I have a table table_One
which has Multiple columns(Column1, Column2, Column3, Column4.....)
both references (Contains PK values for Table_two
) to another table Table_Two
. Is there any efficient way getting joining these two tables rather than joining Table_one back to table_Two Multiple Times.br/>
The Structure of the Two tables and the Desired Result Set is as follows.
Table_One
Following solution (SQLFiddle) reads the rows from the second table just one time:
SET STATISTICS IO ON;
...
PRINT 'Test #1'
SELECT *
FROM
(
SELECT ca.PrimaryKey, ca.[Type], y.ColumnA
FROM @Table1 x
UNPIVOT( Value FOR [Type] IN ([Column1], [Column2]) ) ca
INNER MERGE /*HASH*/ JOIN @Table2 y ON ca.Value = y.ID
) src
PIVOT( MAX(src.ColumnA) FOR src.[Type] IN ([Column1], [Column2]) ) pvt
PRINT 'End of Test #1'
Results:
Test #1
PrimaryKey Column1 Column2
---------- --------- -------
1 ALPHA CHARLIE
2 BETA DELTA
3 CHARLIE ALPHA
4 DELTA CHARLIE
5 ALPHA DELTA
6 CHARLIE ALPHA
7 ALPHA DELTA
8 DELTA CHARLIE
Table 'Worktable'. Scan count 0, logical reads 0, physical reads 0, read-ahead reads 0, lob logical reads 0, lob physical reads 0, lob read-ahead reads 0.
Table '#65B6F546'. Scan count 1, logical reads 2, physical reads 0, read-ahead reads 0, lob logical reads 0, lob physical reads 0, lob read-ahead reads 0.
Table '#61E66462'. Scan count 1, logical reads 2, physical reads 0, read-ahead reads 0, lob logical reads 0, lob physical reads 0, lob read-ahead reads 0.
End of Test #1
Have you tried aliasing your table2 table and joining it twice to Table_One like below?
SELECT
t1.PrimaryKey,
c1.ColumnA AS Column1,
c2.ColumnA AS Column2
FROM Table_One t1
JOIN Table_two c1 ON t1.Column1 = c1.ID
JOIN Table_two c2 ON t1.Column2 = c2.ID;
Select Table_One.PrimaryKey,
T2_Column1.ColumnA As Column1,
T2_Column2.ColumnA As Column2
From Table_One
Inner Join Table_Two As T2_Column1
On Table_One.Column1 = T2_Column1.ID
Inner Join Table_Two As T2_Column2
On Table_One.Column2 = T2_Column2.Id
Basically, you join to table 2 twice. When you do this, you MUST alias at least one of them so that SQL doesn't get confused. As a matter of practice, it's usually best to alias both of them so that when you read this code again in 6 months, it will be easier to understand.