// EntityResolver2.java - Extended SAX entity resolver.
// http://www.saxproject.org
// No warranty; no copyright -- use this as you will.
// $Id: EntityResolver2.java,v 1.1 2004/12/23 22:38:42 mark Exp $

package org.xml.sax.ext;

import java.io.IOException;

import org.xml.sax.EntityResolver;
import org.xml.sax.InputSource;
import org.xml.sax.XMLReader;
import org.xml.sax.SAXException;


/**
 * Extended interface for mapping external entity references to input
 * sources, or providing a missing external subset.  The
 * {@link XMLReader#setEntityResolver XMLReader.setEntityResolver()} method
 * is used to provide implementations of this interface to parsers.
 * When a parser uses the methods in this interface, the
 * {@link EntityResolver2#resolveEntity EntityResolver2.resolveEntity()}
 * method (in this interface) is used <em>instead of</em> the older (SAX 1.0)
 * {@link EntityResolver#resolveEntity EntityResolver.resolveEntity()} method.
 *
 * <blockquote>
 * <em>This module, both source code and documentation, is in the
 * Public Domain, and comes with <strong>NO WARRANTY</strong>.</em>
 * </blockquote>
 *
 * <p>If a SAX application requires the customized handling which this
 * interface defines for external entities, it must ensure that it uses
 * an XMLReader with the
 * <em>http://xml.org/sax/features/use-entity-resolver2</em> feature flag
 * set to <em>true</em> (which is its default value when the feature is
 * recognized).  If that flag is unrecognized, or its value is false,
 * or the resolver does not implement this interface, then only the
 * {@link EntityResolver} method will be used.
 * </p>
 *
 * <p>That supports three categories of application that modify entity
 * resolution.  <em>Old Style</em> applications won't know about this interface;
 * they will provide an EntityResolver.
 * <em>Transitional Mode</em> provide an EntityResolver2 and automatically
 * get the benefit of its methods in any systems (parsers or other tools)
 * supporting it, due to polymorphism.
 * Both <em>Old Style</em> and <em>Transitional Mode</em> applications will
 * work with any SAX2 parser.
 * <em>New style</em> applications will fail to run except on SAX2 parsers
 * that support this particular feature.
 * They will insist that feature flag have a value of "true", and the
 * EntityResolver2 implementation they provide  might throw an exception
 * if the original SAX 1.0 style entity resolution method is invoked.
 * </p>
 *
 * @see org.xml.sax.XMLReader#setEntityResolver
 *
 * @since SAX 2.0 (extensions 1.1 alpha)
 * @author David Brownell
 * @version TBD
 */
public interface EntityResolver2 extends EntityResolver
{
    /**
     * Allows applications to provide an external subset for documents
     * that don't explicitly define one.  Documents with DOCTYPE declarations
     * that omit an external subset can thus augment the declarations
     * available for validation, entity processing, and attribute processing
     * (normalization, defaulting, and reporting types including ID).
     * This augmentation is reported
     * through the {@link LexicalHandler#startDTD startDTD()} method as if
     * the document text had originally included the external subset;
     * this callback is made before any internal subset data or errors
     * are reported.</p>
     *
     * <p>This method can also be used with documents that have no DOCTYPE
     * declaration.  When the root element is encountered,
     * but no DOCTYPE declaration has been seen, this method is
     * invoked.  If it returns a value for the external subset, that root
     * element is declared to be the root element, giving the effect of
     * splicing a DOCTYPE declaration at the end the prolog of a document
     * that could not otherwise be valid.  The sequence of parser callbacks
     * in that case logically resembles this:</p>
     *
     * <pre>
     * ... comments and PIs from the prolog (as usual)
     * startDTD ("rootName", source.getPublicId (), source.getSystemId ());
     * startEntity ("[dtd]");
     * ... declarations, comments, and PIs from the external subset
     * endEntity ("[dtd]");
     * endDTD ();
     * ... then the rest of the document (as usual)
     * startElement (..., "rootName", ...);
     * </pre>
     *
     * <p>Note that the InputSource gets no further resolution.
     * Implementations of this method may wish to invoke
     * {@link #resolveEntity resolveEntity()} to gain benefits such as use
     * of local caches of DTD entities.  Also, this method will never be
     * used by a (non-validating) processor that is not including external
     * parameter entities. </p>
     *
     * <p>Uses for this method include facilitating data validation when
     * interoperating with XML processors that would always require
     * undesirable network accesses for external entities, or which for
     * other reasons adopt a "no DTDs" policy.
     * Non-validation motives include forcing documents to include DTDs so
     * that attributes are handled consistently.
     * For example, an XPath processor needs to know which attibutes have
     * type "ID" before it can process a widely used type of reference.</p>
     * 
     * <p><strong>Warning:</strong> Returning an external subset modifies
     * the input document.  By providing definitions for general entities,
     * it can make a malformed document appear to be well formed.
     * </p>
     *
     * @param name Identifies the document root element.  This name comes
     *	from a DOCTYPE declaration (where available) or from the actual
     *	root element. 
     * @param baseURI The document's base URI, serving as an additional
     *	hint for selecting the external subset.  This is always an absolute
     *	URI, unless it is null because the XMLReader was given an InputSource
     *	without one.
     *
     * @return An InputSource object describing the new external subset
     *	to be used by the parser, or null to indicate that no external
     *	subset is provided.
     *
     * @exception SAXException Any SAX exception, possibly wrapping
     *	another exception.
     * @exception IOException Probably indicating a failure to create
     *	a new InputStream or Reader, or an illegal URL.
     */
    public InputSource getExternalSubset (String name, String baseURI)
    throws SAXException, IOException;

    /**
     * Allows applications to map references to external entities into input
     * sources, or tell the parser it should use conventional URI resolution.
     * This method is only called for external entities which have been
     * properly declared.
     * This method provides more flexibility than the {@link EntityResolver}
     * interface, supporting implementations of more complex catalogue
     * schemes such as the one defined by the <a href=
	"http://www.oasis-open.org/committees/entity/spec-2001-08-06.html"
	>OASIS XML Catalogs</a> specification.</p>
     *
     * <p>Parsers configured to use this resolver method will call it
     * to determine the input source to use for any external entity
     * being included because of a reference in the XML text.
     * That excludes the document entity, and any external entity returned
     * by {@link #getExternalSubset getExternalSubset()}.
     * When a (non-validating) processor is configured not to include
     * a class of entities (parameter or general) through use of feature
     * flags, this method is not invoked for such entities.  </p>
     *
     * <p>Note that the entity naming scheme used here is the same one
     * used in the {@link LexicalHandler}, or in the {@link
	org.xml.sax.ContentHandler#skippedEntity
	ContentHandler.skippedEntity()}
     * method. </p>
     *
     * @param name Identifies the external entity being resolved.
     *	Either "[dtd]" for the external subset, or a name starting
     *	with "%" to indicate a parameter entity, or else the name of
     *	a general entity.  This is never null when invoked by a SAX2
     *	parser.
     * @param publicId The public identifier of the external entity being
     *	referenced (normalized as required by the XML specification), or
     *	null if none was supplied.
     * @param baseURI The URI with respect to which relative systemIDs
     *	are interpreted.  This is always an absolute URI, unless it is
     *	null (likely because the XMLReader was given an InputSource without
     *  one).  This URI is defined by the XML specification to be the one
     *	associated with the "&lt;" starting the relevant declaration.
     * @param systemId The system identifier of the external entity
     *	being referenced; either a relative or absolute URI.
     *  This is never null when invoked by a SAX2 parser; only declared
     *	entities, and any external subset, are resolved by such parsers.
     *
     * @return An InputSource object describing the new input source to
     *	be used by the parser.  Returning null directs the parser to
     *	resolve the system ID against the base URI and open a connection
     *	to resulting URI.
     *
     * @exception SAXException Any SAX exception, possibly wrapping
     *	another exception.
     * @exception IOException Probably indicating a failure to create
     *	a new InputStream or Reader, or an illegal URL.
     */
    public InputSource resolveEntity (
	    String name,
	    String publicId,
	    String baseURI,
	    String systemId
    ) throws SAXException, IOException;
}