1

我在 SQL Server 2008 数据库中有大量行

对于每一行，我有 3 列我关心

典型的行如下所示：

AccountNumber    | basecode                      | subcode
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
9689787209368901 | AQTXG AQTXG AQTXG AQTXG ACC5Z | ZQ596 ZQ596 ZQ596 ZQ655 ZC655

某些基码需要连同其相应的子码一起被删除。basecode 和 supcode 都是由空格分隔的有序列表。我需要一个输入基本代码的查询，它会删除基本代码和子代码，并使列表的其余部分保持不变。

每行具有相同数量的基码和子码项。代码的数量可以从 0 到 45 不等。相同的基码可以出现多次。所有字段都是 varchars。

这样数据最终（动态解析）如下：（因为每个账户有可变数量的基码/子码（对）。

如果我想删除“ACC5Z”，那么我应该有

AccountNumber    | basecode                | subcode
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
9689787209368901 | AQTXG AQTXG AQTXG AQTXG | ZQ596 ZQ596 ZQ596 ZQ655

如果我想删除“AQTXG”，那么我应该

AccountNumber    | basecode   | subcode
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
9689787209368901 | ACC5Z      | ZC655
4

1 回答 1

1

试试这个

DECLARE @t TABLE (AccountNumber VARCHAR(100),Basecode VARCHAR(100),Subcode VARCHAR(100))
INSERT INTO @t SELECT 9689787209368901,'AQTXG AQTXG AQTXG AQTXG ACC5Z','ZQ596 ZQ596 ZQ596 ZQ655 ZC655'

DECLARE @BaseCodetoRemove VARCHAR(10) = 'AQTXG'

;WITH CTE AS(
SELECT  
    AccountNumber
    ,LTRIM(RTRIM(SUBSTRING(Basecode, Number ,CHARINDEX(' ', Basecode + ' ', Number ) - Number))) AS Basecode
    , LTRIM(RTRIM(SUBSTRING(Subcode, Number ,CHARINDEX(' ', Subcode + ' ', Number ) - Number))) AS Subcode
FROM @t 
JOIN master..spt_values ON Number <= DATALENGTH(Basecode) + 1  AND type='P'
AND SUBSTRING(' ' + Basecode, Number , 1) = ' ')
,CTE2 AS(
SELECT * 
FROM CTE 
WHERE BaseCode <> @BaseCodetoRemove)

SELECT AccountNumber
    ,STUFF((SELECT  ' ' + CAST(Basecode AS vARCHAR(MAX))
      FROM CTE2 t2 WHERE t1.AccountNumber = t2.AccountNumber
      FOR XML PATH('')),1,1,'')Basecode

    ,STUFF((SELECT ' ' + CAST(Subcode AS vARCHAR(MAX))
      FROM CTE2 t2 WHERE t1.AccountNumber = t2.AccountNumber
      FOR XML PATH('')),1,1,'')Subcode
FROM CTE2  t1
GROUP BY t1.AccountNumber
于 2012-12-05T03:28:21.767 回答