net.sf.saxon.functions
Class StandardCollectionURIResolver

java.lang.Object
  extended by net.sf.saxon.functions.StandardCollectionURIResolver
All Implemented Interfaces:
Serializable, CollectionURIResolver

public class StandardCollectionURIResolver
extends Object
implements CollectionURIResolver

This class implements the default collection URI Resolver.

This supports two implementations of collections. If the URI supplied uses the "file:/" scheme, and the file that is referenced is a directory, then the collection is the set of files in that directory. Query parameters may be included in the URI:

Otherwise, the resolver attempts to dereference the URI to obtain a catalog file. This is an XML file containing a list of documents, in the format:

 <collection>
   <doc href="doc1.xml"/>
   <doc href="doc2.xml"/>
 </collection>
 

See Also:
Serialized Form

Constructor Summary
StandardCollectionURIResolver()
           
 
Method Summary
 SequenceIterator resolve(String href, String base, XPathContext context)
          Resolve a URI.
 
Methods inherited from class java.lang.Object
equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

StandardCollectionURIResolver

public StandardCollectionURIResolver()
Method Detail

resolve

public SequenceIterator resolve(String href,
                                String base,
                                XPathContext context)
                         throws XPathException
Resolve a URI.

Specified by:
resolve in interface CollectionURIResolver
Parameters:
href - The relative URI of the collection. This corresponds to the argument supplied to the collection() function. If the collection() function was called with no arguments (to get the "default collection") this argument will be null.
base - The base URI that should be used. This is the base URI of the static context in which the call to collection() was made, typically the URI of the stylesheet or query module
context - The dynamic execution context
Returns:
an Iterator over the documents in the collection. The items returned by this iterator must be instances either of xs:anyURI, or of node() (specifically, NodeInfo.). If xs:anyURI values are returned, the corresponding document will be retrieved as if by a call to the doc() function: this means that the system first checks to see if the document is already loaded, and if not, calls the registered URIResolver to dereference the URI. This is the recommended approach to ensure that the resulting collection is stable: however, it has the consequence that the documents will by default remain in memory for the duration of the query or transformation.
Throws:
XPathException


Copyright (C) Michael H. Kay. All rights reserved.