2 * Copyright 2005-2010 the original author or authors.
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
8 * http://www.apache.org/licenses/LICENSE-2.0
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
16 package org.wamblee.xml;
18 import java.io.ByteArrayInputStream;
19 import java.io.ByteArrayOutputStream;
20 import java.io.IOException;
21 import java.io.InputStream;
22 import java.io.OutputStream;
23 import java.util.ArrayList;
24 import java.util.List;
26 import java.util.TreeMap;
27 import java.util.logging.Level;
28 import java.util.logging.Logger;
30 import javax.xml.XMLConstants;
31 import javax.xml.parsers.DocumentBuilder;
32 import javax.xml.parsers.DocumentBuilderFactory;
33 import javax.xml.parsers.ParserConfigurationException;
34 import javax.xml.transform.Transformer;
35 import javax.xml.transform.TransformerException;
36 import javax.xml.transform.TransformerFactory;
37 import javax.xml.transform.dom.DOMSource;
38 import javax.xml.transform.stream.StreamResult;
39 import javax.xml.transform.stream.StreamSource;
40 import javax.xml.validation.Schema;
41 import javax.xml.validation.SchemaFactory;
42 import javax.xml.validation.Validator;
44 import org.w3c.dom.Attr;
45 import org.w3c.dom.Document;
46 import org.w3c.dom.Element;
47 import org.w3c.dom.NamedNodeMap;
48 import org.w3c.dom.Node;
49 import org.w3c.dom.NodeList;
50 import org.w3c.dom.bootstrap.DOMImplementationRegistry;
51 import org.w3c.dom.ls.DOMImplementationLS;
52 import org.w3c.dom.ls.LSException;
53 import org.w3c.dom.ls.LSInput;
54 import org.w3c.dom.ls.LSParser;
55 import org.xml.sax.SAXException;
58 * Some basic XML utilities for common reoccuring tasks for DOM documents.
60 * @author Erik Brakkee
62 public final class DomUtils {
63 private static final Logger LOG = Logger
64 .getLogger(DomUtils.class.getName());
67 * Disabled default constructor.
75 * Gets a dom level 3 implementation.
76 * @return Dom implementation.
77 * @throws ClassNotFoundException
78 * @throws InstantiationException
79 * @throws IllegalAccessException
81 public static DOMImplementationLS getDomImplementationLS() {
82 final String message = "Could not get Dom level 3 implementation";
84 DOMImplementationRegistry registry = DOMImplementationRegistry
87 DOMImplementationLS impl = (DOMImplementationLS) registry
88 .getDOMImplementation("LS");
90 } catch (ClassCastException e) {
91 throw new RuntimeException(message, e);
92 } catch (ClassNotFoundException e) {
93 throw new RuntimeException(message, e);
94 } catch (InstantiationException e) {
95 throw new RuntimeException(message, e);
96 } catch (IllegalAccessException e) {
97 throw new RuntimeException(message, e);
102 * Removes duplicate attributes from a DOM tree.This is useful for
103 * postprocessing the output of JTidy as a workaround for a bug in JTidy.
106 * Node to remove duplicate attributes from (recursively).
107 * Attributes of the node itself are not dealt with. Only the
108 * child nodes are dealt with.
110 public static void removeDuplicateAttributes(Node aNode) {
111 NodeList list = aNode.getChildNodes();
113 for (int i = 0; i < list.getLength(); i++) {
114 Node node = list.item(i);
116 if (node instanceof Element) {
117 removeDuplicateAttributes((Element) node);
118 removeDuplicateAttributes(node);
124 * Removes duplicate attributes from an element.
129 private static void removeDuplicateAttributes(Element aElement) {
130 NamedNodeMap attributes = aElement.getAttributes();
131 Map<String, Attr> uniqueAttributes = new TreeMap<String, Attr>();
132 List<Attr> attlist = new ArrayList<Attr>();
134 for (int i = 0; i < attributes.getLength(); i++) {
135 Attr attribute = (Attr) attributes.item(i);
137 if (uniqueAttributes.containsKey(attribute.getNodeName())) {
138 LOG.info("Detected duplicate attribute (will be removed)'" +
139 attribute.getNodeName() + "'");
142 uniqueAttributes.put(attribute.getNodeName(), attribute);
143 attlist.add(attribute);
146 // Remove all attributes from the element.
147 for (Attr att : attlist) {
148 aElement.removeAttributeNode(att);
151 // Add the unique attributes back to the element.
152 for (Attr att : uniqueAttributes.values()) {
153 aElement.setAttributeNode(att);