Question

I have a fairly easy SQL task at hand and I’d like validation (or guidance) for the solution I came up with. Thank you for helping! (this is my first post)

Here’s the problem I’m facing (simplified):

I’m importing user information from a flat file into a staging table (using SSIS). Each user will have 2 or 3 records. Each line will contain important data. The end result needs to be 1 record per customer that contains info from all 3.

Here’s an example of the data:

PK |    Name   | UniqueCustID | Info1 | Info2  | Info3 |
----------------------
1   | John Doe  |    12345     |  Opt1  |  NULL |  NULL
2   | John Doe  |    12345     |  NULL  |  Opt2 |  NULL
3   | John Doe  |    12345     |  NULL  |  NULL |  Opt3

The final result needs be be something like this:

 PK |    Name   | UniqueCustID | Info1 | Info2  | Info3 |
----------------------
1   | John Doe  |    12345     |  Opt1  |  Opt2 |  Opt3

I’m trying to keep this as simple as possible. I want to handle this with a basic Execute SQL task in SSIS (or a couple). What are people’s jump-to reactions? Can I provide any additional information? Thank you again.

UPDATE – To show the two step process I am trying:

1) This should make all of the individual customer sets match:

WITH CustInfoTMP (UniqueCustID,Info1,Info2,Info3)
AS
    (
    SELECT UniqueCustID,MAX(Info1),MAX(Info2),MAX(Info3)
    FROM CustStaging
    GROUP BY UniqueCustID 
    )
UPDATE CustStaging 
SET 
    CustStaging.Info1 = CustInfoTMP.Info1
    CustStaging.Info2 = CustInfoTMP.Info2
    CustStaging.Info3 = CustInfoTMP.Info3
FROM CustStaging
INNER JOIN CustStagingTMP ON CustStaging.UniqueCustID = CustStagingTMP.UniqueCustID

2) I then using this to delete duplicate records:

DELETE
FROM CustStaging 
WHERE UniqueCustID NOT IN
    (
    SELECT MAX(PK) 
    FROM CustStaging 
    GROUP BY UniqueCustID
    )

I hope everyone is following this. I really appreciate the feedback.

Answer 1

How about this?

CREATE TABLE TargetTable_tbl
( ID INT IDENTITY(1,1) ,
  Name  varchar(100),
  Info1 varchar(10),
  Info2 varchar(10),
  Info3 varchar(10)
)
go

insert TargetTable_tbl 

select distinct Name, Info1, Info2, Info3
from StagingTable_tbl ST
left join 
( select distinct Name, Info1
  from StagingTable_tbl 
  where Info1 is not null ) I1 on ST.Name = I1.Name
left join 
( select distinct Name, Info2
  from StagingTable_tbl 
  where Info2 is not null ) I2 on ST.Name = I2.Name
left join 
( select distinct Name, Info3
  from StagingTable_tbl 
  where Info3 is not null ) I3 on ST.Name = I3.Name
go