Question

我使用以下查询从XML文件的属性ad元素中选择XML的值但我无法从XML页面读取seq, id, reported dated属性所以任何人都请建议如何使用此查询获取属性值。

select a_node.value('(./text())[1]', 'var char(50)') AS c_val,
c1_node.value('(./text())[1]', 'var char(50)') AS c_val 2,
ca_node.value('(./text())[1]', 'var char(50)') AS c_val3, 
d_node.value('(./text())[1]', 'var char(50)') ,
e_node.value('(./text())[1]', 'varchar(50)') ,
f_node.value('(./text())[1]', 'var char(50)') 
FROM @xmlData.nodes('/Reports/x:InquiryResponse/x:ReportData/x:AccountDetails/x:Account') AS b(b_node) 
outer APPLY b.b_node.nodes('./x:primarykey') AS pK_InquiryResponse (a_node) 
outer APPLY b.b_node.nodes('./x:seq') AS CustomerCode (c1_node) 
outer APPLY b.b_node.nodes('./x:id') AS amount (ca_node)
outer APPLY b.b_node.nodes('./x:ReportedDate') AS CustRefField (d_node)
outer APPLY b.b_node.nodes('./x:AccountNumber') AS ReportOrderNO (e_node)
outer apply b.b_node.nodes('./x:CurrentBalance') as additional_id (f_node);

修改：评论中提供的Xml代码段

<sch:Account seq="2" id="345778174" ReportedDate="2014-01-01">
    <sch:AccountNumber>TSTC1595</sch:AccountNumber>
    <sch:CurrentBalance>0</sch:CurrentBalance>
    <sch:Institution>Muthoot Fincorp Limited</sch:Institution>
    <sch:PastDueAmount>0</sch:PastDueAmount>
    <sch:DisbursedAmount>12000</sch:DisbursedAmount>
    <sch:LoanCategory>JOG Group</sch:LoanCategory>
</sch:Account>

<sch:Account seq="2" id="345778174" ReportedDate="2014-01-01">
    <sch:BranchIDMFI>THRISSUR ROAD</sch:BranchIDMFI>
    <sch:KendraIDMFI>COSTCO/RECENT-107</sch:KendraIDMFI>
</sch:Account>

Answer 1

使用Xml Loose解析XQuery @Variable

假设一个类似于此的Xml文档（即一个元素上包含所有属性）：

DECLARE @xmlData XML = 
N'<Reports xmlns:x="http://foo">
    <x:InquiryResponse>
        <x:ReportData>
            <x:AccountDetails>
                <x:Account x:primarykey="pk" x:seq="sq" x:id="id"
                             x:ReportedDate="2014-01-01T00:00:00" />
            </x:AccountDetails>
        </x:ReportData>
    </x:InquiryResponse>
</Reports>';

您可以按如下方式删除属性：

WITH XMLNAMESPACES('http://foo' AS x)
select 
    Nodes.node.value('(@x:primarykey)[1]', 'varchar(50)') AS c_val,
    Nodes.node.value('(@x:seq)[1]', 'varchar(50)') AS c_val2,
    Nodes.node.value('(@x:id)[1]', 'varchar(50)') AS c_val3, 
    Nodes.node.value('(@x:ReportedDate)[1]', 'DATETIME') as someDateTime
FROM 
   @xmlData.nodes('/Reports/x:InquiryResponse/x:ReportData/x:AccountDetails/x:Account') 
   AS Nodes(node);

属性不需要text()因为它们是自动字符串
在命名空间中包含属性是相当罕见的 - 如果它们不是，则删除xmlns别名前缀。

SqlFiddle here

编辑 - 解析Xml列

从属性中删除了命名空间 - 假设您拥有表中的数据，而不是变量，因此需要APPLY。请注意，OUTER APPLY将返回空值，例如仅当您有行时才有用空Xml或缺少Xml元素。 CROSS APPLY是常态（即将xpath应用于LHS表中选择的每一行）
访问元素与属性类似，只是没有@

WITH XMLNAMESPACES('http://foo' AS x)
select 
    Nodes.node.value('(@seq)[1]', 'varchar(50)') AS c_val2,
    Nodes.node.value('(@id)[1]', 'varchar(50)') AS c_val3, 
    Nodes.node.value('(@ReportedDate)[1]', 'DATETIME') as someDateTime,
    Nodes.node.value('(x:AccountNumber)[1]', 'VARCHAR(50)') as accountNumber
FROM 
    MyXmlData z
    CROSS APPLY
z.XmlColumn.nodes('/Reports/x:InquiryResponse/x:ReportData/x:AccountDetails/x:Account') 
      AS Nodes(node);

Updated Fiddle

从磁盘编辑Xml文件

对于从磁盘读取的xml文件，这是一回事。请注意，一旦您拥有XML变量（@MyXmlData）中的数据，您不需要CROSS APPLY任何内容 - 只需提供xpath以选择适当的节点，然后删除元素和属性。

DECLARE @MyXmlData XML;
SET @MyXmlData = 
( SELECT * FROM OPENROWSET ( BULK N'c:\temp\file3098.xml', SINGLE_CLOB ) AS MyXmlData );
-- Assuming all on the one element, no need for all the applies
-- attributes don't have a text axis (they are automatically strings

WITH XMLNAMESPACES('http://foo' AS x)
select 
    Nodes.node.value('(@seq)[1]', 'varchar(50)') AS c_val2,
    Nodes.node.value('(@id)[1]', 'varchar(50)') AS c_val3, 
    Nodes.node.value('(@ReportedDate)[1]', 'DATETIME') as someDateTime,
    Nodes.node.value('(x:AccountNumber)[1]', 'VARCHAR(50)') as accountNumber
FROM 
      @MyXmlData.nodes('/Reports/x:InquiryResponse/x:ReportData/x:AccountDetails/x:Account') 
      AS Nodes(node);

使用Xquery检索xml属性

1 个答案: