 2
5 Node4 2
6 Node5 5
7 Node6 5
8 Node7 7
Each row of the table has a parent row which is indicated by the "Parent_Node_Id" and each row/node can be parent of any number of row/nodes.
How can I write a query which will give me the list of all nodes who have a parent node who all share the same parent node? The Parent_Node_Id will be passed as the parameter. For example, if Parent_Node_Id is passed as 2, the query will return the following -
Node_Id Node_Name Parent_Node_Id
------- --------- --------------
4 Node3 2
5 Node4 2
6 Node5 5
7 Node6 5
8 Node7 7
回答1:
;WITH CTE AS (
SELECT Node_Id,Node_Name, Parent_Node_Id
FROM @a WHERE Parent_Node_Id = 2
UNION ALL
SELECT c.Node_Id,c.Node_Name, c.Parent_Node_Id
FROM @a c
INNER JOIN CTE p ON p.Node_Id = c.Parent_Node_Id
)
Select * from CTE
with example data
Declare @a Table( Node_Id int, Node_Name varchar(100), Parent_Node_Id int)
insert into @a
select 1,'Root',0
UNION SELECT 2,'Node1',1
UNION SELECT 3,'Node2',1
UNION SELECT 4,'Node3',2
UNION SELECT 5,'Node4',2
UNION SELECT 6,'Node5',5
UNION SELECT 7,'Node6',5
UNION SELECT 8,'Node7',7
来源：https://stackoverflow.com/questions/16759496/how-to-traverse-through-a-table-which-has-relation-to-itself