SQL - Join two tables without unique fields - sql

Question

This is my first post in this forum so please be understanding.
I have following issue.
I want to two join two tables:
Table1:
Product | Start Date | End Date
-------------------------------------
Product1 | 01/01/2014 | 01/05/2015
Product2 | 01/03/2014 | 01/01/2015
Table2:
Product | Start Date | End Date | Value
--------------------------------------------
Product1 | 01/01/2014 | 01/02/2015 | 10
Product1 | 02/02/2014 | 01/04/2015 | 15
Product1 | 02/04/2014 | 01/05/2015 | 15
Product2 | 01/03/2014 | 04/05/2014 | 5
Product2 | 05/05/2014 | 01/01/2015 | 5
To have a table with latest value like:
Product | Start Date | End Date | Value
------------------------------------------------
Product1 | 02/04/2014 | 01/05/2015 | 15
Product2 | 05/05/2014 | 01/01/2015 | 5
I need to join them and not use just the second table because both of them have more unique columns that I need to use.
I was thinking about firstly using some kind of IF function on second table to make one row per product (the one with latest start date) and than join it simply then with first table. But I have no idea how to do the first part.
I am really looking forward for your help.
Regards,
Matt

Answer 1

Just use WHERE NOT EXISTS to filter out everything but the latest date from TABLE2 (I am assuming that you are asking for the latest STARTDATE from TABLE2; also I add 'SomeOtherField' to Table1, because otherwise you could just query Table2):
SELECT t1.Product, t1.SomeOtherField, t2.StartDate, t2.EndDate, t2.Value
FROM Table1 t1
JOIN (SELECT a.Product, a.StartDate, a.EndDate, a.Value FROM Table2 a
WHERE NOT EXISTS (SELECT * FROM Table2 b
WHERE b.Product = a.Product AND b.StartDate > a.StartDate)) t2
ON (t2.Product = t1.Product)

Answer 2

This is possible, the query will involve three steps:
Find all the max start date for each product in table 2. Hint: use group by.
Join table 2 with the result from #1 to get the Value.
Join table 1 with the result from #2 to filter out products that are not in table 1.

Answer 3

Not sure you need Table1 at all in your example, you just need to aggregate Table2 to find the MAX([Start Date] for each Product:
SELECT a.*
FROM Table2 a
JOIN (SELECT Product,MAX([Start Date]) AS Mx_Start_Dt
FROM Table2
GROUP BY Product
) b
ON a.Product = b.Product
AND a.[Start Date] = b.Mx_Start_Dt
If you do need to bring in fields from Table you can just add another JOIN:
SELECT a.*,b.*
FROM Table1 a
JOIN (SELECT a.*
FROM Table2 a
JOIN (SELECT Product,MAX([Start Date]) AS Mx_Start_Dt
FROM Table2
GROUP BY Product
) b
ON a.Product = b.Product
AND a.[Start Date] = b.Mx_Start_Dt
) c
ON a.Product = b.Product
If using a database that supports analytic functions, you can make it cleaner via the ROW_NUMBER() function:
;with cte AS (SELECT *,ROW_NUMBER() OVER(PARTITION BY Product ORDER BY [Start Date] DESC) AS RN
FROM Table2
)
SELECT *
FROM Table1 a
JOIN cte b
ON a.Product = b.Product
AND b.RN = 1

Answer 4

Here is a solution using ROW_NUMBER in SQLServer:
SELECT *
FROM (
SELECT *, ROW_NUMBER() OVER(PARTITION BY Product ORDER BY StartDate DESC) RN
FROM #T
) Results
WHERE RN = 1