All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.jvnet.staxex.util.DOMStreamReader Maven / Gradle / Ivy

There is a newer version: 2.1.0
Show newest version
/*
 * Copyright (c) 1997, 2018 Oracle and/or its affiliates. All rights reserved.
 *
 * This program and the accompanying materials are made available under the
 * terms of the Eclipse Distribution License v. 1.0, which is available at
 * http://www.eclipse.org/org/documents/edl-v10.php.
 *
 * SPDX-License-Identifier: BSD-3-Clause
 */

package org.jvnet.staxex.util;

import org.w3c.dom.Attr;
import org.w3c.dom.Element;
import org.w3c.dom.NamedNodeMap;
import org.w3c.dom.Node;
import static org.w3c.dom.Node.*;
import org.w3c.dom.ProcessingInstruction;
import org.w3c.dom.Text;

import javax.xml.namespace.NamespaceContext;
import javax.xml.namespace.QName;
import javax.xml.stream.Location;
import javax.xml.stream.XMLStreamException;
import javax.xml.stream.XMLStreamReader;
import java.util.Collections;
import java.util.Iterator;

/**
 * Create an {@link XMLStreamReader} on top of a DOM tree.
 *
 * 

* Since various libraries as well as users often create "incorrect" DOM node, * this class spends a lot of efforts making sure that broken DOM trees are * nevertheless interpreted correctly. * *

* For example, if a DOM level * 1 tree is passed, each method will attempt to return the correct value * by using {@link Node#getNodeName()}. * *

* Similarly, if DOM is missing explicit namespace declarations, * this class attempts to emulate necessary declarations. * * * @author [email protected] * @author Kohsuke Kawaguchi */ public class DOMStreamReader implements XMLStreamReader, NamespaceContext { /** * Current DOM node being traversed. */ protected Node _current; /** * Starting node of the subtree being traversed. */ private Node _start; /** * Named mapping for attributes and NS decls for the current node. */ private NamedNodeMap _namedNodeMap; /** * If the reader points at {@link #CHARACTERS the text node}, * its whole value. * *

* This is simply a cache of {@link Text#getWholeText()} of {@link #_current}, * but when a large binary data sent as base64 text, this could get very much * non-trivial. */ protected String wholeText; /** * List of attributes extracted from _namedNodeMap. */ private final FinalArrayList _currentAttributes = new FinalArrayList(); /** * {@link Scope} buffer. */ protected Scope[] scopes = new Scope[8]; /** * Depth of the current element. The first element gets depth==0. * Also used as the index to {@link #scopes}. */ protected int depth = 0; /** * State of this reader. Any of the valid states defined in StAX' * XMLStreamConstants class. */ protected int _state; /** * Namespace declarations on one element. * * Instances are reused. */ protected static final class Scope { /** * Scope for the parent element. */ final Scope parent; /** * List of namespace declarations extracted from _namedNodeMap */ final FinalArrayList currentNamespaces = new FinalArrayList(); /** * Additional namespace declarations obtained as a result of "fixing" DOM tree, * which were not part of the original DOM tree. * * One entry occupies two spaces (prefix followed by URI.) */ final FinalArrayList additionalNamespaces = new FinalArrayList(); Scope(Scope parent) { this.parent = parent; } void reset() { currentNamespaces.clear(); additionalNamespaces.clear(); } int getNamespaceCount() { return currentNamespaces.size()+additionalNamespaces.size()/2; } String getNamespacePrefix(int index) { int sz = currentNamespaces.size(); if(index< sz) { Attr attr = currentNamespaces.get(index); String result = attr.getLocalName(); if (result == null) { result = QName.valueOf(attr.getNodeName()).getLocalPart(); } return result.equals("xmlns") ? null : result; } else { return additionalNamespaces.get((index-sz)*2); } } String getNamespaceURI(int index) { int sz = currentNamespaces.size(); if(index< sz) { return currentNamespaces.get(index).getValue(); } else { return additionalNamespaces.get((index-sz)*2+1); } } /** * Returns the prefix bound to the given URI, or null. * This method recurses to the parent. */ String getPrefix(String nsUri) { for( Scope sp=this; sp!=null; sp=sp.parent ) { for( int i=sp.currentNamespaces.size()-1; i>=0; i--) { String result = getPrefixForAttr(sp.currentNamespaces.get(i),nsUri); if(result!=null) return result; } for( int i=sp.additionalNamespaces.size()-2; i>=0; i-=2 ) if(sp.additionalNamespaces.get(i+1).equals(nsUri)) return sp.additionalNamespaces.get(i); } return null; } /** * Returns the namespace URI bound by the given prefix. * * @param prefix * Prefix to look up. */ String getNamespaceURI(String prefix) { String nsDeclName = prefix.length()==0 ? "xmlns" : "xmlns:"+prefix; for( Scope sp=this; sp!=null; sp=sp.parent ) { for( int i=sp.currentNamespaces.size()-1; i>=0; i--) { Attr a = sp.currentNamespaces.get(i); if(a.getNodeName().equals(nsDeclName)) return a.getValue(); } for( int i=sp.additionalNamespaces.size()-2; i>=0; i-=2 ) if(sp.additionalNamespaces.get(i).equals(prefix)) return sp.additionalNamespaces.get(i+1); } return null; } } public DOMStreamReader() { } public DOMStreamReader(Node node) { setCurrentNode(node); } public void setCurrentNode(Node node) { scopes[0] = new Scope(null); depth=0; _start = _current = node; _state = START_DOCUMENT; // verifyDOMIntegrity(node); // displayDOM(node, System.out); } public void close() throws XMLStreamException { } /** * Called when the current node is {@link Element} to look at attribute list * (which contains both ns decl and attributes in DOM) and split them * to attributes-proper and namespace decls. */ protected void splitAttributes() { // Clear attribute and namespace lists _currentAttributes.clear(); Scope scope = allocateScope(); _namedNodeMap = _current.getAttributes(); if (_namedNodeMap != null) { final int n = _namedNodeMap.getLength(); for (int i = 0; i < n; i++) { final Attr attr = (Attr) _namedNodeMap.item(i); final String attrName = attr.getNodeName(); if (attrName.startsWith("xmlns:") || attrName.equals("xmlns")) { // NS decl? scope.currentNamespaces.add(attr); } else { _currentAttributes.add(attr); } } } // verify that all the namespaces used in element and attributes are indeed available ensureNs(_current); for( int i=_currentAttributes.size()-1; i>=0; i-- ) { Attr a = _currentAttributes.get(i); if(fixNull(a.getNamespaceURI()).length()>0) ensureNs(a); // no need to declare "" for attributes in the default namespace } } /** * Sub-routine of {@link #splitAttributes()}. * *

* Makes sure that the namespace URI/prefix used in the given node is available, * and if not, declare it on the current scope to "fix" it. * * It's often common to create DOM trees without putting namespace declarations, * and this makes sure that such DOM tree will be properly marshalled. */ private void ensureNs(Node n) { String prefix = fixNull(n.getPrefix()); String uri = fixNull(n.getNamespaceURI()); Scope scope = scopes[depth]; String currentUri = scope.getNamespaceURI(prefix); if(prefix.length()==0) { currentUri = fixNull(currentUri); if(currentUri.equals(uri)) return; // declared correctly } else { if(currentUri!=null && currentUri.equals(uri)) return; // declared correctly } if(prefix.equals("xml") || prefix.equals("xmlns")) return; // implicitly declared namespaces // needs to be declared scope.additionalNamespaces.add(prefix); scope.additionalNamespaces.add(uri); } /** * Allocate new {@link Scope} for {@link #splitAttributes()}. */ private Scope allocateScope() { if(scopes.length==++depth) { Scope[] newBuf = new Scope[scopes.length*2]; System.arraycopy(scopes,0,newBuf,0,scopes.length); scopes = newBuf; } Scope scope = scopes[depth]; if(scope==null) { scope = scopes[depth] = new Scope(scopes[depth-1]); } else { scope.reset(); } return scope; } public int getAttributeCount() { if (_state == START_ELEMENT) return _currentAttributes.size(); throw new IllegalStateException("DOMStreamReader: getAttributeCount() called in illegal state"); } /** * Return an attribute's local name. Handle the case of DOM level 1 nodes. */ public String getAttributeLocalName(int index) { if (_state == START_ELEMENT) { String localName = _currentAttributes.get(index).getLocalName(); return (localName != null) ? localName : QName.valueOf(_currentAttributes.get(index).getNodeName()).getLocalPart(); } throw new IllegalStateException("DOMStreamReader: getAttributeLocalName() called in illegal state"); } /** * Return an attribute's qname. Handle the case of DOM level 1 nodes. */ public QName getAttributeName(int index) { if (_state == START_ELEMENT) { Node attr = _currentAttributes.get(index); String localName = attr.getLocalName(); if (localName != null) { String prefix = attr.getPrefix(); String uri = attr.getNamespaceURI(); return new QName(fixNull(uri), localName, fixNull(prefix)); } else { return QName.valueOf(attr.getNodeName()); } } throw new IllegalStateException("DOMStreamReader: getAttributeName() called in illegal state"); } public String getAttributeNamespace(int index) { if (_state == START_ELEMENT) { String uri = _currentAttributes.get(index).getNamespaceURI(); return fixNull(uri); } throw new IllegalStateException("DOMStreamReader: getAttributeNamespace() called in illegal state"); } public String getAttributePrefix(int index) { if (_state == START_ELEMENT) { String prefix = _currentAttributes.get(index).getPrefix(); return fixNull(prefix); } throw new IllegalStateException("DOMStreamReader: getAttributePrefix() called in illegal state"); } public String getAttributeType(int index) { if (_state == START_ELEMENT) { return "CDATA"; } throw new IllegalStateException("DOMStreamReader: getAttributeType() called in illegal state"); } public String getAttributeValue(int index) { if (_state == START_ELEMENT) { return _currentAttributes.get(index).getNodeValue(); } throw new IllegalStateException("DOMStreamReader: getAttributeValue() called in illegal state"); } public String getAttributeValue(String namespaceURI, String localName) { if (_state == START_ELEMENT) { if (_namedNodeMap != null) { Node attr = _namedNodeMap.getNamedItemNS(namespaceURI, localName); return attr != null ? attr.getNodeValue() : null; } return null; } throw new IllegalStateException("DOMStreamReader: getAttributeValue() called in illegal state"); } public String getCharacterEncodingScheme() { return null; } public String getElementText() throws javax.xml.stream.XMLStreamException { throw new RuntimeException("DOMStreamReader: getElementText() not implemented"); } public String getEncoding() { return null; } public int getEventType() { return _state; } /** * Return an element's local name. Handle the case of DOM level 1 nodes. */ public String getLocalName() { if (_state == START_ELEMENT || _state == END_ELEMENT) { String localName = _current.getLocalName(); return localName != null ? localName : QName.valueOf(_current.getNodeName()).getLocalPart(); } else if (_state == ENTITY_REFERENCE) { return _current.getNodeName(); } throw new IllegalStateException("DOMStreamReader: getAttributeValue() called in illegal state"); } public Location getLocation() { return DummyLocation.INSTANCE; } /** * Return an element's qname. Handle the case of DOM level 1 nodes. */ public javax.xml.namespace.QName getName() { if (_state == START_ELEMENT || _state == END_ELEMENT) { String localName = _current.getLocalName(); if (localName != null) { String prefix = _current.getPrefix(); String uri = _current.getNamespaceURI(); return new QName(fixNull(uri), localName, fixNull(prefix)); } else { return QName.valueOf(_current.getNodeName()); } } throw new IllegalStateException("DOMStreamReader: getName() called in illegal state"); } public NamespaceContext getNamespaceContext() { return this; } /** * Verifies the current state to see if we can return the scope, and do so * if appropriate. * * Used to implement a bunch of StAX API methods that have the same usage restriction. */ private Scope getCheckedScope() { if (_state == START_ELEMENT || _state == END_ELEMENT) { return scopes[depth]; } throw new IllegalStateException("DOMStreamReader: neither on START_ELEMENT nor END_ELEMENT"); } public int getNamespaceCount() { return getCheckedScope().getNamespaceCount(); } public String getNamespacePrefix(int index) { return getCheckedScope().getNamespacePrefix(index); } public String getNamespaceURI(int index) { return getCheckedScope().getNamespaceURI(index); } public String getNamespaceURI() { if (_state == START_ELEMENT || _state == END_ELEMENT) { String uri = _current.getNamespaceURI(); return fixNull(uri); } return null; } /** * This method is not particularly fast, but shouldn't be called very * often. If we start to use it more, we should keep track of the * NS declarations using a NamespaceContext implementation instead. */ public String getNamespaceURI(String prefix) { if (prefix == null) { throw new IllegalArgumentException("DOMStreamReader: getNamespaceURI(String) call with a null prefix"); } else if (prefix.equals("xml")) { return "http://www.w3.org/XML/1998/namespace"; } else if (prefix.equals("xmlns")) { return "http://www.w3.org/2000/xmlns/"; } // check scopes String nsUri = scopes[depth].getNamespaceURI(prefix); if(nsUri!=null) return nsUri; // then ancestors above start node Node node = findRootElement(); String nsDeclName = prefix.length()==0 ? "xmlns" : "xmlns:"+prefix; while (node.getNodeType() != DOCUMENT_NODE) { // Is ns declaration on this element? NamedNodeMap namedNodeMap = node.getAttributes(); Attr attr = (Attr) namedNodeMap.getNamedItem(nsDeclName); if (attr != null) return attr.getValue(); node = node.getParentNode(); } return null; } public String getPrefix(String nsUri) { if (nsUri == null) { throw new IllegalArgumentException("DOMStreamReader: getPrefix(String) call with a null namespace URI"); } else if (nsUri.equals("http://www.w3.org/XML/1998/namespace")) { return "xml"; } else if (nsUri.equals("http://www.w3.org/2000/xmlns/")) { return "xmlns"; } // check scopes String prefix = scopes[depth].getPrefix(nsUri); if(prefix!=null) return prefix; // then ancestors above start node Node node = findRootElement(); while (node.getNodeType() != DOCUMENT_NODE) { // Is ns declaration on this element? NamedNodeMap namedNodeMap = node.getAttributes(); for( int i=namedNodeMap.getLength()-1; i>=0; i-- ) { Attr attr = (Attr)namedNodeMap.item(i); prefix = getPrefixForAttr(attr,nsUri); if(prefix!=null) return prefix; } node = node.getParentNode(); } return null; } /** * Finds the root element node of the traversal. */ private Node findRootElement() { int type; Node node = _start; while ((type = node.getNodeType()) != DOCUMENT_NODE && type != ELEMENT_NODE) { node = node.getParentNode(); } return node; } /** * If the given attribute is a namespace declaration for the given namespace URI, * return its prefix. Otherwise null. */ private static String getPrefixForAttr(Attr attr, String nsUri) { String attrName = attr.getNodeName(); if (!attrName.startsWith("xmlns:") && !attrName.equals("xmlns")) return null; // not nsdecl if(attr.getValue().equals(nsUri)) { if(attrName.equals("xmlns")) return ""; String localName = attr.getLocalName(); return (localName != null) ? localName : QName.valueOf(attrName).getLocalPart(); } return null; } public Iterator getPrefixes(String nsUri) { // This is an incorrect implementation, // but AFAIK it's not used in the JAX-WS runtime String prefix = getPrefix(nsUri); if(prefix==null) return Collections.emptyList().iterator(); else return Collections.singletonList(prefix).iterator(); } public String getPIData() { if (_state == PROCESSING_INSTRUCTION) { return ((ProcessingInstruction) _current).getData(); } return null; } public String getPITarget() { if (_state == PROCESSING_INSTRUCTION) { return ((ProcessingInstruction) _current).getTarget(); } return null; } public String getPrefix() { if (_state == START_ELEMENT || _state == END_ELEMENT) { String prefix = _current.getPrefix(); return fixNull(prefix); } return null; } public Object getProperty(String str) throws IllegalArgumentException { return null; } public String getText() { if (_state == CHARACTERS) return wholeText; if(_state == CDATA || _state == COMMENT || _state == ENTITY_REFERENCE) return _current.getNodeValue(); throw new IllegalStateException("DOMStreamReader: getTextLength() called in illegal state"); } public char[] getTextCharacters() { return getText().toCharArray(); } public int getTextCharacters(int sourceStart, char[] target, int targetStart, int targetLength) throws XMLStreamException { String text = getText(); int copiedSize = Math.min(targetLength, text.length() - sourceStart); text.getChars(sourceStart, sourceStart + copiedSize, target, targetStart); return copiedSize; } public int getTextLength() { return getText().length(); } public int getTextStart() { if (_state == CHARACTERS || _state == CDATA || _state == COMMENT || _state == ENTITY_REFERENCE) { return 0; } throw new IllegalStateException("DOMStreamReader: getTextStart() called in illegal state"); } public String getVersion() { return null; } public boolean hasName() { return (_state == START_ELEMENT || _state == END_ELEMENT); } public boolean hasNext() throws javax.xml.stream.XMLStreamException { return (_state != END_DOCUMENT); } public boolean hasText() { if (_state == CHARACTERS || _state == CDATA || _state == COMMENT || _state == ENTITY_REFERENCE) { return getText().trim().length() > 0; } return false; } public boolean isAttributeSpecified(int param) { return false; } public boolean isCharacters() { return (_state == CHARACTERS); } public boolean isEndElement() { return (_state == END_ELEMENT); } public boolean isStandalone() { return true; } public boolean isStartElement() { return (_state == START_ELEMENT); } public boolean isWhiteSpace() { if (_state == CHARACTERS || _state == CDATA) return getText().trim().length()==0; return false; } private static int mapNodeTypeToState(int nodetype) { switch (nodetype) { case CDATA_SECTION_NODE: return CDATA; case COMMENT_NODE: return COMMENT; case ELEMENT_NODE: return START_ELEMENT; case ENTITY_NODE: return ENTITY_DECLARATION; case ENTITY_REFERENCE_NODE: return ENTITY_REFERENCE; case NOTATION_NODE: return NOTATION_DECLARATION; case PROCESSING_INSTRUCTION_NODE: return PROCESSING_INSTRUCTION; case TEXT_NODE: return CHARACTERS; default: throw new RuntimeException("DOMStreamReader: Unexpected node type"); } } public int next() throws XMLStreamException { while(true) { int r = _next(); switch (r) { case CHARACTERS: // if we are currently at text node, make sure that this is a meaningful text node. Node prev = _current.getPreviousSibling(); if(prev!=null && prev.getNodeType()==Node.TEXT_NODE) continue; // nope. this is just a continuation of previous text that should be invisible Text t = (Text)_current; wholeText = t.getWholeText(); if(wholeText.length()==0) continue; // nope. this is empty text. return CHARACTERS; case START_ELEMENT: splitAttributes(); return START_ELEMENT; default: return r; } } } protected int _next() throws XMLStreamException { Node child; switch (_state) { case END_DOCUMENT: throw new IllegalStateException("DOMStreamReader: Calling next() at END_DOCUMENT"); case START_DOCUMENT: // Don't skip document element if this is a fragment if (_current.getNodeType() == ELEMENT_NODE) { return (_state = START_ELEMENT); } child = _current.getFirstChild(); if (child == null) { return (_state = END_DOCUMENT); } else { _current = child; return (_state = mapNodeTypeToState(_current.getNodeType())); } case START_ELEMENT: child = _current.getFirstChild(); if (child == null) { return (_state = END_ELEMENT); } else { _current = child; return (_state = mapNodeTypeToState(_current.getNodeType())); } case END_ELEMENT: case CHARACTERS: case COMMENT: case CDATA: case ENTITY_REFERENCE: case PROCESSING_INSTRUCTION: if (_state == END_ELEMENT) depth--; // If at the end of this fragment, then terminate traversal if (_current == _start) { return (_state = END_DOCUMENT); } Node sibling = _current.getNextSibling(); if (sibling == null) { _current = _current.getParentNode(); // getParentNode() returns null for fragments _state = (_current == null || _current.getNodeType() == DOCUMENT_NODE) ? END_DOCUMENT : END_ELEMENT; return _state; } else { _current = sibling; return (_state = mapNodeTypeToState(_current.getNodeType())); } case DTD: case ATTRIBUTE: case NAMESPACE: default: throw new RuntimeException("DOMStreamReader: Unexpected internal state"); } } public int nextTag() throws javax.xml.stream.XMLStreamException { int eventType = next(); while (eventType == CHARACTERS && isWhiteSpace() || eventType == CDATA && isWhiteSpace() || eventType == SPACE || eventType == PROCESSING_INSTRUCTION || eventType == COMMENT) { eventType = next(); } if (eventType != START_ELEMENT && eventType != END_ELEMENT) { throw new XMLStreamException("DOMStreamReader: Expected start or end tag"); } return eventType; } public void require(int type, String namespaceURI, String localName) throws javax.xml.stream.XMLStreamException { if (type != _state) { throw new XMLStreamException("DOMStreamReader: Required event type not found"); } if (namespaceURI != null && !namespaceURI.equals(getNamespaceURI())) { throw new XMLStreamException("DOMStreamReader: Required namespaceURI not found"); } if (localName != null && !localName.equals(getLocalName())) { throw new XMLStreamException("DOMStreamReader: Required localName not found"); } } public boolean standaloneSet() { return true; } // -- Debugging ------------------------------------------------------ /* private static void displayDOM(Node node, java.io.OutputStream ostream) { try { System.out.println("\n====\n"); XmlUtil.newTransformer().transform( new DOMSource(node), new StreamResult(ostream)); System.out.println("\n====\n"); } catch (Exception e) { e.printStackTrace(); } } private static void verifyDOMIntegrity(Node node) { switch (node.getNodeType()) { case ELEMENT_NODE: case ATTRIBUTE_NODE: // DOM level 1? if (node.getLocalName() == null) { System.out.println("WARNING: DOM level 1 node found"); System.out.println(" -> node.getNodeName() = " + node.getNodeName()); System.out.println(" -> node.getNamespaceURI() = " + node.getNamespaceURI()); System.out.println(" -> node.getLocalName() = " + node.getLocalName()); System.out.println(" -> node.getPrefix() = " + node.getPrefix()); } if (node.getNodeType() == ATTRIBUTE_NODE) return; NamedNodeMap attrs = node.getAttributes(); for (int i = 0; i < attrs.getLength(); i++) { verifyDOMIntegrity(attrs.item(i)); } case DOCUMENT_NODE: NodeList children = node.getChildNodes(); for (int i = 0; i < children.getLength(); i++) { verifyDOMIntegrity(children.item(i)); } } } */ private static String fixNull(String s) { if(s==null) return ""; else return s; } }





© 2015 - 2024 Weber Informatics LLC | Privacy Policy