SQL 字符串操作以添加 xml 节点

Question

我正在使用 sql 2008R2，该表具有类似 xml 列-

<New>
   <From>
        <Scale>Tony</Scale>
        <ScaleName>Name</ScaleName>
    </From>
</New>
<New>
    <From>
        <Scale>Tom</Scale>
        <ScaleName>Name</ScaleName>
    </From>
</New>

 <New>
    <From>
        <Scale>Seven</Scale>
        <ScaleName>Height</ScaleName>
    </From>
</New>
<New>
    <From>
        <Scale>Ten</Scale>
        <ScaleName>Height</ScaleName>
    </From>
</New>
<New>
    <From>
        <Scale>***XXX***</Scale>
        <ScaleName>Height</ScaleName>
   </From>
</New>

.......等等

我需要编写一个 SQL，它可以检查 ScaleName 的 Scale 没有 XXX 作为值的所有节点，然后添加/插入以下文本，为 2次。当只有一个 ***XXX**** 条目时，它应该只添加/插入一次

<New>
    <From>
        <Scale>***XXX***</Scale>
        <ScaleName>Respective Scalename</ScaleName>
    </From>
</New>  

预期结果-----

<New>
   <From>
        <Scale>Tony</Scale>
        <ScaleName>Name</ScaleName>
    </From>

<New>
    <From>
        <Scale>Tom</Scale>
        <ScaleName>Name</ScaleName>
    </From>
</New>
<New>
   <From>
        <Scale>***XXX***</Scale>
        <ScaleName>Name</ScaleName>
    </From>
</New>
<New>

   <From>
        <Scale>***XXX***</Scale>
        <ScaleName>Name</ScaleName>
    </From>
</New>


 <New>
    <From>
        <Scale>Seven</Scale>
        <ScaleName>Height</ScaleName>
    </From>
 </New>
 <New>
    <From>
        <Scale>Ten</Scale>
        <ScaleName>Height</ScaleName>
    </From>
 </New>
 <New>
    <From>
        <Scale>***XXX***</Scale>
        <ScaleName>Height</ScaleName>
   </From>
 </New>
 <New>
    <From>
        <Scale>***XXX***</Scale>
        <ScaleName>Height</ScaleName>
    </From>
</New>

Answer 1

我不知道我是否完全理解您的需要，但这可能会有所帮助：

注意：大多数情况下是这样！ - 使用 magic 值（例如 ***XXX***...

是个坏主意

这是您的示例 XML。标尺Name 没有***XXX*** 条目，而标尺height 有一个...

DECLARE @xml XML=
(N'<New>
  <From>
    <Scale>Tony</Scale>
    <ScaleName>Name</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>Tom</Scale>
    <ScaleName>Name</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>Seven</Scale>
    <ScaleName>Height</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>Ten</Scale>
    <ScaleName>Height</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>***XXX***</Scale>
    <ScaleName>Height</ScaleName>
  </From>
</New>');

--CTE 将 XML 读入派生表中，省略带有 ***XXX*** 的条目

WITH ScaleNames AS
(
    SELECT  fr.value('(Scale)[1]','nvarchar(100)') AS Scale
           ,fr.value('(ScaleName)[1]','nvarchar(100)') AS ScaleName
    FROM @xml.nodes('/New/From') AS A(fr)
    WHERE fr.value('(Scale)[1]','nvarchar(100)')<>'***XXX***'
)

--SELECT 将使用 real 值重建整个 XML，并添加两倍的 ***XXX*** 节点。

SELECT (
            SELECT x.Scale AS [From/Scale]
                  ,x.ScaleName AS [From/ScaleName]
            FROM ScaleNames AS x
            WHERE x.ScaleName=ScaleNames.ScaleName
            FOR XML PATH('New'),TYPE
       )
      ,(SELECT
        (SELECT '***XXX***' AS Scale, ScaleName FOR XML PATH('From'),ROOT('New'),TYPE )
        ,(SELECT '***XXX***' AS Scale, ScaleName FOR XML PATH('From'),ROOT('New'),TYPE )
        FOR XML PATH(''),TYPE
       ) AS [node()]
FROM ScaleNames
GROUP BY ScaleName
FOR XML PATH('')

结果

<New>
  <From>
    <Scale>Seven</Scale>
    <ScaleName>Height</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>Ten</Scale>
    <ScaleName>Height</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>***XXX***</Scale>
    <ScaleName>Height</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>***XXX***</Scale>
    <ScaleName>Height</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>Tony</Scale>
    <ScaleName>Name</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>Tom</Scale>
    <ScaleName>Name</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>***XXX***</Scale>
    <ScaleName>Name</ScaleName>
  </From>
</New>
<New>
  <From>
    <Scale>***XXX***</Scale>
    <ScaleName>Name</ScaleName>
  </From>
</New>