If value of 1 table exists in the 2nd table, update the 2nd table value with sum of 1st and 2nd tables, else insert 1st table data to 2nd table

Question

In SQL Server, I have tables TAB and STG as follows:

CREATE TABLE [Tab]
(
     [tab_Client] [VARCHAR](30) NULL,
     [tab_Security] [VARCHAR](30) NULL,
     [tab_Quantity] [FLOAT] NULL,
     [tab_Cost] [FLOAT] NULL
)

CREATE TABLE [Stg]
(
    [stg_client] [VARCHAR](30) NULL,
    [stg_security] [VARCHAR](30) NULL,
    [stg_Quantity] [FLOAT] NULL,
    [stg_Cost] [FLOAT] NULL
)

I need to either

1. if stg_client/stg_security doesn't exist in the Tab table, insert the stg_Client/stg_Security data from Stg table to Tab table

2. if stg_client/stg_security exist in the Tab table:

   • update the Tab table's tab_Quantity with the sum of Tab.tab_Quantity & Stg.stg_Quantity

   • update the Tab table's tab_Cost with the sum of Tab.tab_Cost & Stg.stg_Cost

How can I do that ?

TAB table

Client  Security   Quantity     Cost
-------------------------------------
JP       L1         1           100
JP       L2         2           200
JP       L3         3           300

STG table

Client     Security   Quantity   Cost
-------------------------------------
JP         L1         10         1000
JP         L3         30         3000
JP         L4         40         4000

Desired result:

TAB table

Client    Security   Quantity     Cost
-----------------------------------------
JP        L1         11           1100  -> Sum of Tab and Stg table
JP        L2         2             200
JP        L3         33           3300  -> Sum of Tab and Stg table
JP        L4         40           4000  -> From Stg table

The MERGE works

MERGE TAB AS target
USING STG AS source ON target.tab_client = source.stg_client 
                    AND target.tab_security = source.stg_security

WHEN MATCHED THEN
    UPDATE 
        SET target.tab_quantity = source.stg_quantity + target.tab_quantity,
            target.tab_cost = source.stg_cost + target.tab_cost

WHEN NOT MATCHED BY target THEN
    INSERT (tab_client, tab_security, tab_quantity, tab_cost)
    VALUES (source.stg_client, source.stg_security, source.stg_quantity, source.stg_cost);

Thank you.

Answer 1

you could use a join in updated

update  TAB 
set TAB.cost  = TAB.cost + STG.cost ,
    TAB.Quantity   = TAB.Quantity  + STG.Quantity     
from TAB  
INNER JOIN STG ON TAB.clinet = STG.client 
  AND   TAB.security = STG.security 

Answer 2

Another correct way to do this with SQL is with a LEFT OUTER JOIN for inserting the missing rows and an INNER JOIN for updating the existing rows -

Tip: Either Update matching before inserting missing or only insert the client/security fields in the insert statement to avoid updating the qty and cost of the newly inserted rows...

Here's a snippet:

UPDATE  t
SET tab_Cost = t.tab_Cost + s.stg_Cost ,
    tab_Quantity = t.tab_Quantity  + s.stg_Quantity     
FROM TAB t  
INNER JOIN STG s ON t.clinet = s.client AND t.security = s.security

INSERT INTO TAB (tab_client, tab_security, tab_quantity, tab_cost)
SELECT s.stg_Client, s.stg_Security, s.quantity, s.Cost
FROM STG s LEFT OUTER JOIN
TAB t ON t.clinet = s.client 
  AND  t.security = s.security
WHERE t.tab_client IS NULL```

(assuming tab_client is a non nullable field)