Question

Very new SQL Server, am using SQL server 2008. I have two tables, Table A and Table B

I want to update Table A with the count of a matching strings of Table B. Here is what I came up by declaring a static varchar. I would like it to do it in a procedure for all Records in Table A anytime a record is inserted/updated in Table B.

TABLE A: **PO**, Count, Closed         Table B: **LD**
            24A, 0,     0,                      24A-1
            25A, 0,     0,                      24A-2
            26A, 0,     0,                      25A-3
                                                26A-1
                                                26A-2

Code I tried:

 Declare @POTableA AS VARCHAR(15)
 SET @POTableA = '24A'
   Update TABLE A
   SET TableA.Count =(Select Count(*) AS 'Count_LD' FROM TABLE B 
   WHERE TableB.LD LIKE '%'+@POTableA+'-%') 
   FROM TABLE B WHERE TABLEA.PO LIKE '%'+@POTableA+'%'

Current result:

TABLE A: **PO**, Count, Closed         Table B: **LD**
            24A, 2,     0,                      24A-1
            25A, 0,     0,                      24A-2
            26A, 0,     0,                      25A-3
                                                26A-1
                                                26A-2

Desired result:

TABLE A: **PO**, Count, Closed         Table B: **LD**
            24A, 2,     0,                      24A-1
            25A, 1,     0,                      24A-2
            26A, 2,     0,                      25A-3
                                                26A-1
                                                26A-2

Answer 1

While I’m not sure why you need to keep the count stored in table A – you can always determine this count at runtime, then you don’t have to use triggers etc. to maintain that information (which is redundant). But here is a demonstration using table variables about how such an update with an aggregate could be achieved (there are other ways, likely).

DECLARE @A TABLE(PO VARCHAR(32), [count] INT);

DECLARE @B TABLE(LD VARCHAR(32));

INSERT @A VALUES
('24A',0), ('25A',0), ('26A',0);

INSERT @B VALUES
('24A-1'), ('24A-2'), ('25A-3'),
('26A-1'), ('26A-2');

UPDATE T
SET T.[count] = S.[count]
FROM @A AS T
INNER JOIN
(
 SELECT A.PO, [count] = COUNT(B.LD)
   FROM @A AS A INNER JOIN @B AS B
   ON B.LD LIKE A.PO + '%'
   GROUP BY A.PO
) AS S
ON T.PO = S.PO;

SELECT PO, [count] FROM @A;

Results:

PO     count
----   -----
24A    2
25A    1
26A    2

Now if you need to do this in a trigger, it would be something like:

CREATE TRIGGER dbo.MaintainRedundantCount
ON dbo.TableB
FOR INSERT, UPDATE
AS
BEGIN
  SET NOCOUNT ON;

  UPDATE T
   SET T.[count] = S.[count]
   FROM dbo.TableA AS T
   INNER JOIN
   (
     SELECT A.PO, [count] = COUNT(B.LD)
       FROM dbo.TableA AS A 
       INNER JOIN dbo.TableB AS B
       ON B.LD LIKE A.PO + '%'
       GROUP BY A.PO
   ) AS S
   ON T.PO = S.PO;
END
GO

You might want to modify one of those lines to:

       ON B.LD LIKE A.PO + '-%'

But since you only gave three sample values in the question, and they were all the exact same format, it’s tough to tell what your true data set looks like.