We need to provide a handler for resolving references and obtain the content they refer. In our case the element which has references is ref and the attribute indicating the referred resource is location. We will have to implement a Java extension class for obtaining the referred resources.
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 {
The method hasReferences
verifies if the
handler considers the node to have references. It takes as argument an
AuthorNode
that represents the node which
will be verified. The method will return true
if the node
is considered to have references. In our case, 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; }
The method getDisplayName
returns the
display name of the node that contains the expanded referred content.
It takes as argument an AuthorNode
that
represents the node for which the display name is needed. The referred
content engine will ask this
AuthorReferenceResolver
implementation what is the display name for each node which is
considered a reference. In our case 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; }
The method resolveReference
resolves the
reference of the node and returns a SAXSource
with the parser and the parser's input source. It takes as arguments
an AuthorNode
that represents the node for
which the reference needs resolving, the systemID
of the node, the AuthorAccess
with access
methods to the Author data model and a SAX
EntityResolver
which resolves resources
that are already opened in another editor or resolve resources through
the XML catalog. In our implementation we 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.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; }
The method getReferenceUniqueID
should return
an unique identifier for the node reference. The unique identifier is
used to avoid resolving the references recursively. It takes as
argument an AuthorNode
that represents the node
with the reference. In our implementation the unique identifier 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; }
The method getReferenceSystemID
should return
the systemID
of the referred content. It takes as
arguments an AuthorNode
that represents the
node with the reference and the AuthorAccess
with access methods to the Author data model. In our implementation we
use the value of the location attribute from the
ref element and resolve it 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.correctURL(attrStringVal)); systemID = absoluteUrl.toString(); } catch (MalformedURLException e) { logger.error(e, e); } } } } return systemID; }
The complete source code of our implementation is found in the Example Files Listings, the Java Files section.
In the listing below, the XML document contains the ref element:
<ref location="referred.xml">Reference</ref>
When no reference resolver is specified, the reference has the following layout:
When the above implementation is configured, the reference has the expected layout: