package org.orbeon.dom.saxon;
import org.orbeon.dom.*;
import org.orbeon.saxon.Configuration;
import org.orbeon.saxon.event.Receiver;
import org.orbeon.saxon.om.*;
import org.orbeon.saxon.pattern.AnyNodeTest;
import org.orbeon.saxon.pattern.NodeTest;
import org.orbeon.saxon.trans.XPathException;
import org.orbeon.saxon.type.Type;
import org.orbeon.saxon.value.AtomicValue;
import org.orbeon.saxon.value.StringValue;
import org.orbeon.saxon.value.UntypedAtomicValue;
import org.orbeon.saxon.value.Value;
import java.util.*;
import java.util.ListIterator;
/**
* A node in the XML parse tree representing an XML element, character content, or attribute.
* This is the implementation of the NodeInfo interface used as a wrapper for Orbeon DOM nodes.
*
* History: this started life as the NodeWrapper for JDOM nodes; it was then modified by the
* Orbeon team to act as a wrapper for DOM4J nodes, and was shipped with the Orbeon product;
* it has now been absorbed back into Saxon. And it is now back to its own life
*
* @author Michael H. Kay
* @author Orbeon
*/
public class NodeWrapper implements NodeInfo, VirtualNode, SiblingCountingNode {
protected Node node;
private NodeWrapper parent; // null means unknown
DocumentWrapper docWrapper;
// Beware: this is an index over the result of content(), which may contain Namespace nodes
protected int index; // -1 means unknown
/**
* This constructor is protected: nodes should be created using the wrap
* factory method on the DocumentWrapper class
*
* @param node The node to be wrapped
* @param parent The NodeWrapper that wraps the parent of this node
* @param index Position of this node among its siblings
*/
protected NodeWrapper(Node node, NodeWrapper parent, int index) {
this.node = node;
this.parent = parent;
this.index = index;
}
/**
* Factory method to wrap a node with a wrapper that implements the Saxon
* NodeInfo interface.
*
* @param node The node
* @param docWrapper The wrapper for the Document containing this node
* @return The new wrapper for the supplied node
*/
NodeWrapper makeWrapper(Node node, DocumentWrapper docWrapper) {
return makeWrapper(node, docWrapper, null, -1);
}
/**
* Factory method to wrap a node with a wrapper that implements the Saxon
* NodeInfo interface.
*
* @param node The node
* @param docWrapper The wrapper for the Document containing this node
* @param parent The wrapper for the parent of the node
* @param index The position of this node relative to its siblings
* @return The new wrapper for the supplied node
*/
protected NodeWrapper makeWrapper(Node node, DocumentWrapper docWrapper, NodeWrapper parent, int index) {
return makeWrapperImpl(node, docWrapper, parent, index);
}
static NodeWrapper makeWrapperImpl(Node node, DocumentWrapper docWrapper, NodeWrapper parent, int index) {
if (node instanceof Document) {
return docWrapper;
} else {
final NodeWrapper wrapper = new NodeWrapper(node, parent, index);
wrapper.docWrapper = docWrapper;
return wrapper;
}
}
public Object getUnderlyingNode() {
return node;
}
public NamePool getNamePool() {
return docWrapper.getNamePool();
}
public int getNodeKind() {
// ORBEON: Should profile to see whether this is called often compared with the previous
// implementation where the node type was directly stored into each underlying node.
// Try to go from most frequently used to least
if (node instanceof Element)
return Type.ELEMENT;
else if (node instanceof Attribute)
return Type.ATTRIBUTE;
else if (node instanceof Text)
return Type.TEXT;
else if (node instanceof Document)
return Type.DOCUMENT;
else if (node instanceof Comment)
return Type.COMMENT;
else if (node instanceof ProcessingInstruction)
return Type.PROCESSING_INSTRUCTION;
else if (node instanceof Namespace)
return Type.NAMESPACE;
else
throw new IllegalStateException();
}
public SequenceIterator getTypedValue() throws XPathException {
return SingletonIterator.makeIterator((AtomicValue) atomize());
}
public Value atomize() throws XPathException {
if (node instanceof Comment || node instanceof ProcessingInstruction)
return new StringValue(getStringValueCS());
else
return new UntypedAtomicValue(getStringValueCS());
}
// UNTYPED or UNTYPED_ATOMIC
public int getTypeAnnotation() {
if (node instanceof Attribute) {
return StandardNames.XS_UNTYPED_ATOMIC;
}
return StandardNames.XS_UNTYPED;
}
public String getSystemId() {
return docWrapper.baseURI;
}
public void setSystemId(String uri) {
docWrapper.baseURI = uri;
}
// In this model, base URIs are held only an the document level. We don't currently take any account of xml:base attributes.
public String getBaseURI() {
if (node instanceof Namespace) {
return null;
}
NodeInfo n = this;
if (!(node instanceof Element)) {
n = n.getParent();
}
// Look for an xml:base attribute
while (n != null) {
String xmlbase = n.getAttributeValue(StandardNames.XML_BASE);
if (xmlbase != null) {
return xmlbase;
}
n = n.getParent();
}
// if not found, return the base URI of the document node
return docWrapper.baseURI;
}
public int getLineNumber() {
return -1;
}
public int getColumnNumber() {
return -1;
}
public int compareOrder(NodeInfo other) {
return Navigator.compareOrder(this, (SiblingCountingNode) other);
}
public String getStringValue() {
return getStringValue(node);
}
public CharSequence getStringValueCS() {
return getStringValue(node);
}
private static String getStringValue(Node node) {
return node.getStringValue();
}
public int getNameCode() {
switch (getNodeKind()) {
case Type.ELEMENT:
case Type.ATTRIBUTE:
case Type.PROCESSING_INSTRUCTION:
case Type.NAMESPACE:
return docWrapper.getNamePool().allocate(getPrefix(), getURI(), getLocalPart());
default:
return -1;
}
}
public int getFingerprint() {
return getNameCode() & 0xfffff;
}
public String getLocalPart() {
switch (getNodeKind()) {
case Type.ELEMENT:
return node.getName();
case Type.ATTRIBUTE:
return node.getName();
case Type.TEXT:
case Type.COMMENT:
case Type.DOCUMENT:
return "";
case Type.PROCESSING_INSTRUCTION:
return ((ProcessingInstruction) node).getTarget();
case Type.NAMESPACE:
return ((Namespace) node).prefix();
default:
return null;
}
}
public String getPrefix() {
if (node instanceof Element)
return ((Element) node).getNamespacePrefix();
else if (node instanceof Attribute)
return ((Attribute) node).getNamespacePrefix();
else
return "";
}
public String getURI() {
if (node instanceof Element)
return ((Element) node).getNamespaceURI();
else if (node instanceof Attribute)
return ((Attribute) node).getNamespaceURI();
else
return "";
}
public String getDisplayName() {
switch (getNodeKind()) {
case Type.ELEMENT:
return ((Element) node).getQualifiedName();
case Type.ATTRIBUTE:
return ((Attribute) node).getQualifiedName();
case Type.PROCESSING_INSTRUCTION:
case Type.NAMESPACE:
return getLocalPart();
default:
return "";
}
}
public NodeInfo getParent() {
if (parent == null) {
if (node instanceof Element) {
if (((Element) node).isRootElement()) {
parent = makeWrapper(node.getDocument(), docWrapper);
} else {
final Node parentNode = node.getParent();
// This checks the case of an element detached from a Document
if (parentNode != null)
parent = makeWrapper(parentNode, docWrapper);
}
} else if (node instanceof Text) {
parent = makeWrapper(node.getParent(), docWrapper);
} else if (node instanceof Comment) {
parent = makeWrapper(node.getParent(), docWrapper);
} else if (node instanceof ProcessingInstruction) {
parent = makeWrapper(node.getParent(), docWrapper);
} else if (node instanceof Attribute) {
parent = makeWrapper(node.getParent(), docWrapper);
} else if (node instanceof Document) {
parent = null;
} else if (node instanceof Namespace) {
throw new UnsupportedOperationException("Cannot find parent of a Namespace node");
} else {
throw new IllegalStateException();
}
}
return parent;
}
// Get the index position of this node among its siblings (starting from 0)
public int getSiblingPosition() {
if (index == -1) {
int ix = 0;
getParent();
AxisIterator iter;
switch (getNodeKind()) {
case Type.ELEMENT:
case Type.TEXT:
case Type.COMMENT:
case Type.PROCESSING_INSTRUCTION: {
final NodeWrapper parent = (NodeWrapper) getParent();
final List<Node> children;
if (parent.getNodeKind() == Type.DOCUMENT) {
// This is an attempt to work around a DOM4J bug
// ORBEON: What bug was that? Can we remove this and fix the issue in org.orbeon.dom?
final Document document = (Document) parent.node;
final List<Node> content = document.content();
if (content.size() == 0 && document.getRootElement() != null)
children = Collections.<Node>singletonList(document.getRootElement());
else
children = content;
} else {
// Beware: content() contains Namespace nodes (which is broken)!
children = ((Element) parent.node).content();
}
for (ListIterator iterator = children.listIterator(); iterator.hasNext(); ) {
final Object n = iterator.next();
if (n == node) {
index = ix;
return index;
}
ix++;
}
throw new IllegalStateException("DOM node not linked to parent node");
}
case Type.ATTRIBUTE:
iter = parent.iterateAxis(Axis.ATTRIBUTE);
break;
case Type.NAMESPACE:
iter = parent.iterateAxis(Axis.NAMESPACE);
break;
default:
index = 0;
return index;
}
while (true) {
NodeInfo n = (NodeInfo) iter.next();
if (n == null) {
break;
}
if (n.isSameNodeInfo(this)) {
index = ix;
return index;
}
ix++;
}
throw new IllegalStateException("DOM node not linked to parent node");
}
return index;
}
public AxisIterator iterateAxis(byte axisNumber) {
return iterateAxis(axisNumber, AnyNodeTest.getInstance());
}
public AxisIterator iterateAxis(byte axisNumber, NodeTest nodeTest) {
final int nodeKind = getNodeKind();
switch (axisNumber) {
case Axis.ANCESTOR:
if (nodeKind == Type.DOCUMENT) {
return EmptyIterator.getInstance();
}
return new Navigator.AxisFilter(
new Navigator.AncestorEnumeration(this, false),
nodeTest);
case Axis.ANCESTOR_OR_SELF:
if (nodeKind == Type.DOCUMENT) {
return Navigator.filteredSingleton(this, nodeTest);
}
return new Navigator.AxisFilter(
new Navigator.AncestorEnumeration(this, true),
nodeTest);
case Axis.ATTRIBUTE:
if (nodeKind != Type.ELEMENT) return EmptyIterator.getInstance();
return new Navigator.AxisFilter(
new AttributeEnumeration(this),
nodeTest);
case Axis.CHILD:
if (hasChildNodes()) {
return new Navigator.AxisFilter(
new ChildEnumeration(this, true, true),
nodeTest);
} else {
return EmptyIterator.getInstance();
}
case Axis.DESCENDANT:
if (hasChildNodes()) {
return new Navigator.AxisFilter(
new Navigator.DescendantEnumeration(this, false, true),
nodeTest);
} else {
return EmptyIterator.getInstance();
}
case Axis.DESCENDANT_OR_SELF:
return new Navigator.AxisFilter(
new Navigator.DescendantEnumeration(this, true, true),
nodeTest);
case Axis.FOLLOWING:
return new Navigator.AxisFilter(
new Navigator.FollowingEnumeration(this),
nodeTest);
case Axis.FOLLOWING_SIBLING:
switch (nodeKind) {
case Type.DOCUMENT:
case Type.ATTRIBUTE:
case Type.NAMESPACE:
return EmptyIterator.getInstance();
default:
return new Navigator.AxisFilter(
new ChildEnumeration(this, false, true),
nodeTest);
}
case Axis.NAMESPACE:
if (nodeKind != Type.ELEMENT) {
return EmptyIterator.getInstance();
}
return new Navigator.AxisFilter(
new NamespaceEnumeration(this),
nodeTest);
case Axis.PARENT:
getParent();
return Navigator.filteredSingleton(parent, nodeTest);
case Axis.PRECEDING:
return new Navigator.AxisFilter(
new Navigator.PrecedingEnumeration(this, false),
nodeTest);
case Axis.PRECEDING_SIBLING:
switch (nodeKind) {
case Type.DOCUMENT:
case Type.ATTRIBUTE:
case Type.NAMESPACE:
return EmptyIterator.getInstance();
default:
return new Navigator.AxisFilter(
new ChildEnumeration(this, false, false),
nodeTest);
}
case Axis.SELF:
return Navigator.filteredSingleton(this, nodeTest);
case Axis.PRECEDING_OR_ANCESTOR:
return new Navigator.AxisFilter(
new Navigator.PrecedingEnumeration(this, true),
nodeTest);
default:
throw new IllegalArgumentException("Unknown axis number " + axisNumber);
}
}
public String getAttributeValue(int fingerprint) {
if (node instanceof Element) {
Iterator list = ((Element) node).attributes().iterator();
NamePool pool = docWrapper.getNamePool();
while (list.hasNext()) {
Attribute att = (Attribute) list.next();
int nameCode = pool.allocate(att.getNamespacePrefix(),
att.getNamespaceURI(),
att.getName());
if (fingerprint == (nameCode & 0xfffff)) {
return att.getValue();
}
}
}
return null;
}
public NodeInfo getRoot() {
return docWrapper;
}
public DocumentInfo getDocumentRoot() {
return docWrapper;
}
public boolean hasChildNodes() {
if (node instanceof Document) {
return true;
} else if (node instanceof Element) {
// Beware: content() contains Namespace nodes (which is broken)!
List<Node> content = ((Element) node).content();
for (int i = 0; i < content.size(); i++) {
if (!(content.get(i) instanceof Namespace)) {
return true;
}
}
return false;
} else {
return false;
}
}
public void generateId(FastStringBuffer buffer) {
Navigator.appendSequentialKey(this, buffer, true);
}
public int getDocumentNumber() {
// NOTE: We used to call getParent().getDocumentNumber(), but all other implementations use
// docWrapper.getDocumentNumber() so we now harmonize with them.
// This also has another benefit: if a node gets detached from its parent, and getParent() has not yet been
// cached, getParent() can return null and getDocumentNumber() fails. By using docWrapper.getDocumentNumber()
// we avoid this issue, although arguably 1) a detached node should not point back to a DocumentWrapper and 2)
// one should not keep using a NodeInfo created to a node which is then detached.
return docWrapper.getDocumentNumber();
}
public void copy(Receiver out, int whichNamespaces, boolean copyAnnotations, int locationId) throws XPathException {
Navigator.copy(this, out, docWrapper.getNamePool(), whichNamespaces, copyAnnotations, locationId);
}
public boolean isId() {
return false;
}
public boolean isIdref() {
return false;
}
public boolean isNilled() {
return false;
}
///////////////////////////////////////////////////////////////////////////////
// Axis enumeration classes
///////////////////////////////////////////////////////////////////////////////
private final class AttributeEnumeration extends Navigator.BaseEnumeration {
private Iterator<Attribute> atts;
private int ix = 0;
private NodeWrapper start;
AttributeEnumeration(NodeWrapper start) {
this.start = start;
atts = ((Element) start.node).attributes().iterator();
}
public void advance() {
if (atts.hasNext()) {
current = makeWrapper(atts.next(), docWrapper, start, ix++);
} else {
current = null;
}
}
public SequenceIterator getAnother() {
return new AttributeEnumeration(start);
}
}
private final class NamespaceEnumeration extends Navigator.BaseEnumeration {
private HashMap<String, Namespace> nslist = new HashMap<String, Namespace>();
private Iterator<String> prefixes;
private int ix = 0;
private NodeWrapper start;
NamespaceEnumeration(NodeWrapper start) {
this.start = start;
NodeWrapper curr = start;
// build the complete list of namespaces
do {
final Element elem = (Element) curr.node;
final Namespace ns = elem.getNamespace();
final String prefix = ns.prefix();
final String uri = ns.uri();
if (!(prefix.length() == 0 && uri.length() == 0)) {
if (!nslist.containsKey(prefix)) {
nslist.put(prefix, ns);
}
}
final List<Namespace> additionalNamespaces = elem.additionalNamespaces();
if (!additionalNamespaces.isEmpty()) {
for (final Namespace additionalNamespace : additionalNamespaces) {
if (!nslist.containsKey(additionalNamespace.prefix())) {
nslist.put(additionalNamespace.prefix(), additionalNamespace);
}
}
}
curr = (NodeWrapper) curr.getParent();
} while (curr != null && curr.getNodeKind() == Type.ELEMENT);// NOTE: support elements detached from document
nslist.put("xml", Namespace$.MODULE$.XMLNamespace());
prefixes = nslist.keySet().iterator();
}
public void advance() {
if (prefixes.hasNext()) {
final String prefix = prefixes.next();
final Namespace ns = nslist.get(prefix);
current = makeWrapper(ns, docWrapper, start, ix++);
} else {
current = null;
}
}
public SequenceIterator getAnother() {
return new NamespaceEnumeration(start);
}
// NB: namespace nodes in the implementation do not support all
// XPath functions, for example namespace nodes have no parent.
}
/**
* The class ChildEnumeration handles not only the child axis, but also the
* following-sibling and preceding-sibling axes. It can also iterate the children
* of the start node in reverse order, something that is needed to support the
* preceding and preceding-or-ancestor axes (the latter being used by xsl:number)
*/
private final class ChildEnumeration extends Navigator.BaseEnumeration {
private NodeWrapper start;
private NodeWrapper commonParent;
private ListIterator<Node> children;
private int ix = 0;
private boolean downwards; // iterate children of start node (not siblings)
private boolean forwards; // iterate in document order (not reverse order)
ChildEnumeration(NodeWrapper start, boolean downwards, boolean forwards) {
this.start = start;
this.downwards = downwards;
this.forwards = forwards;
if (downwards) {
commonParent = start;
} else {
commonParent = (NodeWrapper) start.getParent();
}
if (commonParent.getNodeKind() == Type.DOCUMENT) {
// This is an attempt to work around a DOM4J bug
// ORBEON: What bug was that? Can we remove this and fix the issue in org.orbeon.dom?
final Document document = (Document) commonParent.node;
final List<Node> content = document.content();
if (content.size() == 0 && document.getRootElement() != null)
children = Collections.<Node>singletonList(document.getRootElement()).listIterator();
else
children = content.listIterator();
} else {
children = ((Element) commonParent.node).content().listIterator();
}
if (downwards) {
if (!forwards) {
// backwards enumeration: go to the end
while (children.hasNext()) {
children.next();
ix++;
}
}
} else {
ix = start.getSiblingPosition();
// find the start node among the list of siblings
if (forwards) {
for (int i = 0; i <= ix; i++) {
children.next();
}
ix++;
} else {
for (int i = 0; i < ix; i++) {
children.next();
}
ix--;
}
}
//advance();
}
public void advance() {
if (forwards) {
if (children.hasNext()) {
Node nextChild = children.next();
if (nextChild instanceof Namespace) {
ix++; // increment anyway so that makeWrapper() passes the correct index)
advance();
return;
}
// if (isAtomizing()) {
// current = new UntypedAtomicValue(getStringValue(node));
// } else {
current = makeWrapper(nextChild, docWrapper, commonParent, ix++);
// }
} else {
current = null;
}
} else { // backwards
if (children.hasPrevious()) {
Node nextChild = children.previous();
if (nextChild instanceof Namespace) {
ix--; // decrement anyway so that makeWrapper() passes the correct index)
advance();
return;
}
// if (isAtomizing()) {
// current = new UntypedAtomicValue(getStringValue(node));
// } else {
current = makeWrapper(nextChild, docWrapper, commonParent, ix--);
// }
} else {
current = null;
}
}
}
public SequenceIterator getAnother() {
return new ChildEnumeration(start, downwards, forwards);
}
}
public boolean isSameNodeInfo(NodeInfo other) {
if (other instanceof NodeWrapper) {
final NodeWrapper otherWrapper = (NodeWrapper) other;
if (node instanceof Namespace) {
final Namespace thisNamespace = (Namespace) node;
if (otherWrapper.node instanceof Namespace) {
final Namespace otherNamespace = (Namespace) otherWrapper.node;
// `Namespace` doesn't have a parent, but when `Namespace` is wrapped within `NodeWrapper`
// a parent is set on the wrapper, so we can compare the parents' identity.
return thisNamespace.prefix().equals(otherNamespace.prefix()) && getParent().isSameNodeInfo(otherWrapper.getParent());
} else {
return false;
}
} else {
// This check that `this.node eq other.node`
return node == ((NodeWrapper) other).node;
}
} else {
return false;
}
}
public Configuration getConfiguration() {
return docWrapper.getConfiguration();
}
/**
* Get all namespace undeclarations and undeclarations defined on this element.
*
* @param buffer If this is non-null, and the result array fits in this buffer, then the result
* may overwrite the contents of this array, to avoid the cost of allocating a new array on the heap.
* @return An array of integers representing the namespace declarations and undeclarations present on
* this element. For a node other than an element, return null. Otherwise, the returned array is a
* sequence of namespace codes, whose meaning may be interpreted by reference to the name pool. The
* top half word of each namespace code represents the prefix, the bottom half represents the URI.
* If the bottom half is zero, then this is a namespace undeclaration rather than a declaration.
* The XML namespace is never included in the list. If the supplied array is larger than required,
* then the first unused entry will be set to -1.
*
* For a node other than an element, the method returns null.
*/
public int[] getDeclaredNamespaces(int[] buffer) {
if (node instanceof Element) {
final Element elem = (Element) node;
final List namespaces = elem.declaredNamespaces();
if (namespaces == null || namespaces.isEmpty()) {
return EMPTY_NAMESPACE_LIST;
}
final int count = namespaces.size();
if (count == 0) {
return EMPTY_NAMESPACE_LIST;
} else {
int[] result = (buffer == null || count > buffer.length ? new int[count] : buffer);
NamePool pool = getNamePool();
int n = 0;
for (Iterator i = namespaces.iterator(); i.hasNext(); ) {
final Namespace namespace = (Namespace) i.next();
final String prefix = namespace.prefix();
final String uri = namespace.uri();
result[n++] = pool.allocateNamespaceCode(prefix, uri);
}
if (count < result.length) {
result[count] = -1;
}
return result;
}
} else {
return null;
}
}
public boolean equals(Object other) {
if (other instanceof NodeInfo) {
return isSameNodeInfo((NodeInfo) other);
} else {
return false;
}
}
public int hashCode() {
return node.hashCode();
}
}