SQL Server 2014 Join Doubling Line Items - sql

Question

I can usually find answers to my questions if I search hard enough but I don't know how to word my question well enough to get the results I desire.
The issue I am having is that I am trying to Join my Order Charge Table to my other Query so that I can see if there was a Discount or a Shipping charge applied.
There are 3 scenarios that can happen per order in the Order Charge Table (Discount AND Shipping charge applied) OR (Discount applied) OR (Shipping applied)
NULL values are NOT allowed so if there is no discount or shipping for the order, it does not show up in this table.
My query without the order charges applied is:
SELECT
[Order].[OrderNumber],
CASE
WHEN [ShopifyOrder].[PaymentStatusCode] = '2' THEN 'Paid'
WHEN [ShopifyOrder].[PaymentStatusCode] = '4' THEN 'Refunded'
WHEN [ShopifyOrder].[PaymentStatusCode] = '5' THEN 'Voided'
WHEN [ShopifyOrder].[PaymentStatusCode] = '6' THEN 'Partially Refunded'
END AS 'PaymentStatus',
[Store].[StoreName] as 'MarketplaceNames',
[OrderItem].[SKU],
[LookupList].[MainSKU], [LookupList].[ProductName],
[LookupList].[Classification] as 'Classification',
[LookupList].[Cost],
([OrderItem].[Quantity] * [OrderItem].[UnitPrice]) AS 'Sales',
(([OrderItem].[Quantity] * [LookupList].[Quantity]) * [LookupList].[Cost]) AS 'Total Cost',
[OrderItem].[Quantity] * [LookupList].[Quantity] AS 'Total Qty'
FROM
[SHIPSERVER].[dbo].[Order]
JOIN
[SHIPSERVER].[dbo].[ShopifyOrder] ON [Order].[OrderID] = [ShopifyOrder].[OrderID]
JOIN
[SHIPSERVER].[dbo].[OrderItem] ON [OrderItem].[OrderID] = [Order].[OrderID]
JOIN
[SHIPSERVER].[dbo].[Store] ON [Order].[StoreID] = [Store].[StoreID]
LEFT JOIN
[SHIPSERVER].[dbo].[LookupList] ON [OrderItem].[SKU] = [LookupList].[SKU]
WHERE
([Store].[StoreName]= 'Shopify')
AND ([Order].[OrderDate] BETWEEN '2015-09-01 00:00:00.000' AND '2015-09-30 23:59:59.999')
AND ([Order].[IsManual] = '0')
I am going to give 1 order as an example in the Results
+-------------+---------------+------------------+---------------+---------------+----------------+-------+-------+------------+-----------+
| OrderNumber | PaymentStatus | MarketplaceNames | SKU | MainSKU | Classification | Cost | Sales | Total Cost | Total Qty |
| 7177 | Paid | Shopify | 300TLSH SL PL | 300TLSH SL PL | Sheet Set | 21.00 | 48.99 | 21 | 1 |
+-------------+---------------+------------------+---------------+---------------+----------------+-------+-------+------------+-----------+
Here is my query when I join the OrderCharge table:
SELECT
[Order].[OrderNumber],
CASE
WHEN [ShopifyOrder].[PaymentStatusCode] = '2' THEN 'Paid'
WHEN [ShopifyOrder].[PaymentStatusCode] = '4' THEN 'Refunded'
WHEN [ShopifyOrder].[PaymentStatusCode] = '5' THEN 'Voided'
WHEN [ShopifyOrder].[PaymentStatusCode] = '6' THEN 'Partially Refunded'
END AS 'PaymentStatus',
[Store].[StoreName] as 'MarketplaceNames',
[OrderItem].[SKU],
[LookupList].[MainSKU],
[OrderCharge].[Type], [OrderCharge].[Description], [OrderCharge].[Amount],
[LookupList].[Classification] as 'Classification', [LookupList].[Cost],
([OrderItem].[Quantity]* [OrderItem].[UnitPrice]) AS 'Sales',
(([OrderItem].[Quantity] * [LookupList].[Quantity]) * [LookupList].[Cost]) AS 'Total Cost',
[OrderItem].[Quantity] * [LookupList].[Quantity] AS 'Total Qty'
FROM
[SHIPSERVER].[dbo].[Order]
JOIN
[SHIPSERVER].[dbo].[ShopifyOrder] ON [Order].[OrderID] = [ShopifyOrder].[OrderID]
JOIN
[SHIPSERVER].[dbo].[OrderItem] ON [OrderItem].[OrderID] = [Order].[OrderID]
JOIN
[SHIPSERVER].[dbo].[Store] ON [Order].[StoreID] = [Store].[StoreID]
LEFT JOIN
[SHIPSERVER].[dbo].[LookupList] ON [OrderItem].[SKU] = [LookupList].[SKU]
JOIN
[SHIPSERVER].[dbo].[OrderCharge] ON [Order].[OrderID] = [OrderCharge].[OrderID]
WHERE
([Store].[StoreName]= 'Shopify')
AND ([Order].[OrderDate] BETWEEN '2015-09-01 00:00:00.000' AND '2015-09-30 23:59:59.999')
AND ([Order].[IsManual] = '0')
Again I am going to give the same order as an example in the Results
+-------------+---------------+------------------+---------------+---------------+----------+------------------------+--------+----------------+-------+-------+------------+-----------+
| OrderNumber | PaymentStatus | MarketplaceNames | SKU | MainSKU | Type | Description | Amount | Classification | Cost | Sales | Total Cost | Total Qty |
| 7177 | Paid | Shopify | 300TLSH SL PL | 300TLSH SL PL | DISCOUNT | 15chance | -7.35 | Sheet Set | 21.00 | 48.99 | 21 | 1 |
| 7177 | Paid | Shopify | 300TLSH SL PL | 300TLSH SL PL | SHIPPING | FREE Standard Shipping | 0.00 | Sheet Set | 21.00 | 48.99 | 21 | 1 |
+-------------+---------------+------------------+---------------+---------------+----------+------------------------+--------+----------------+-------+-------+------------+-----------+
If I were to export this into an excel file, the Cost, Sales, and Total Qty fields are now all doubled for this order, this becomes an even bigger issue if there are multiple line items in an order.
I thought a solution would be to make the Discount and Shipping Fields their own columns but all that did was put Discount and Shipping on the same line but all the line items were still doubled.
I have to be over looking something.

Answer 1

SELECT [Order].[OrderNumber]
,CASE WHEN [ShopifyOrder].[PaymentStatusCode] = '2' THEN 'Paid'
WHEN [ShopifyOrder].[PaymentStatusCode] = '4' THEN 'Refunded'
WHEN [ShopifyOrder].[PaymentStatusCode] = '5' THEN 'Voided'
WHEN [ShopifyOrder].[PaymentStatusCode] = '6' THEN 'Partially Refunded'
END AS 'PaymentStatus'
,[Store].[StoreName] as 'MarketplaceNames'
,[OrderItem].[SKU]
,[LookupList].[MainSKU]
,[ShippingCharge].[Description] as shippingDescription
,[ShippingCharge].[Amount] as shippingAmount
,[DiscountCharge].[Description] as discountDescription
,[DiscountCharge].[Amount] as discountAmount
,[LookupList].[Classification] as 'Classification'
,[LookupList].[Cost]
,([OrderItem].[Quantity]* [OrderItem].[UnitPrice]) AS 'Sales'
,(([OrderItem].[Quantity] * [LookupList].[Quantity]) * [LookupList].[Cost]) AS 'Total Cost'
,[OrderItem].[Quantity] * [LookupList].[Quantity] AS 'Total Qty'
FROM [SHIPSERVER].[dbo].[Order]
JOIN [SHIPSERVER].[dbo].[ShopifyOrder]
ON [Order].[OrderID]=[ShopifyOrder].[OrderID]
JOIN [SHIPSERVER].[dbo].[OrderItem]
ON [OrderItem].[OrderID]=[Order].[OrderID]
JOIN [SHIPSERVER].[dbo].[Store]
ON [Order].[StoreID]=[Store].[StoreID]
LEFT JOIN [SHIPSERVER].[dbo].[LookupList]
ON [OrderItem].[SKU]=[LookupList].[SKU]
LEFT JOIN [SHIPSERVER].[dbo].[OrderCharge] [ShippingCharge]
ON [Order].[OrderID]=[ShippingCharge].[OrderID] AND [ShippingCharge].[Type] = 'SHIPPING'
LEFT JOIN [SHIPSERVER].[dbo].[OrderCharge] [DiscountCharge]
ON [Order].[OrderID]=[DiscountCharge].[OrderID] AND [DiscountCharge].[Type] = 'DISCOUNT'
WHERE ([Store].[StoreName]= 'Shopify')
AND ([Order].[OrderDate] BETWEEN '2015-09-01 00:00:00.000' AND '2015-09-30 23:59:59.999')
AND ([Order].[IsManual] = '0')
The differences are :
,[ShippingCharge].[Description] as shippingDescription
,[ShippingCharge].[Amount] as shippingAmount
,[DiscountCharge].[Description] as discountDescription
,[DiscountCharge].[Amount] as discountAmount
and
LEFT JOIN [SHIPSERVER].[dbo].[OrderCharge] [ShippingCharge]
ON [Order].[OrderID]=[ShippingCharge].[OrderID] AND [ShippingCharge].[Type] = 'SHIPPING'
LEFT JOIN [SHIPSERVER].[dbo].[OrderCharge] [DiscountCharge]
ON [Order].[OrderID]=[DiscountCharge].[OrderID] AND [DiscountCharge].[Type] = 'DISCOUNT'
Basically, what I did is I left joined on OrderCharge twice, once for Discount and once for Shipping, with a different alias each time. This means that you're potentially linked to a discount row and potentially linked to a shipping row, and from there getting the data is incredibly easy.
As #thab pointed out in comments though, there are glaring issues with this. First of all, having more than one Shipping or Discount entries will duplicate rows, at which point you would have to use a sum on the [Amount] (and probably an XML concatenation on the description). This also means that the query must be altered whenever a new Type of ChargeOrder appears.
The idea solution would be using Pivot, but I haven't dabbled with that yet so I can't help you with that one. I do believe that Pivot tables run slower though (well, at least dynamic ones do), so as long as your problem doesn't change you should be fine.