java中通过xpath处理含有命名空间

    声明：以下的测试是用的jdk1.4.2，xalan7.0对于如下xml文档片断：

<ML:MREML xmlns:ML="MREML">
    <EnvelopEntity>
        <EnvelopID>GUID2006102000002</EnvelopID>
        <EnvelopVersion>1.0.1R</EnvelopVersion>
        <ExchangeTime>2006-10-25 13:12:10</ExchangeTime>
        <SendFrom>
            <SFID>SOBEY_NEWS</SFID>
                           <SFDescription>SOBEY新闻系统</SFDescription>
        </SendFrom>
        <SendTo>
            <STID>DAYANG_MAM</STID>
            <STDescription>DAYANG媒资系统</STDescription>
        </SendTo>
        <Priority>2</Priority>
        <EntityInfo>
            <EntityID>REQUEST_ID_01</EntityID>
            <EntityType>4</EntityType>
        </EntityInfo>
        <EntityInfo>
            <EntityID>节目GUID</EntityID>
            <EntityType>3</EntityType>
        </EntityInfo>
        <EntityInfo>
            <EntityID>管理信息实体ID</EntityID>
            <EntityType>2</EntityType>
        </EntityInfo>
    </EnvelopEntity>
</ML:MREML >
    对于上述含有命名空间的xml文档，如果想通过xpath查找EnvelopID的值，有两种方式：1.通过利用xpath的函数local-name（）

    如上述查找内容的xpath表达式可以写为://*[local-name（）='MREML']/EnvelopEntity/EnvelopID/text（）

    2.通过在java程序中处理上述文档的命名空间是定义的，如果要使xpath表达式能正确地被解析需要在java程序中建立起prefix和uri二者的映射关系
    public static Node parseXPath(String expression, Object obj, QName qname)
            throws Exception {
        javax.xml.xpath.XPath xpath = javax.xml.xpath.XPathFactory.newInstance().newXPath();
        xpath.setNamespaceContext(getNamespaceContext());
        if (qname.equals(XPathConstants.NODE))
            return (Node) xpath.evaluate(expression, obj, qname);
        return null;
    }

  public static NamespaceContext getNamespaceContext() throws Exception {
        return new NamespaceContext() {
            public String getNamespaceURI(String prefix) {
/*
                // 一种方式:
                //这里可以采用配置文件的方式,先将所需要使用的xmlNamespace配置好,
                //采用注册的方式供应用使用,这种方式效率应该高一些,不用每次都要从文档中提取namespace
                //不过没有第二种方便
                String uri;
                if (prefix.equalsIgnoreCase("ML"))
                    uri = "MREML";
                else if (prefix.equalsIgnoreCase("RE"))
                    uri = "http://herry.com.cn";
                else if (prefix.equalsIgnoreCase("RID"))
                    uri = "ResourceID";
                else if (prefix.equalsIgnoreCase("RUI"))
                    uri = "ResourceUniqueID";
                else if (prefix.equalsIgnoreCase("RMDI"))
                    uri = "ResourceMetaDataInfo";
                else
                    uri = null;
                System.out.println(ParseXMLUtil.class + "::getNamespaceURI:prefix= " + prefix);
                return uri;
*/
                /*
                                //                另外一种方式: 通过PrefixResolver来提取出prefix和Namespace的对应关系
                                                final PrefixResolver resolver =
                                                        new PrefixResolverDefault(doc.getDocumentElement());

                                                return resolver.getNamespaceForPrefix(prefix);

                */
                
            // Dummy implementation - not used!
            public java.util.Iterator getPrefixes(String val) {
                return null;
            }

            // Dummy implemenation - not used!
            public String getPrefix(String uri) {
                return null;
            }
        };
    }
    若要查找ResourceUniqueID的值，则xpath应该书写为：//ML：MREML/ML：ResourceEntity/RMDI：ResourceMetaDataInfo/RMDI：ResourceID/RUI：ResourceUniqueID/text（）

    这里的ML（prefix）和URI（MERML）等已经通过

    对于含有默认命名空间地文档如果采用NamespaceContext的方式，即将prefix和uri已经映射起来，如对于以下的文档：
<MREML xmlns="MREML"><ResourceEntity >
        <ResourceMetaDataInfo>
            <ResourceID xmlns="ResourceID">
                <RUI:ResourceUniqueID xmlns:RUI="ResourceUniqueID">节目GUID</RUI:ResourceUniqueID>
                <UserDefID>UserDefID填写节目代码</UserDefID>
            </ResourceID>
            <Title>
                <ResourceName>911新闻素材</ResourceName>
            </Title>
            <ResourceType>1</ResourceType>
            <Subject>
                <KeyWords>911 袭击</KeyWords>
            </Subject>
            <Description>
                <ContentDescription>美国遭遇911袭击现场30S画面及同期声素材</ContentDescription>
                <Column>
                    <ColumnName>今日世界</ColumnName>
                </Column>
            </Description>
            <Date>
                <CreateDate>2006-10-25 13:12:10</CreateDate>
            </Date>
            <Format>
                <TotalLength>00:04:35:12</TotalLength>
                <NumberofElements>2</NumberofElements>
                <MarkPoint>
                    <StartPoint>00:00:00:05</StartPoint>
                    <EndPoint>00:04:35:10</EndPoint>
                </MarkPoint>
            </Format>
        </ResourceMetaDataInfo>
</ResourceEntity>
</MREML>

    若要查找ResourceUniqueID的值，则xpath应该书写为：//ML：MREML/ML：ResourceEntity/RMDI：ResourceMetaDataInfo/RMDI：ResourceID/RUI：ResourceUniqueID/text（）

    这里的ML（prefix）和URI（MERML）等已经通过
if (prefix.equalsIgnoreCase("ML"))
                                    uri = "MREML";
                                else if (prefix.equalsIgnoreCase("RE"))
                                    uri = "http://herry.com.cn";
                                else if (prefix.equalsIgnoreCase("RID"))
                                    uri = "ResourceID";
                                else if (prefix.equalsIgnoreCase("RUI"))
                                    uri= "ResourceUniqueID";
                                else if (prefix.equalsIgnoreCase("RMDI"))
                                    uri = "ResourceMetaDataInfo";
                                else
uri = null;
    映射。

    其中的ML和RMDI都是default namespace，它的作用范围包含它的子元素，直至有新的default namespace出现为止。而namespace只对它自身起作用。如将上述文档中：<ResourceMetaDataInfo>
            <ResourceID xmlns="ResourceID">
改为:
<RMDI:ResourceMetaDataInfo xmlns:RMDI=” ResourceMetaDataInfo”>
            <ResourceID>
…
…
</ RMDI:ResourceMetaDataInfo>
    查找ResourceUniqueID的值，则xpath应该改为：//ML：MREML/ML：ResourceEntity/RMDI：ResourceMetaDataInfo/ResourceID/RUI：ResourceUniqueID/text（）

    大家也可参考这篇文章：
http://blog.davber.com/2006/09/17/xpath-with-namespaces-in-java/

查看本文来源