Question

我从消息表中获得了以下column（name-xml）值（XML格式）

<Line>
    <ID>12345</ID>
    <ErrorText>sky is red</ErrorText>
</Line>
<Line>
    <ID>54321</ID>
    <ErrorText>roses are white</ErrorText>
</Line>
<Line>
    <ID>98765</ID>
    <ErrorText>Sun rises in the east</ErrorText>
</Line>

从这里我需要以下数据

ID      ErrorText
12345   sky is red
54321   roses are white
98765   Sun rises in the east

我需要从同一数据库表列中分离出2个单独的列。

Answer 1

尝试一下

DECLARE @mockup TABLE(ID INT IDENTITY, YourXmlColumn XML);
INSERT INTO @mockup VALUES
(
N'<Line><ID>12345</ID><ErrorText>sky is red</ErrorText></Line><Line><ID>54321</ID><ErrorText>roses are white</ErrorText></Line><Line><ID>98765</ID><ErrorText>Sun rises in the east</ErrorText></Line>'
);

SELECT ln.value('(ID/text())[1]','int') AS ID
      ,ln.value('(ErrorText/text())[1]','nvarchar(max)') AS ID
FROM @mockup m
CROSS APPLY m.YourXmlColumn.nodes('/Line') A(ln);

CROSS APPLY允许您调用本地XML.method .nodes()。 XPath“ / Line”将每个重复的<Line>元素作为单独的行返回。方法.value()使您可以选择元素的内容。

更新

如何转换不推荐使用的TEXT类型的示例。我使用CTE进行投射。

WITH casted AS
(
    SELECT t.*
          ,CAST(CAST(t.YourXmlColumn AS VARCHAR(MAX)) AS XML) AS ToXml
    FROM YourTable t
    --WHERE...  (filters go here)
)
SELECT ln.value('(ID/text())[1]','int') AS ID
      ,ln.value('(ErrorText/text())[1]','nvarchar(max)') AS ID
FROM casted
CROSS APPLY casted.ToXml.nodes('/Line') A(ln);

Answer 2

尝试：

DECLARE @XML XML = '
    <Line>
        <ID>12345</ID>
        <ErrorText>sky is red</ErrorText>
    </Line>
    <Line>
        <ID>54321</ID>
        <ErrorText>roses are white</ErrorText>
    </Line>
    <Line>
        <ID>98765</ID>
        <ErrorText>Sun rises in the east</ErrorText>
    </Line>'

SELECT 
    Tbl.Col.value('ID[1]', 'INT') ID
    ,Tbl.Col.value('ErrorText[1]', 'VARCHAR(100)') ErrorText
FROM 
    @XML.nodes('//Line') Tbl(Col)

sql-从同一列中选择多个子字符串

2 个答案:

更新