nokogiri-maven 1.5.0-java

data/ext/java/nokogiri/XmlDocumentFragment.java

@@ -0,0 +1,223 @@
+/**
+ * (The MIT License)
+ *
+ * Copyright (c) 2008 - 2011:
+ *
+ * * {Aaron Patterson}[http://tenderlovemaking.com]
+ * * {Mike Dalessio}[http://mike.daless.io]
+ * * {Charles Nutter}[http://blog.headius.com]
+ * * {Sergio Arbeo}[http://www.serabe.com]
+ * * {Patrick Mahoney}[http://polycrystal.org]
+ * * {Yoko Harada}[http://yokolet.blogspot.com]
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining
+ * a copy of this software and associated documentation files (the
+ * 'Software'), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sublicense, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ * 
+ * The above copyright notice and this permission notice shall be
+ * included in all copies or substantial portions of the Software.
+ * 
+ * THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+ * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+ * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+ * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+ * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+package nokogiri;
+
+import static nokogiri.internals.NokogiriHelpers.getLocalNameForNamespace;
+import static nokogiri.internals.NokogiriHelpers.getLocalPart;
+import static nokogiri.internals.NokogiriHelpers.getNokogiriClass;
+import static nokogiri.internals.NokogiriHelpers.getPrefix;
+import static nokogiri.internals.NokogiriHelpers.isNamespace;
+import static nokogiri.internals.NokogiriHelpers.rubyStringToString;
+
+import java.util.HashMap;
+import java.util.Map;
+import java.util.Set;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+
+import org.jruby.Ruby;
+import org.jruby.RubyArray;
+import org.jruby.RubyClass;
+import org.jruby.RubyString;
+import org.jruby.anno.JRubyClass;
+import org.jruby.anno.JRubyMethod;
+import org.jruby.javasupport.util.RuntimeHelpers;
+import org.jruby.runtime.ThreadContext;
+import org.jruby.runtime.builtin.IRubyObject;
+import org.w3c.dom.Attr;
+import org.w3c.dom.Document;
+import org.w3c.dom.NamedNodeMap;
+
+/**
+ * Class for Nokogiri::XML::DocumentFragment
+ * 
+ * @author sergio
+ * @author Yoko Harada <yokolet@gmail.com>
+ */
+@JRubyClass(name="Nokogiri::XML::DocumentFragment", parent="Nokogiri::XML::Node")
+public class XmlDocumentFragment extends XmlNode {
+    private XmlElement fragmentContext = null;
+
+    public XmlDocumentFragment(Ruby ruby) {
+        this(ruby, getNokogiriClass(ruby, "Nokogiri::XML::DocumentFragment"));
+    }
+
+    public XmlDocumentFragment(Ruby ruby, RubyClass klazz) {
+        super(ruby, klazz);
+    }
+
+    @JRubyMethod(name="new", meta = true, required=1, optional=2)
+    public static IRubyObject rbNew(ThreadContext context, IRubyObject cls, IRubyObject[] args) {
+        
+        if(args.length < 1) {
+            throw context.getRuntime().newArgumentError(args.length, 1);
+        }
+
+        if(!(args[0] instanceof XmlDocument)){
+            throw context.getRuntime().newArgumentError("first parameter must be a Nokogiri::XML::Document instance");
+        }
+
+        XmlDocument doc = (XmlDocument) args[0];
+        
+        // make wellformed fragment, ignore invalid namespace, or add appropriate namespace to parse
+        if (args.length > 1 && args[1] instanceof RubyString) {
+            args[1] = trim(context, doc, (RubyString)args[1]);
+            args[1] = RubyString.newString(context.getRuntime(), ignoreNamespaceIfNeeded(doc, rubyStringToString(args[1])));
+            args[1] = RubyString.newString(context.getRuntime(), addNamespaceDeclIfNeeded(doc, rubyStringToString(args[1])));
+        }
+
+        XmlDocumentFragment fragment = (XmlDocumentFragment) NokogiriService.XML_DOCUMENT_FRAGMENT_ALLOCATOR.allocate(context.getRuntime(), (RubyClass)cls);
+        fragment.setDocument(context, doc);
+        fragment.setNode(context, doc.getDocument().createDocumentFragment());
+
+        //TODO: Get namespace definitions from doc.
+        if (args.length == 3 && args[2] != null && args[2] instanceof XmlElement) {
+            fragment.fragmentContext = (XmlElement)args[2];
+        }
+        RuntimeHelpers.invoke(context, fragment, "initialize", args);
+        return fragment;
+    }
+    
+    private static IRubyObject trim(ThreadContext context, XmlDocument xmlDocument, RubyString str) {
+        // checks whether schema is given. if exists, allows whitespace processing to a parser
+        Document document = (Document)xmlDocument.node;
+        if (document.getDoctype() != null) return str;
+        // strips trailing \n off forcefully
+        // not to return new object in case of no chomp needed, chomp! is used here.
+        IRubyObject result;
+        if (context.getRuntime().is1_9()) result = str.chomp_bang19(context);
+        else result = str.chomp_bang(context);
+        return result.isNil() ? str : result;
+    }
+
+    private static Pattern qname_pattern = Pattern.compile("[^</:>\\s]+:[^</:>=\\s]+");
+    private static Pattern starttag_pattern = Pattern.compile("<[^</>]+>");
+    
+    private static String ignoreNamespaceIfNeeded(XmlDocument doc, String tags) {
+        if (doc.getDocument() == null) return tags;
+        if (doc.getDocument().getDocumentElement() == null) return tags;
+        Matcher matcher = qname_pattern.matcher(tags);
+        Map<String, String> rewriteTable = new HashMap<String, String>();
+        while(matcher.find()) {
+            String qName = matcher.group();
+            NamedNodeMap nodeMap = doc.getDocument().getDocumentElement().getAttributes();
+            if (!isNamespaceDefined(qName, nodeMap)) {
+                rewriteTable.put(qName, getLocalPart(qName));
+            }
+        }
+        Set<String> keys = rewriteTable.keySet();
+        for (String key : keys) {
+            tags = tags.replace(key, rewriteTable.get(key));
+        }
+        return tags;
+    }
+    
+    private static boolean isNamespaceDefined(String qName, NamedNodeMap nodeMap) {
+        if (isNamespace(qName.intern())) return true;
+        for (int i=0; i < nodeMap.getLength(); i++) {
+            Attr attr = (Attr)nodeMap.item(i);
+            if (isNamespace(attr.getNodeName())) {
+                String localPart = getLocalNameForNamespace(attr.getNodeName());
+                if (getPrefix(qName).equals(localPart)) {
+                    return true;
+                }
+            }
+        }
+        return false;
+    }
+    
+    private static String addNamespaceDeclIfNeeded(XmlDocument doc, String tags) {
+        if (doc.getDocument() == null) return tags;
+        if (doc.getDocument().getDocumentElement() == null) return tags;
+        Matcher matcher = starttag_pattern.matcher(tags);
+        Map<String, String> rewriteTable = new HashMap<String, String>();
+        while(matcher.find()) {
+            String start_tag = matcher.group();
+            Matcher matcher2 = qname_pattern.matcher(start_tag);
+            while(matcher2.find()) {
+                String qName = matcher2.group();
+                NamedNodeMap nodeMap = doc.getDocument().getDocumentElement().getAttributes();
+                if (isNamespaceDefined(qName, nodeMap)) {
+                    String namespaceDecl = getNamespceDecl(getPrefix(qName), nodeMap);
+                    if (namespaceDecl != null) {
+                        rewriteTable.put("<"+qName+">", "<"+qName + " " + namespaceDecl+">");
+                    }
+                }
+            }
+        }
+        Set<String> keys = rewriteTable.keySet();
+        for (String key : keys) {
+            tags = tags.replace(key, rewriteTable.get(key));
+        }
+        
+        return tags;
+    }
+    
+    private static String getNamespceDecl(String prefix, NamedNodeMap nodeMap) {
+        for (int i=0; i < nodeMap.getLength(); i++) {
+            Attr attr = (Attr)nodeMap.item(i);
+            if (prefix.equals(attr.getLocalName())) {
+                return attr.getName() + "=\"" + attr.getValue() + "\"";
+            }
+        }
+        return null;
+    }
+
+    public XmlElement getFragmentContext() {
+        return fragmentContext;
+    }
+
+    //@Override
+    public void add_child(ThreadContext context, XmlNode child) {
+        // Some magic for DocumentFragment
+
+        Ruby ruby = context.getRuntime();
+        XmlNodeSet children = (XmlNodeSet) child.children(context);
+
+        long length = children.length();
+
+        RubyArray childrenArray = children.convertToArray();
+
+        if(length != 0) {
+            for(int i = 0; i < length; i++) {
+                XmlNode item = (XmlNode) ((XmlNode) childrenArray.aref(ruby.newFixnum(i))).dup_implementation(context, true);
+                add_child(context, item);
+            }
+        }
+    }
+
+    @Override
+    public void relink_namespace(ThreadContext context) {
+        ((XmlNodeSet) children(context)).relink_namespace(context);
+    }
+}

data/ext/java/nokogiri/XmlDtd.java

@@ -0,0 +1,469 @@
+/**
+ * (The MIT License)
+ *
+ * Copyright (c) 2008 - 2011:
+ *
+ * * {Aaron Patterson}[http://tenderlovemaking.com]
+ * * {Mike Dalessio}[http://mike.daless.io]
+ * * {Charles Nutter}[http://blog.headius.com]
+ * * {Sergio Arbeo}[http://www.serabe.com]
+ * * {Patrick Mahoney}[http://polycrystal.org]
+ * * {Yoko Harada}[http://yokolet.blogspot.com]
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining
+ * a copy of this software and associated documentation files (the
+ * 'Software'), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sublicense, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ * 
+ * The above copyright notice and this permission notice shall be
+ * included in all copies or substantial portions of the Software.
+ * 
+ * THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+ * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+ * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+ * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+ * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+package nokogiri;
+
+import static nokogiri.internals.NokogiriHelpers.getNokogiriClass;
+import static nokogiri.internals.NokogiriHelpers.nonEmptyStringOrNil;
+import static nokogiri.internals.NokogiriHelpers.stringOrNil;
+import static org.jruby.javasupport.util.RuntimeHelpers.invoke;
+import nokogiri.internals.NokogiriHelpers;
+import nokogiri.internals.SaveContextVisitor;
+
+import org.apache.xerces.xni.QName;
+import org.cyberneko.dtd.DTDConfiguration;
+import org.jruby.Ruby;
+import org.jruby.RubyArray;
+import org.jruby.RubyClass;
+import org.jruby.RubyHash;
+import org.jruby.anno.JRubyClass;
+import org.jruby.anno.JRubyMethod;
+import org.jruby.runtime.ThreadContext;
+import org.jruby.runtime.builtin.IRubyObject;
+import org.w3c.dom.Document;
+import org.w3c.dom.DocumentType;
+import org.w3c.dom.Element;
+import org.w3c.dom.Node;
+
+/**
+ * Class for Nokogiri::XML::DTD
+ * 
+ * @author sergio
+ * @author Patrick Mahoney <pat@polycrystal.org>
+ * @author Yoko Harada <yokolet@gmail.com>
+ */
+
+@JRubyClass(name="Nokogiri::XML::DTD", parent="Nokogiri::XML::Node")
+public class XmlDtd extends XmlNode {
+    protected RubyArray allDecls = null;
+
+    /** cache of children, Nokogiri::XML::NodeSet */
+    protected IRubyObject children = null;
+
+    /** cache of name => XmlAttributeDecl */
+    protected RubyHash attributes = null;
+
+    /** cache of name => XmlElementDecl */
+    protected RubyHash elements = null;
+
+    /** cache of name => XmlEntityDecl */
+    protected RubyHash entities = null;
+
+    /** cache of name => Nokogiri::XML::Notation */
+    protected RubyHash notations = null;
+    protected RubyClass notationClass;
+
+    /** temporary store of content models before they are added to
+     * their XmlElementDecl. */
+    protected RubyHash contentModels;
+
+    /** node name */
+    protected IRubyObject name;
+
+    /** public ID (or external ID) */
+    protected IRubyObject pubId;
+
+    /** system ID */
+    protected IRubyObject sysId;
+
+    public XmlDtd(Ruby ruby, RubyClass rubyClass) {
+        super(ruby, rubyClass);
+    }
+    
+    public void setNode(Ruby runtime, Node dtd) {
+        this.node = dtd;
+        notationClass = (RubyClass) runtime.getClassFromPath("Nokogiri::XML::Notation");
+
+        name = pubId = sysId = runtime.getNil();
+        if (dtd == null) return;
+
+        // This is the dtd declaration stored in the document; it
+        // contains the DTD name (root element) and public and system
+        // ids. The actual declarations are in the NekoDTD 'dtd'
+        // variable. I don't know of a way to consolidate the two.
+
+        DocumentType otherDtd = dtd.getOwnerDocument().getDoctype();
+        if (otherDtd != null) {
+            name = stringOrNil(runtime, otherDtd.getNodeName());
+            pubId = nonEmptyStringOrNil(runtime, otherDtd.getPublicId());
+            sysId = nonEmptyStringOrNil(runtime, otherDtd.getSystemId());
+        }
+    }
+
+    public XmlDtd(Ruby ruby, RubyClass rubyClass, Node dtd) {
+        super(ruby, rubyClass, dtd);
+        setNode(ruby, dtd);
+    }
+
+    public static XmlDtd newEmpty(Ruby runtime,
+                                  Document doc,
+                                  IRubyObject name,
+                                  IRubyObject external_id,
+                                  IRubyObject system_id) {
+        Element placeHolder = doc.createElement("dtd_placeholder");
+        XmlDtd dtd = (XmlDtd) NokogiriService.XML_DTD_ALLOCATOR.allocate(runtime, getNokogiriClass(runtime, "Nokogiri::XML::DTD"));
+        dtd.setNode(runtime, placeHolder);
+        dtd.name = name;
+        dtd.pubId = external_id;
+        dtd.sysId = system_id;
+        return dtd;
+    }
+
+
+    /**
+     * Create an unparented element that contains DTD declarations
+     * parsed from the internal subset attached as user data to
+     * <code>doc</code>.  The attached dtd must be the tree from
+     * NekoDTD. The owner document of the returned tree will be
+     * <code>doc</doc>.
+     *
+     * NekoDTD parser returns a new document node containing elements
+     * representing the dtd declarations. The plan is to get the root
+     * element and adopt it into the correct document, stipping the
+     * Document provided by NekoDTD.
+     *
+     */
+    public static XmlDtd newFromInternalSubset(Ruby runtime, Document doc) {
+        Object dtdTree_ = doc.getUserData(XmlDocument.DTD_RAW_DOCUMENT);
+        if (dtdTree_ == null) {
+            XmlDtd xmlDtd = (XmlDtd) NokogiriService.XML_DTD_ALLOCATOR.allocate(runtime, getNokogiriClass(runtime, "Nokogiri::XML::DTD"));
+            xmlDtd.setNode(runtime, null);
+            return xmlDtd;
+        }
+
+        Node dtdTree = (Node) dtdTree_;
+        Node dtd = getInternalSubset(dtdTree);
+        if (dtd == null) {
+            XmlDtd xmlDtd = (XmlDtd) NokogiriService.XML_DTD_ALLOCATOR.allocate(runtime, getNokogiriClass(runtime, "Nokogiri::XML::DTD"));
+            xmlDtd.setNode(runtime, null);
+            return xmlDtd;
+        } else {
+            // Import the node into doc so it has the correct owner document.
+            dtd = doc.importNode(dtd, true);
+            XmlDtd xmlDtd = (XmlDtd) NokogiriService.XML_DTD_ALLOCATOR.allocate(runtime, getNokogiriClass(runtime, "Nokogiri::XML::DTD"));
+            xmlDtd.setNode(runtime, dtd);
+            return xmlDtd;
+        }
+    }
+
+    public static IRubyObject newFromExternalSubset(Ruby runtime, Document doc) {
+        Object dtdTree_ = doc.getUserData(XmlDocument.DTD_RAW_DOCUMENT);
+        if (dtdTree_ == null) {
+            return runtime.getNil();
+        }
+
+        Node dtdTree = (Node) dtdTree_;
+        Node dtd = getExternalSubset(dtdTree);
+        if (dtd == null) {
+            return runtime.getNil();
+        } else if (!dtd.hasChildNodes()) {
+            return runtime.getNil();
+        } else {
+            // Import the node into doc so it has the correct owner document.
+            dtd = doc.importNode(dtd, true);
+            XmlDtd xmlDtd = (XmlDtd) NokogiriService.XML_DTD_ALLOCATOR.allocate(runtime, getNokogiriClass(runtime, "Nokogiri::XML::DTD"));
+            xmlDtd.setNode(runtime, dtd);
+            return xmlDtd;
+        }
+    }
+
+    /*
+     * <code>dtd</code> is the document node of a NekoDTD tree.
+     * NekoDTD tree looks like this:
+     *
+     * <code><pre>
+     * [#document: null]
+     *   [#comment: ...]
+     *   [#comment: ...]
+     *   [dtd: null]   // a DocumentType; isDTD(node) => false
+     *   [dtd: null]   // root of dtd, an Element node; isDTD(node) => true
+     *     ... decls, content models, etc. ...
+     *     [externalSubset: null] pubid="the pubid" sysid="the sysid"
+     *       ... external subset decls, etc. ...
+     * </pre></code>
+     */
+    protected static Node getInternalSubset(Node dtdTree) {
+        Node root;
+        for (root = dtdTree.getFirstChild(); ; root = root.getNextSibling()) {
+            if (root == null)
+                return null;
+            else if (isDTD(root))
+                return root;      // we have second dtd which is root
+        }
+    }
+
+    protected static Node getExternalSubset(Node dtdTree) {
+        Node dtd = getInternalSubset(dtdTree);
+        if (dtd == null) return null;
+        for (Node ext = dtd.getFirstChild(); ; ext = ext.getNextSibling()) {
+            if (ext == null)
+                return null;
+            else if (isExternalSubset(ext))
+                return ext;
+        }
+    }
+
+    /**
+     * This overrides the #attributes method defined in
+     * lib/nokogiri/xml/node.rb.
+     */
+    @JRubyMethod
+    public IRubyObject attributes(ThreadContext context) {
+        if (attributes == null) extractDecls(context);
+
+        return attributes;
+    }
+
+    @JRubyMethod
+    public IRubyObject elements(ThreadContext context) {
+        if (elements == null) extractDecls(context);
+
+        return elements;
+    }
+
+    @JRubyMethod
+    public IRubyObject entities(ThreadContext context) {
+        if (entities == null) extractDecls(context);
+
+        return entities;
+    }
+
+    @JRubyMethod
+    public IRubyObject notations(ThreadContext context) {
+        if (notations == null) extractDecls(context);
+
+        return notations;
+    }
+
+    /**
+     * Our "node" object is as-returned by NekoDTD.  The actual
+     * "children" that we're interested in (Attribute declarations,
+     * etc.) are a few layers deep.
+     */
+    @Override
+    @JRubyMethod
+    public IRubyObject children(ThreadContext context) {
+        if (children == null) extractDecls(context);
+
+        return children;
+    }
+
+    /**
+     * Returns the name of the dtd.
+     */
+    @Override
+    @JRubyMethod
+    public IRubyObject node_name(ThreadContext context) {
+        return name;
+    }
+
+    @Override
+    @JRubyMethod(name = "node_name=")
+    public IRubyObject node_name_set(ThreadContext context, IRubyObject name) {
+        throw context.getRuntime()
+            .newRuntimeError("cannot change name of DTD");
+    }
+
+    @JRubyMethod
+    public IRubyObject system_id(ThreadContext context) {
+        return sysId;
+    }
+
+    @JRubyMethod
+    public IRubyObject external_id(ThreadContext context) {
+        return pubId;
+    }
+    
+    @JRubyMethod
+    public IRubyObject validate(ThreadContext context, IRubyObject doc) {
+        RubyArray errors = RubyArray.newArray(context.getRuntime());
+        if (doc instanceof XmlDocument) {
+          errors = (RubyArray) ((XmlDocument)doc).getInstanceVariable("@errors");
+        }
+        return errors;
+    }
+
+    public static boolean nameEquals(Node node, QName name) {
+        return name.localpart.equals(node.getNodeName());
+    }
+
+    public static boolean isExternalSubset(Node node) {
+        return nameEquals(node, DTDConfiguration.E_EXTERNAL_SUBSET);
+    }
+
+    /**
+     * Checks instanceof Element so we return false for a DocumentType
+     * node (NekoDTD uses Element for all its nodes).
+     */
+    public static boolean isDTD(Node node) {
+        return (node instanceof Element &&
+                nameEquals(node, DTDConfiguration.E_DTD));
+    }
+
+    public static boolean isAttributeDecl(Node node) {
+        return nameEquals(node, DTDConfiguration.E_ATTRIBUTE_DECL);
+    }
+
+    public static boolean isElementDecl(Node node) {
+        return nameEquals(node, DTDConfiguration.E_ELEMENT_DECL);
+    }
+
+    public static boolean isEntityDecl(Node node) {
+        return (nameEquals(node, DTDConfiguration.E_INTERNAL_ENTITY_DECL) ||
+                nameEquals(node, DTDConfiguration.E_UNPARSED_ENTITY_DECL));
+    }
+
+    public static boolean isNotationDecl(Node node) {
+        return nameEquals(node, DTDConfiguration.E_NOTATION_DECL);
+    }
+
+    public static boolean isContentModel(Node node) {
+        return nameEquals(node, DTDConfiguration.E_CONTENT_MODEL);
+    }
+
+    /**
+     * Recursively extract various DTD declarations and store them in
+     * the various collections.
+     */
+    protected void extractDecls(ThreadContext context) {
+        Ruby runtime = context.getRuntime();
+
+        // initialize data structures
+        allDecls = RubyArray.newArray(runtime);
+        attributes = RubyHash.newHash(runtime);
+        elements = RubyHash.newHash(runtime);
+        entities = RubyHash.newHash(runtime);
+        notations = RubyHash.newHash(runtime);
+        contentModels = RubyHash.newHash(runtime);
+        children = runtime.getNil();
+
+        // recursively extract decls
+        if (node == null) return; // leave all the decl hash's empty
+        extractDecls(context, node.getFirstChild());
+
+        // convert allDecls to a NodeSet
+        children = XmlNodeSet.newXmlNodeSet(context, allDecls);
+
+        // add attribute decls as attributes to the matching element decl
+        RubyArray keys = attributes.keys();
+        for (int i = 0; i < keys.getLength(); ++i) {
+            IRubyObject akey = keys.entry(i);
+            IRubyObject val;
+
+            val = attributes.op_aref(context, akey);
+            if (val.isNil()) continue;
+            XmlAttributeDecl attrDecl = (XmlAttributeDecl) val;
+            IRubyObject ekey = attrDecl.element_name(context);
+            val = elements.op_aref(context, ekey);
+            if (val.isNil()) continue;
+            XmlElementDecl elemDecl = (XmlElementDecl) val;
+
+            elemDecl.appendAttrDecl(attrDecl);
+        }
+
+        // add content models to the matching element decl
+        keys = contentModels.keys();
+        for (int i = 0; i < keys.getLength(); ++i) {
+            IRubyObject key = keys.entry(i);
+            IRubyObject cm = contentModels.op_aref(context, key);
+
+            IRubyObject elem = elements.op_aref(context, key);
+            if (elem.isNil()) continue;
+            if (((XmlElementDecl)elem).isEmpty()) continue;
+            ((XmlElementDecl) elem).setContentModel(cm);
+        }
+    }
+
+    /**
+     * The <code>node</code> is either the first child of the root dtd
+     * node (as returned by getInternalSubset()) or the first child of
+     * the external subset node (as returned by getExternalSubset()).
+     *
+     * This recursive function will not descend into an
+     * 'externalSubset' node, thus for an internal subset it only
+     * extracts nodes in the internal subset, and for an external
+     * subset it extracts everything and assumess <code>node</code>
+     * and all children are part of the external subset.
+     */
+    protected void extractDecls(ThreadContext context, Node node) {
+        while (node != null) {
+            if (isExternalSubset(node)) {
+                return;
+            } else if (isAttributeDecl(node)) {
+                XmlAttributeDecl decl = (XmlAttributeDecl)
+                    XmlAttributeDecl.create(context, node);
+                attributes.op_aset(context, decl.attribute_name(context), decl);
+                allDecls.append(decl);
+            } else if (isElementDecl(node)) {
+                XmlElementDecl decl = (XmlElementDecl)
+                    XmlElementDecl.create(context, node);
+                elements.op_aset(context, decl.element_name(context), decl);
+                allDecls.append(decl);
+            } else if (isEntityDecl(node)) {
+                XmlEntityDecl decl = (XmlEntityDecl)
+                    XmlEntityDecl.create(context, node);
+                entities.op_aset(context, decl.node_name(context), decl);
+                allDecls.append(decl);
+            } else if (isNotationDecl(node)) {
+                XmlNode tmp = (XmlNode)
+                    NokogiriHelpers.constructNode(context.getRuntime(), node);
+                IRubyObject decl = invoke(context, notationClass, "new",
+                                          tmp.getAttribute(context, "name"),
+                                          tmp.getAttribute(context, "pubid"),
+                                          tmp.getAttribute(context, "sysid"));
+                notations.op_aset(context,
+                                  tmp.getAttribute(context, "name"), decl);
+                allDecls.append(decl);
+            } else if (isContentModel(node)) {
+                XmlElementContent cm =
+                    new XmlElementContent(context.getRuntime(),
+                                          (XmlDocument) document(context),
+                                          node);
+                contentModels.op_aset(context, cm.element_name(context), cm);
+            } else {
+                // recurse
+                extractDecls(context, node.getFirstChild());
+            }
+
+            node = node.getNextSibling();
+        }
+    }
+    
+    @Override
+    public void accept(ThreadContext context, SaveContextVisitor visitor) {
+        // since we use nekoDTD to parse dtd, node might be ElementImpl type
+        // An external subset doesn't need to show up, so this method just see docType.
+        DocumentType docType = node.getOwnerDocument().getDoctype();
+        visitor.enter(docType);
+        visitor.leave(docType);
+    }
+}