INSERT or UPDATE table from another table with composite primary key

Question

I am looking for the correct syntax and way to do the following directly from SQL: insert or update (if data already exists inside) TableMain from data contained in TableA with both having same composite primary key.

Both tables are defined as :

CREATE TABLE TableA (
[TID0] [int] NOT NULL,
[TID1] [int] NOT NULL,
[language] [nvarchar](2) NOT NULL,
[TID2] [nvarchar](200) NOT NULL,
[text] [nvarchar](max) NULL,
[updatedOn] [datetime] NOT NULL  DEFAULT (getdate())
PRIMARY KEY (
    [TID0],
    [TID1],
    [language],
    [TID2],
)
) ON [PRIMARY] TEXTIMAGE_ON [PRIMARY]

TableA will be periodically deleted and filled.

TableMain as the same definition but will contain many more rows of data and what I need is to insert never seen values from TableA into TableMain, and update already existing rows.

I used to do this kind of insert but I do not know how to handle update and composite primary keys :

INSERT INTO TableMain 
    SELECT * FROM TableA

EDIT : i am using SQL Server 9.00.5000

EDIT : another way inspired by MERGE and mimick it

DECLARE @updatedIDs TABLE(
    [TID0] [int],
    [TID1] [int],
    [language] [nvarchar](2),
    [TID2] [nvarchar](200),
                PRIMARY KEY ([TID0], [TID1], [language], [TID2])  -- as stated by Nikola Markovinović above, thanks
);

-- First update records 
update TableMain 
   set [text]      = source.[text], 
       [updatedOn] = source.[updatedOn] 
       OUTPUT
         inserted.[TID0]
         inserted.[TID1]
         inserted.[language]
         inserted.[TID2]
       INTO @updatedIDs
  from 
       TableMain AS main
       , TableA AS source 
  WHERE    
   TableMain.[TID0]         = source.[TID0] 
   and TableMain.[TID1]     = source.[TID1] 
   and TableMain.[language] = source.[language] 
   and TableMain.[TID2]     = source.[TID2] 

-- And then insert 

insert into TableMain
select *
  from TableA AS source 
 where not exists 
       ( 
      select 1
        from @updatedIDs AS i
       where i.[TID0]     = source.[TID0] 
         and i.[TID1]     = source.[TID1] 
         and i.[language] = source.[language] 
         and i.[TID2]     = source.[TID2] 
       )

Answer 1

you should use a merge statment

something like this:

merge TableMain  AS target
using TableA as source 
ON <join tables here>
WHEN MATCHED THEN <update>
WHEN NOT MATCHED BY TARGET <Insert>
WHEN NOT MATCHED BY SOURCE <delete>

Answer 2

Here is a script you might use to upsert your data:

-- On error transaction is automatically rolled back
set xact_abort on

begin transaction
-- First update records
update TableMain
   set [text]      = source.[text],
       [updatedOn] = source.[updatedOn]
  from TableMain
 inner join TableA source
    on TableMain.[TID0]     = source.[TID0]
   and TableMain.[TID1]     = source.[TID1]
   and TableMain.[language] = source.[language]
   and TableMain.[TID2]     = source.[TID2]

-- And then insert

insert into TableMain ([TID0], [TID1], [language], [TID2], [text], [updatedOn])
select [TID0], [TID1], [language], [TID2], [text], [updatedOn]
  from TableA source
 where not exists
       (
          select *
            from TableMain
           where TableMain.[TID0]     = source.[TID0]
             and TableMain.[TID1]     = source.[TID1]
             and TableMain.[language] = source.[language]
             and TableMain.[TID2]     = source.[TID2]
       )

commit transaction

You might rewrite not exists() as left join ... where TableMain.TID0 is null if performance is not satisfactory.