您如何将其转换为 SQL Server 2008 语句？

Question

我不确定如何将其从 Access 转换为 SQL。我们目前正在使用 SQL Server 2008。我正在尝试将 Access 报告迁移到 SQL 视图中。

SELECT 
  First(IIf(IsNull([dbo_CUST_ADDRESS]![STATE]),IIf(IsNull([dbo_CUSTOMER]![BILL_TO_STATE]),[dbo_CUSTOMER]![STATE],[dbo_CUSTOMER]![BILL_TO_STATE]),[dbo_CUST_ADDRESS]![STATE])) AS STATE
  , dbo_RECEIVABLE.INVOICE_ID
  , dbo_RECEIVABLE.INVOICE_DATE
  , dbo_RECEIVABLE.CUSTOMER_ID
  , First(IIf(IsNull([dbo_CUST_ADDRESS]![NAME]),IIf(IsNull([dbo_CUSTOMER]![BILL_TO_NAME]),[dbo_CUSTOMER]![NAME],[dbo_CUSTOMER]![BILL_TO_NAME]),[dbo_CUST_ADDRESS]![NAME])) AS COMPANY
  , Sum(IIf([dbo_RECEIVABLE_LINE]![GL_ACCOUNT_ID]='4005-008',0,[dbo_RECEIVABLE_LINE]![AMOUNT])) AS Inv_Total_Amt
  , Sum(IIf([dbo_RECEIVABLE_LINE]![REFERENCE]<>"LA-COUNTY" And [dbo_RECEIVABLE_LINE]![REFERENCE]<>"FREIGHT" And [dbo_RECEIVABLE_LINE]![GL_ACCOUNT_ID]<>'4005-008',[dbo_RECEIVABLE_LINE]![AMOUNT],0)) AS Subtotal
  , Sum(IIf([dbo_RECEIVABLE_LINE]![REFERENCE]="LA-COUNTY",[dbo_RECEIVABLE_LINE]![AMOUNT],0)) AS TAX
  , Sum(IIf([dbo_RECEIVABLE_LINE]![REFERENCE]="FREIGHT",[dbo_RECEIVABLE_LINE]![AMOUNT],0)) AS Freight
  , dbo_RECEIVABLE.SALESREP_ID    
FROM (((dbo_RECEIVABLE RIGHT JOIN dbo_RECEIVABLE_LINE ON dbo_RECEIVABLE.INVOICE_ID = dbo_RECEIVABLE_LINE.INVOICE_ID) LEFT JOIN dbo_CUSTOMER_ORDER ON dbo_RECEIVABLE_LINE.CUST_ORDER_ID = dbo_CUSTOMER_ORDER.ID) LEFT JOIN dbo_CUST_ADDRESS ON (dbo_CUSTOMER_ORDER.CUSTOMER_ID = dbo_CUST_ADDRESS.CUSTOMER_ID) AND (dbo_CUSTOMER_ORDER.SHIP_TO_ADDR_NO = dbo_CUST_ADDRESS.ADDR_NO)) LEFT JOIN dbo_CUSTOMER ON dbo_RECEIVABLE.CUSTOMER_ID = dbo_CUSTOMER.ID    
GROUP BY 
  dbo_RECEIVABLE.INVOICE_ID
  , dbo_RECEIVABLE.INVOICE_DATE
  , dbo_RECEIVABLE.CUSTOMER_ID
  , dbo_RECEIVABLE.SALESREP_ID    
ORDER BY 
  First(IIf(IsNull([dbo_CUST_ADDRESS]![STATE]),IIf(IsNull([dbo_CUSTOMER]![BILL_TO_STATE]),[dbo_CUSTOMER]![STATE],[dbo_CUSTOMER]![BILL_TO_STATE]),[dbo_CUST_ADDRESS]![STATE]))
  , dbo_RECEIVABLE.INVOICE_ID;

这是我在 2015 年 10 月 21 日创建的内容。我更改了 WHERE 以查找发票编号以对其进行测试并添加嵌套 CASE。它显示了该发票的 2 行，并且没有将它们分组为 1。

选择
(SELECT TOP 1 CASE WHEN CUST_ADDRESS.STATE is NULL THEN CASE WHEN CUSTOMER.BILL_TO_STATE is NULL THEN CUSTOMER.STATE ELSE CUSTOMER.BILL_TO_STATE END ELSE CUST_ADDRESS.STATE END FROM dbo.RECEIVABLE RECEIVABLE ERIVEINABLE dbo.RECEIVABLE . dbo.RECEIVABLE_LINE.INVOICE_ID LEFT OUTER JOIN dbo.CUSTOMER_ORDER ON dbo.RECEIVABLE_LINE.CUST_ORDER_ID = dbo.CUSTOMER_ORDER.ID LEFT OUTER JOIN dbo.CUST_ADDRESS ON dbo.CUSTOMER.CUST_ORDER_Dbo.CUSTOMER.CUST_ORDER_ID = dbo.CUSTOMER.CUST_ORDER_Dbo.CUSTOMER. CUST_ADDRESS。ADDR_NO LEFT OUTER JOIN dbo.CUSTOMER ON dbo.RECEIVABLE.CUSTOMER_ID = dbo.CUSTOMER.ID WHERE dbo.RECEIVABLE.INVOICE_ID = 'iv184546') 作为状态, dbo.RECEIVABLE.RECINVOICE_ID,IVABLE_DBOICEDVOICE_ID. , dbo.RECEIVABLE.CUSTOMER_ID, (SELECT TOP 1 CASE WHEN dbo.CUST_ADDRESS.NAME 为 NULL THEN dbo.CUSTOMER.NAME ELSE dbo.CUST_ADDRESS.NAME END FROM dbo.RECEIVABLE RIGHT OUTER JOIN dbo.RECEIVABLEDLINE_VOIN dbo.RECEIVABLE_LINE.INVOICE_ID LEFT OUTER JOIN dbo.CUSTOMER_ORDER ON dbo.RECEIVABLE_LINE.CUST_ORDER_ID = dbo.CUSTOMER_ORDER。ID LEFT OUTER JOIN dbo.CUST_ADDRESS ON dbo.CUSTOMER_ORDER.CUSTOMER_ID = dbo.CUST_ADDRESS.CUSTOMER_ID AND dbo.CUSTOMER_ORDER.SHIP_TO_ADDR_NO = dbo.CUST_ADDRESS.RECADDR_NO LEFTINSTOMER_ORDER.CUSTOMER_ID = dbo.CUST_ADDRESS.CUSTOMER_ID dbo.RECEIVABLE.INVOICE_ID = 'iv184546') AS COMPANY, SUM(CASE WHEN RECEIVABLE_LINE.GL_ACCOUNT_ID = '4005-008' THEN 0 ELSE RECEIVABLE_LINE.AMOUNT END) AS Inv_Total_Amt, SUM-CASE WHEN CASE ' AND RECEIVABLE_LINE.REFERENCE <> 'FREIGHT' AND RECEIVABLE_LINE.GL_ACCOUNT_ID <> '4005-008' THEN RECEIVABLE_LINE.AMOUNT ELSE 0 END) 作为小计，SUM（CASE WHEN RECEIVABLE_LINE.REFERENCE = 'LA-COUNTY' THEN RECEIVABLE_LINE.AMOUNT ELSE 0 END）作为税款，SUM（CASE WHEN RECEIVABLE_LINE.REFERENCE = 'FREIGHT' THEN RECEIVABLE_LINE.AMOUNT ELSE 0 END。作为税款） SALESREP_ID

从 dbo.RECEIVABLE RIGHT OUTER JOIN dbo.RECEIVABLE_LINE ON dbo.RECEIVABLE.INVOICE_ID = dbo.RECEIVABLE_LINE.INVOICE_ID LEFT OUTER JOIN dbo.CUSTOMER_ORDER ON dbo.RECEIVABLE_LINEID.CUST_ORDERdbo.RECEIVABLE_LINEID.CUST_ORDERDBOFTA.CUST_ORDERdbo.RECEIVABLE_LINE.INVOICE_ID .CUSTOMER_ID = dbo.CUST_ADDRESS.CUSTOMER_ID AND dbo.CUSTOMER_ORDER.SHIP_TO_ADDR_NO = dbo.CUST_ADDRESS.ADDR_NO LEFT OUTER JOIN dbo.CUSTOMER ON dbo.RECEIVABLE.CUSTOMER_ID = .dboID6E4MER_ID=.dboID6E4MER_IDBOE4MER_4MERICUSTOMER_IDABLEIVECUID6 CUST_ADDRESS.STATE 为 NULL 时的情况 CUSTOMER.BILL_TO_STATE 为 NULL 时的情况则 CUSTOMER.STATE ELSE 客户。BILL_TO_STATE END ELSE CUST_ADDRESS.STATE END, dbo.CUSTOMER.NAME, dbo.RECEIVABLE.INVOICE_ID, dbo.RECEIVABLE.INVOICE_DATE, dbo.RECEIVABLE.POSTING_DATE, dbo.RECEIVABLE.CUSTOMER_SREIVABLEd.CUSTOM.INVOICE_ID, dbo.RECEIVABLE.INVOICE_DATE. NAME IS NULL THEN dbo.CUSTOMER.NAME ELSE dbo.CUST_ADDRESS.NAME END当 dbo.CUST_ADDRESS.NAME 为 NULL 时的情况，则 dbo.CUSTOMER.NAME ELSE dbo.CUST_ADDRESS.NAME END当 dbo.CUST_ADDRESS.NAME 为 NULL 时的情况，则 dbo.CUSTOMER.NAME ELSE dbo.CUST_ADDRESS.NAME END
ORDER BY dbo.RECEIVABLE.INVOICE_ID, dbo.RECEIVABLE.INVOICE_DATE;

Answer 1

而不是重写整个查询，我会给你一些关于如何做的提示：

• IIf(condition, a, b) 变成： CASE WHEN condition THEN a ELSE b END
• IsNull(x) 变成： x IS NULL
• IIf(IsNull(x), y, x))然后可以将组合写为： CASE WHEN x IS NULL THEN y ELSE x END并进一步简化为COALESCE(x, y)。
• 表列表达式:[tab]![col]变为:[tab].[col]如果名称未保留，则更具可读性：tab.col
• 对字符串文字使用单引号（而不是双引号）。例如，替换"FREIGHT"为'FREIGHT'。双引号在标准 SQL（以及 SQL Server 中，除了括号之外）用于引用表和列的名称和别名，而不是用于文字。
• FROM可以安全地删除子句中的括号。SQL Server 的解析器不管有没有它们都很满意。

• 此外，如果您为表使用别名，可读性会增加：

  FROM (((dbo_RECEIVABLE RIGHT JOIN dbo_RECEIVABLE_LINE ON dbo_RECEIVABLE.INVOICE_ID = dbo_RECEIVABLE_LINE.INVOICE_ID) LEFT JOIN dbo_CUSTOMER_ORDER ON dbo_RECEIVABLE_LINE.CUST_ORDER_ID = dbo_CUSTOMER_ORDER.ID) LEFT JOIN dbo_CUST_ADDRESS ON (dbo_CUSTOMER_ORDER.CUSTOMER_ID = dbo_CUST_ADDRESS.CUSTOMER_ID) AND (dbo_CUSTOMER_ORDER.SHIP_TO_ADDR_NO = dbo_CUST_ADDRESS.ADDR_NO)) LEFT JOIN dbo_CUSTOMER ON dbo_RECEIVABLE.CUSTOMER_ID = dbo_CUSTOMER.ID 
  

  变成：

  FROM dbo_RECEIVABLE 
    RIGHT JOIN dbo_RECEIVABLE_LINE 
      ON  dbo_RECEIVABLE.INVOICE_ID = dbo_RECEIVABLE_LINE.INVOICE_ID
    LEFT JOIN dbo_CUSTOMER_ORDER 
      ON  dbo_RECEIVABLE_LINE.CUST_ORDER_ID = dbo_CUSTOMER_ORDER.ID
    LEFT JOIN dbo_CUST_ADDRESS 
      ON  dbo_CUSTOMER_ORDER.CUSTOMER_ID = dbo_CUST_ADDRESS.CUSTOMER_ID 
      AND dbo_CUSTOMER_ORDER.SHIP_TO_ADDR_NO = dbo_CUST_ADDRESS.ADDR_NO
    LEFT JOIN dbo_CUSTOMER 
      ON  dbo_RECEIVABLE.CUSTOMER_ID = dbo_CUSTOMER.ID 
  

  进而：

  FROM dbo_RECEIVABLE AS rec
    RIGHT JOIN dbo_RECEIVABLE_LINE AS line
      ON  rec.INVOICE_ID = line.INVOICE_ID
    LEFT JOIN dbo_CUSTOMER_ORDER AS ord
      ON  line.CUST_ORDER_ID = ord.ID
    LEFT JOIN dbo_CUST_ADDRESS AS addr
      ON  ord.CUSTOMER_ID = addr.CUSTOMER_ID 
      AND ord.SHIP_TO_ADDR_NO = addr.ADDR_NO
    LEFT JOIN dbo_CUSTOMER AS cust
      ON  rec.CUSTOMER_ID = cust.ID 
  

  相应地调整SELECT和中的所有引用GROUP BY。

• 在ORDER BY你可以使用包含或定义的列别名SELECT列表。无需重复复杂的表达式或表别名：

  ORDER BY STATE, INVOICE_ID; ; 
  

• 顺便说RIGHT JOIN一下，这看起来不对。特别是因为GROUP BY是基于可为空的侧列（dbo_RECEIVABLE表）。它可能应该是INNER或LEFT JOIN。

• 该First()功能是最困难的部分。你可以用MIN()or替换它，MAX()但你可能不会得到相同的结果，并且可能无法在所有查询中使用这个技巧，只有那些带有GROUP BY.
  等价物将非常复杂，使用其中之一ROW_NUMBER()并更改整个查询的结构（将其包装在派生表或 CTE 中并将其GROUP BY从内部级别移动到外部级别）。在 2012+ 版本中，FIRST_VALUE()也可能会使用该功能。