使用左连接获取所有日期

Question

我正在努力库存管理系统我在哪里追踪产品的发布日期。所以，我的条件是 - 如果存储ID为1，今年是2014年，那么它应该返回所有的产品针对特定年份的发行日期如下：

Product - Date - TotalIssued 
Keyboard - 2014-10-10 - 1 -- This product has been issued or given 1 time on that date of the year 2014 
Mouse - 2014-10-11 - 1 
Keyboard - 2014-10-12 - 0 
Mouse - 2014-10-12 - 0 
------------------------- 
------------------------- 
Keyboard - 2014-12-31 - 0 
Mouse - 2014-12-31 - 0 -- This will continue till the end of the year 

我试图使用日历表使用下面的链接，并创建一个：

Calendar Table

我用下面的查询使用日历表以获得所需的输出，但它仅返回已发出，这意味着仅存储在该日期的产品在ReturnToStore表：

SELECT l.ItemName AS Product, m.PKDate AS Date, COUNT(m.PKDate) AS TotalIssued 
FROM [days] m 
LEFT JOIN ReturnToStore k ON CONVERT(DATE, k.IssuedDate) = m.PKDate 
LEFT JOIN Item l ON l.ItemID = k.ItemID 
WHERE m.calendar_year = 2014 AND k.StoreID = 1 
GROUP BY k.StoreID, l.ItemName, m.PKDate 

现在这是我得到的输出：正如我我已经更新后：

Product - Date - TotalIssued 
Keyboard - 2014-10-10 - 1 
Mouse - 2014-10-11 - 1 

下面是表的结构：

CREATE TABLE [dbo].[Item](
     [ItemID] [int] IDENTITY(1,1) NOT NULL, 
     [ItemName] [varchar](50) NULL, 
     [Description] [varchar](150) NULL, 
     [Tag] [varchar](50) NULL, 
    CONSTRAINT [PK_Item] PRIMARY KEY CLUSTERED 
    (
     [ItemID] ASC 
    ) 
    ) ON [PRIMARY] 

    INSERT INTO [dbo].[Item] 
       ([ItemName] 
       ,[Description] 
       ,[Tag]) 
     VALUES 
       ('Keyboard', 'IT Equipment', ''), 
       ('Mouse', 'IT Equipment', '') 

    CREATE TABLE [dbo].[ReturnToStore](
     [ID] [int] IDENTITY(1,1) NOT NULL, 
     [StoreID] [int] NULL, 
     [CategoryID] [int] NULL, 
     [ItemID] [int] NULL, 
     [Quantity] [float] NULL, 
     [IssuedDate] [datetime] NULL, 
     [Description] [varchar](150) NULL, 
     [OrderID] [varchar](50) NULL, 
    CONSTRAINT [PK_ReturnToStore] PRIMARY KEY CLUSTERED 
    (
     [ID] ASC 
    ) 
    ) ON [PRIMARY] 

    INSERT INTO [dbo].[ReturnToStore] 
       ([StoreID] 
       ,[CategoryID] 
       ,[ItemID] 
       ,[Quantity] 
       ,[IssuedDate] 
       ,[Description] 
       ,[OrderID]) 
     VALUES 
       (1, 1, 1, 10, '2014-10-10 00:00:00.000', '', 'PO-02'), 
       (1, 1, 2, 20, '2014-10-11 00:00:00.000', '', 'PO-03') 

**Note:** I guess, this could be done easily with LEFT JOIN using calendar table. But got stuck here. 

编辑将查询转换为以下几点：

CREATE PROCEDURE [dbo].[MonthlyConsumption] 
@StoreID int, 
@Year int 

AS 

SELECT Distinct ItemName, 
    [1] AS M1, 
    [2] AS M2, 
    [3] AS M3, 
    [4] AS M4, 
    [5] AS M5, 
    [6] AS M6, 
    [7] AS M7, 
    [8] AS M8, 
    [9] AS M9, 
    [10] AS M10, 
    [11] AS M11, 
    [12] AS M12 
FROM 
(SELECT MONTH(m.PKDate) as YEAR, 
p.ItemName, k.IssuedDate, k.ItemID 
FROM [days] m 
CROSS JOIN (SELECT DISTINCT itemName,itemID FROM Item) p 
LEFT JOIN ReturnToStore k 
ON CONVERT(DATE, k.IssuedDate) = m.PKDate AND 
    p.itemid = k.itemid AND 
    k.StoreID = @StoreID 
LEFT JOIN Item l 
ON l.ItemID = k.ItemID 
WHERE m.calendar_year = @Year 
GROUP BY p.ItemName, m.PKDate, k.ItemID, k.IssuedDate) 

source 
PIVOT 
(
    COUNT(ItemID) 
    FOR YEAR 
    IN ([1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12]) 
) AS PivotDay 

返回下面的输出：

我不知道为什么它是重复的，如果我添加了同样的产品比一次在特定日期较多，那么就没有在枢轴所示变化并且它与给定的输出保持一致。

Answer 1

你的第一个错误实际上是由于你的WHERE条款，而不是ON子句中放置在LEFT JOIN的右表的条件。

第二件事情，你也应该使用派生表项目名称，就像日期：

SELECT p.ItemName AS Product, m.PKDate AS Date, COUNT(l.ItemID) AS TotalIssued 
FROM [days] m 
CROSS JOIN (SELECT DISTINCT itemName,itemID FROM Item) p 
LEFT JOIN ReturnToStore k 
ON CONVERT(DATE, k.IssuedDate) = m.PKDate AND 
    p.itemid = k.itemid AND 
    k.StoreID = 1 
LEFT JOIN Item l 
ON l.ItemID = k.ItemID 
WHERE m.calendar_year = 2014 
GROUP BY p.ItemName, m.PKDate 

Answer 2

我不知道，如果这是解决方案，但至少这是一种进步...

当LEFT JOIN，把右侧表的条件ON子句中得到真正的LEFT JOIN行为。 （当在WHERE时，你会得到经常的INNER JOIN结果。） move k.StoreID = 1到ON条款：

SELECT l.ItemName AS Product, m.PKDate AS Date, COUNT(m.PKDate) AS TotalIssued 
FROM [days] m 
LEFT JOIN ReturnToStore k ON CONVERT(DATE, k.IssuedDate) = m.PKDate AND k.StoreID = 1 
LEFT JOIN Item l ON l.ItemID = k.ItemID 
WHERE m.calendar_year = 2014 
GROUP BY l.ItemName, m.PKDate