Question

How can I efficiently update the table based on values from a join table
only when ID – identifier I use to join both of the tables matches perfectly
1 to 1. I mean when joined table has only one ID to the updated table?

DECLARE @T1 TABLE (
  ID INT,
  NAME VARCHAR(10),
  Age int
)
INSERT INTO @T1 VALUES (1, 'Name', null)

DECLARE @T2 TABLE (
  ID INT,  
  Age int
)
INSERT INTO @T2 VALUES (1, 28)
INSERT INTO @T2 VALUES (1, 29)
INSERT INTO @T2 VALUES (1, 30)

In this example table T2 has three records of the ID = 1 which corresponds to one ID
in T1.

And I would like to update T1 only when in T2 there is one record of ID = 1.

(I would like to avoid joining twice table t2 to solve this task …)

Thanks!

Answer 1

;WITH T2
     AS (SELECT ID,
                MAX(Age) AS Age
         FROM   @T2
         GROUP  BY ID
         HAVING COUNT(*) = 1)
UPDATE @T1
SET    [@T1].Age = T2.Age
FROM   @T1
       JOIN T2
         ON [@T1].ID = T2.ID