Sql 仅当所有列都为空时才从其他表中选择_Sql_Sql Server_Sql Server 2008

Sql 仅当所有列都为空时才从其他表中选择

sql sql-server sql-server-2008

Sql 仅当所有列都为空时才从其他表中选择,sql,sql-server,sql-server-2008,Sql,Sql Server,Sql Server 2008,我使用以下SQL来获取员工的地址和其他信息 SELECT emp_id, emp_name, @addr1 = ISNULL(ma.addr1, ba.addr1), @addr2 = ISNULL(ma.addr2, ba.addr2), @addr3 = ISNULL(ISNULL(ma.addr3, ba.addr3),''), @addr4 = ISNULL(ISNULL(ma.addr4, ba.addr4),''), @addr5

我使用以下SQL来获取员工的地址和其他信息

SELECT   
  emp_id,  
  emp_name,
  @addr1 = ISNULL(ma.addr1, ba.addr1),  
  @addr2 = ISNULL(ma.addr2, ba.addr2),  
  @addr3 = ISNULL(ISNULL(ma.addr3, ba.addr3),''),  
  @addr4 = ISNULL(ISNULL(ma.addr4, ba.addr4),''),  
  @addr5 = ISNULL(ISNULL(ma.addr5, ba.addr5),''),  
  @postCode = ISNULL(ma.postcode, ba.postcode)  
 FROM dbo.employee e
 INNER JOIN MainAddress ma ON ma.AddressId=e.AddressId
 INNER JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

问题是，当主地址中的Addr1或其他字段为空时，它将采用辅助地址中的Addr3或其他字段，并显示这两个地址的混合。这是完全无效的

当主地址中的所有字段都为空（不可用）时，如何获取辅助地址？

注意：用户可以同时更新两个表，也可以只更新一个表

如果“不可用”表示

main address

表中没有记录，则

ma.AddressId

为空，因此，您可以对每个字段使用此条件：

  CASE WHEN ma.AddressId IS NULL THEN ba.addr1 ELSE ma.addr1 END,

如果确实需要检查所有字段是否为空，则必须按如下方式更改案例条件：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

注意：如果有两种可能的情况，一种是根本没有行，另一种是行存在，但值都为空，则需要如下检查：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

正如评论中所说，我忘记告诉您必须使用

左连接

，而不是

内部连接

。如果不是，则如果在任何adress表中都没有员工的记录，则查询不会返回该员工

顺便说一句，如果您有一个地址表，并且有一个用于指定地址类型的鉴别器列：主地址或辅助地址，那么就容易多了。如果您这样做了，您只需从employee表中选择一个

左联接

，并从address表中选择一个查询来获取

TOP 1

。这将使您的应用程序开发人员在许多其他情况下变得更容易，例如查找一个您不知道是否是主地址的地址

更新：使用CTE

至于OP注释，有必要使用长表达式。为了使其更加清晰和易于维护，可以像这样使用CTE：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

如果“不可用”表示

main address

表中没有记录，则

ma.AddressId

为空，因此，您可以对每个字段使用此条件：

  CASE WHEN ma.AddressId IS NULL THEN ba.addr1 ELSE ma.addr1 END,

如果确实需要检查所有字段是否为空，则必须按如下方式更改案例条件：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

注意：如果有两种可能的情况，一种是根本没有行，另一种是行存在，但值都为空，则需要如下检查：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

正如评论中所说，我忘记告诉您必须使用

左连接

，而不是

内部连接

。如果不是，则如果在任何adress表中都没有员工的记录，则查询不会返回该员工

左联接

，并从address表中选择一个查询来获取

TOP 1

。这将使您的应用程序开发人员在许多其他情况下变得更容易，例如查找一个您不知道是否是主地址的地址

更新：使用CTE

至于OP注释，有必要使用长表达式。为了使其更加清晰和易于维护，可以像这样使用CTE：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

如果“不可用”表示

main address

表中没有记录，则

ma.AddressId

为空，因此，您可以对每个字段使用此条件：

  CASE WHEN ma.AddressId IS NULL THEN ba.addr1 ELSE ma.addr1 END,

如果确实需要检查所有字段是否为空，则必须按如下方式更改案例条件：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

注意：如果有两种可能的情况，一种是根本没有行，另一种是行存在，但值都为空，则需要如下检查：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

正如评论中所说，我忘记告诉您必须使用

左连接

，而不是

内部连接

。如果不是，则如果在任何adress表中都没有员工的记录，则查询不会返回该员工

左联接

，并从address表中选择一个查询来获取

TOP 1

。这将使您的应用程序开发人员在许多其他情况下变得更容易，例如查找一个您不知道是否是主地址的地址

更新：使用CTE

至于OP注释，有必要使用长表达式。为了使其更加清晰和易于维护，可以像这样使用CTE：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

如果“不可用”表示

main address

表中没有记录，则

ma.AddressId

为空，因此，您可以对每个字段使用此条件：

  CASE WHEN ma.AddressId IS NULL THEN ba.addr1 ELSE ma.addr1 END,

如果确实需要检查所有字段是否为空，则必须按如下方式更改案例条件：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

注意：如果有两种可能的情况，一种是根本没有行，另一种是行存在，但值都为空，则需要如下检查：

  CASE WHEN 
    ma.addres1 IS NULL AND ma.addres2 IS NULL ...
  THEN ba.addr1 ELSE ma.addr1 END,

WITH ea AS -- the CTE, ea includes "employee + MainAdressExists column"
(
SELECT 
    *, -- select the required columns
    -- plus an extra calculated column
    CASE WHEN 
       ma.address1 IS NOT NULL 
       OR ma.address2 IS NOT NULL
       -- add here all the column checks
    THEN 1 ELSE 0 END AS MainAddressExists 
FROM employee e
LEFT JOIN MainAddres ma ON ma.IdAddress = e.IdAdress
)
SELECT
  -- choose the needed fields from ea
  CASE WHEN P.MainAddressExists = 1 THEN ma.addres1 ELSE ba.address1 END 
  AS address1,
  -- choose all the other address fields
FROM ea
LEFT JOIN MainAddress ma ON ma.AddressId=e.AddressId
LEFT JOIN SecondaryAddress ba ON ba.AddressId=e.AddressId

正如评论中所说，我忘记告诉您必须使用

左连接

，而不是

内部连接

。如果不是，则如果在任何adress表中都没有员工的记录，则查询不会返回该员工

左联接

，并从address表中选择一个查询来获取

TOP 1

。这将使您的应用程序开发人员在许多其他情况下变得更容易，例如查找一个您不知道是否是主地址的地址

更新：使用CTE

至于OP