Edit online

Configuring a Reference Resolver

This information is helpful if you need to provide a handler for resolving references and obtain the content they reference. For example, suppose the element that has references is ref and the attribute indicating the referenced resource is location. You need to implement a Java extension class for obtaining the referenced resources.

  1. Create the class simple.documentation.framework.ReferencesResolver. This class must implement the ro.sync.ecss.extensions.api.AuthorReferenceResolver interface.
    import ro.sync.ecss.extensions.api.AuthorReferenceResolver;
    import ro.sync.ecss.extensions.api.AuthorAccess;
    import ro.sync.ecss.extensions.api.node.AttrValue;
    import ro.sync.ecss.extensions.api.node.AuthorElement;
    import ro.sync.ecss.extensions.api.node.AuthorNode;
    
    public class ReferencesResolver 
          implements AuthorReferenceResolver {
  2. The hasReferences method verifies if the handler considers the node to have references. It takes AuthorNode as an argument that represents the node that will be verified. The method will return true if the node is considered to have references. In the following example, to be a reference, the node must be an element with the name ref and it must have an attribute named location.
    public boolean hasReferences(AuthorNode node) {
      boolean hasReferences = false;
      if (node.getType() == AuthorNode.NODE_TYPE_ELEMENT) {
        AuthorElement element = (AuthorElement) node;
        if ("ref".equals(element.getLocalName())) {
          AttrValue attrValue = element.getAttribute("location");
          hasReferences = attrValue != null;
        }
      }
      return hasReferences;
    }
  3. The method getDisplayName returns the display name of the node that contains the expanded referenced content. It takes AuthorNode as an argument that represents the node that needs the display name. The referenced content engine will ask this AuthorReferenceResolver implementation for the display name for each node that is considered a reference. In the following example, the display name is the value of the location attribute from the ref element.
    public String getDisplayName(AuthorNode node) {
      String displayName = "ref-fragment";
      if (node.getType() == AuthorNode.NODE_TYPE_ELEMENT) {
        AuthorElement element = (AuthorElement) node;
        if ("ref".equals(element.getLocalName())) {
          AttrValue attrValue = element.getAttribute("location");
          if (attrValue != null) {
            displayName = attrValue.getValue();
          }
        }
      }
      return displayName;
    }
  4. The method resolveReference resolves the reference of the node and returns a SAXSource with the parser and its input source. It takes AuthorNode as an argument that represents the node that needs the reference resolved, the systemID of the node, the AuthorAccess with access methods to the Author mode data model and a SAX EntityResolver that resolves resources that are already opened in another editor or resolve resources through the XML Catalog. In the implementation, you need to resolve the reference relative to the systemID, and create a parser and an input source over the resolved reference.
    public SAXSource resolveReference(
        AuthorNode node, 
        String systemID, 
        AuthorAccess authorAccess,
        EntityResolver  entityResolver) {
      SAXSource saxSource = null;
      
      if (node.getType() == AuthorNode.NODE_TYPE_ELEMENT) {
        AuthorElement element = (AuthorElement) node;
        if ("ref".equals(element.getLocalName())) {
          AttrValue attrValue = element.getAttribute("location");
          if (attrValue != null) {
            String attrStringVal = attrValue.getValue();
            try {
              URL absoluteUrl = new URL(new URL(systemID), 
              authorAccess.getUtilAccess().correctURL(attrStringVal));
              
              InputSource inputSource = entityResolver.resolveEntity(null, 
                  absoluteUrl.toString());
              if(inputSource == null) {
                inputSource = new InputSource(absoluteUrl.toString());
              }
              
              XMLReader xmlReader = authorAccess.newNonValidatingXMLReader();
              xmlReader.setEntityResolver(entityResolver);
              
              saxSource = new SAXSource(xmlReader, inputSource);
            } catch (MalformedURLException e) {
              logger.error(e, e);
            } catch (SAXException e) {
              logger.error(e, e);
            } catch (IOException e) {
              logger.error(e, e);
            }
          }
        }
      }
    
      return saxSource;
    }
  5. The method getReferenceUniqueID should return a unique identifier for the node reference. The unique identifier is used to avoid resolving the references recursively. The method takes AuthorNode as an argument that represents the node with the reference. In the following example, the unique identifier is the value of the location attribute from the ref element.
    public String getReferenceUniqueID(AuthorNode node) {
      String id = null;
      if (node.getType() == AuthorNode.NODE_TYPE_ELEMENT) {
        AuthorElement element = (AuthorElement) node;
        if ("ref".equals(element.getLocalName())) {
          AttrValue attrValue = element.getAttribute("location");
          if (attrValue != null) {
            id = attrValue.getValue();
          }
        }
      }
      return id;
    }
  6. The method getReferenceSystemIDshould return the systemID of the referenced content. It takes AuthorNode as an argument that represents the node with the reference and the AuthorAccess with access methods to the Author mode data model. For example, the value of the location attribute is used from the ref element and resolved relatively to the XML base URL of the node.
    public String getReferenceSystemID(AuthorNode node, 
                                       AuthorAccess authorAccess) {
      String systemID = null;
      if (node.getType() == AuthorNode.NODE_TYPE_ELEMENT) {
        AuthorElement element = (AuthorElement) node;
        if ("ref".equals(element.getLocalName())) {
          AttrValue attrValue = element.getAttribute("location");
          if (attrValue != null) {
            String attrStringVal = attrValue.getValue();
            try {
              URL absoluteUrl = new URL(node.getXMLBaseURL(), 
                  authorAccess.getUtilAccess().correctURL(attrStringVal));
              systemID = absoluteUrl.toString();
            } catch (MalformedURLException e) {
              logger.error(e, e);
            }
          }
        }
      }
      return systemID;
    }

    In the listing below, the XML document contains the ref element:

    <ref location="referenced.xml">Reference</ref>

When no reference resolver is specified, the reference has the following layout:

Figure 1. Reference with no specified reference resolver

When the above implementation is configured, the reference has the expected layout:

Figure 2. Reference with reference resolver
Note: The complete source code for framework customization examples can be found in the oxygen-sample-framework module of the Oxygen SDK, available as a Maven archetype on the Oxygen XML Editor website.