使用Solr Dataimporthandler索引XML属性值

Question

I need to index some XML documents which make use of namespaces, such as:

<pm:Kroot>
<pm:root>
    <pm:meta>
        <dc:id xmlns:dc="http://purl.org/dc/elements/1.1/">1</dc:id>
        <dc:source>
            <dc:source>
                <pm:link pm:description="Tele" pm:source="8326"/>
            </dc:source>
        </dc:source>
    </pm:meta>
</pm:root>
</pm:Kroot>

Now when I use the the below DataImport, Solr manages to get the ID but it fails to index the Attributes values:

<dataConfig>
 <dataSource  type="FileDataSource" encoding="UTF-8" />
 <document>
    <entity name="article"
            url="/sample.xml"
            processor="XPathEntityProcessor"
            stream="true"                               
            forEach="/Kroot/root" >


        <field column="id"              xpath="/Kroot/root/meta/id" />            
        <field column="news_id"    xpath="/Kroot/root/meta/source/source/link/@source" />
        <field column="news_name"  xpath="/Kroot/root/meta/source/source/link/@description" />

   </entity>
 </document>

By removing the attribute namespaces from the XML file, Solr manages to index all the data! Now I am searching for a solution and I cant find an explanation for this behavior. The Solr wiki says that in case of namespaces we should use only the attribute name without the namespace, just like I do. I am using Solr 4.1 btw.

Answer 1

您可以尝试例如/Kroot/root/meta/source/source/link[@description]