Content Engine includes an XML Classification Plug-in, called the XML Classifier in Enterprise Manager. When properly configured, this XML plug-in automatically populates document object properties with values based on data extracted from an incoming XML document's content. Administrators build mapping scripts that associate XML tags in the incoming document to properties, thus enabling the automatic classification of any XML document.
XML Property Mapping Scripts are versionable objects that are based on the root Document class. They are managed by Enterprise Manager, and are available for subclassing from the Document Class\XML Property Mapping Script. The main purpose of the XML Property Mapping Scripts is to provide a container for user-authored XSL scripts that transform XML documents of a specified type. The script contains instructions for extracting content from the type of XML documents identified by the XML Document Type property. This script enables you to map the content onto property values for the properties associated with the target document class identified by the Target Class property.
Classification is exposed via the server API as a flag on Document.Checkin. This appears in the Create New Document Wizard exposed by the Enterprise Manager. The XML Classifier only operates on the first content element of a document; additional content elements are ignored.
If there is a schema reference present, the XML Classifier uses it to validate the document. If there is no schema, validation will not occur. In either case, the XML document must be well formed.
The XML Classifier lets you customize its behavior for specific XML document types by creating XML Property Mapping Scripts using the Create New XML Property Mapping Script Wizard. This wizard creates objects whose content elements are property-mapping scripts which are registered for a specific XML document type. The XML Classifier invokes the script when classifying a document of that type. The script identifies which properties on the classified document object are updated as part of the classification process and what the values are.
The important classification-related properties of the XML Property Mapping Script class are:
The XML Classifier does not support binary properties, multi-value properties, or object-valued properties.
The XML Classifier utilizes the XML features provided by the JDK and must be compliant with the XSLT Version 1.0 standards. Prior to P8 4.0, non-normative XSLT features might have worked, but they were not supported. If you used non-normative features in the .xsl for XML classification in a P8 3.x release, the .xsl might need to be updated to be compliant with the XSLT Version 1.0 standard.