SELECTt.COUNTRY_CODE,t.COMPANY_CODE,t.ITEM_NBR,t.Department,t.[Week],d.[Day],q.QTY
FROM t
CROSS JOIN (
VALUES ('SUN'),('MON'),('TUE'),('WED'),('THU'),('FRI'),('SAT')
) d ([Day])
CROSS APPLY (
SELECT CASE d.[Day]
WHEN 'SUN' THEN t.SUN_QTY
WHEN 'MON' THEN t.MON_QTY
WHEN 'TUE' THEN t.TUE_QTY
WHEN 'WED' THEN t.WED_QTY
WHEN 'THU' THEN...

Don't you mean login as the default language of the login dictates date conversion as described by Thom and could be different...

Phil's solution is what I would have done.

Just for completeness

SELECTCOUNTRY_CODE,COMPANY_CODE,ITEM_NBR,Department,[Week],[Day],QTY
FROM(
SELECTCOUNTRY_CODE,COMPANY_CODE,ITEM_NBR,Department,[Week],
SUN_QTY,MON_QTY,TUE_QTY,WED_QTY,THU_QTY,FRI_QTY,SAT_QTY
FROM
) p (COUNTRY_CODE,COMPANY_CODE,ITEM_NBR,Department,[Week],
[SUN],[MON],[TUE],[WED],[THU],[FRI],[SAT])
UNPIVOT(
[QTY] FOR [Day] IN ([SUN],[MON],[TUE],[WED],[THU],[FRI],[SAT])
) a;

SELECTt.COUNTRY_CODE,t.COMPANY_CODE,t.ITEM_NBR,t.Department,t.[Week],d.[Day],q.QTY
FROM t
CROSS JOIN (
VALUES ('SUN'),('MON'),('TUE'),('WED'),('THU'),('FRI'),('SAT')
) d ([Day])
CROSS APPLY...

I could not see any attachment.

However two samples that may help

SELECTID, DayNo, DayValue
FROM(
SELECTID, Val1, Val2, Val3, Val4, Val5, Val6, Val7
FROM
) p (ID, [1], [2], [3], [4], [5],...

Great explanation Jeff (as always) 🙂

Actually, the best thing to do would be to have an "Order_Header" table where the ORDER TYPE only needs to be noted once and then store the details in...

Just make sure the number of seconds does not exceed one day otherwise the time will be wrong

Don't know Oracle but if the result of your math is number of seconds since midnight then

CAST(DATEADD(second,(626400-432000)/10,0) as time)

I'm assuming by now...

I also have AK's, which are also unique and are what you call "natural keys".  For me, they exist mostly to enforce the very uniqueness you speak of. ...

IMHO, you are on the right path.  I'm currently doing exactly what you described.  Importing flat, delimited files to a 'staging' table where all fields are defined...

USE LAG instead of SAMPLING_POINT in the select like this

CASE WHEN SAMPLE.SAMPLING_POINT = LAG(SAMPLE.SAMPLING_POINT,1,'') OVER (ORDER BY SAMPLE.SAMPLING_POINT ASC)
THEN ''
ELSE SAMPLE.SAMPLING_POINT
END AS [SAMPLING_POINT]

You would if you only wanted to affect a subset of TRANSDETAIL

Otherwise the...

You need to produce a cartesion of unique employees (SELECT DISTINCT) and DateRange (CROSS JOIN) and LEFT JOIN the employees table

Actually...