ohair@286: /*
alanb@368:  * Copyright (c) 1997, 2013, Oracle and/or its affiliates. All rights reserved.
ohair@286:  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
ohair@286:  *
ohair@286:  * This code is free software; you can redistribute it and/or modify it
ohair@286:  * under the terms of the GNU General Public License version 2 only, as
ohair@286:  * published by the Free Software Foundation.  Oracle designates this
ohair@286:  * particular file as subject to the "Classpath" exception as provided
ohair@286:  * by Oracle in the LICENSE file that accompanied this code.
ohair@286:  *
ohair@286:  * This code is distributed in the hope that it will be useful, but WITHOUT
ohair@286:  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
ohair@286:  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
ohair@286:  * version 2 for more details (a copy is included in the LICENSE file that
ohair@286:  * accompanied this code).
ohair@286:  *
ohair@286:  * You should have received a copy of the GNU General Public License version
ohair@286:  * 2 along with this work; if not, write to the Free Software Foundation,
ohair@286:  * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
ohair@286:  *
ohair@286:  * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
ohair@286:  * or visit www.oracle.com if you need additional information or have any
ohair@286:  * questions.
ohair@286:  */
ohair@286: 
ohair@286: package com.sun.tools.internal.ws.wsdl.parser;
ohair@286: 
ohair@286: import com.sun.istack.internal.NotNull;
alanb@368: import com.sun.tools.internal.ws.util.xml.XmlUtil;
ohair@286: import com.sun.tools.internal.ws.wscompile.ErrorReceiver;
ohair@286: import com.sun.tools.internal.ws.wscompile.WsimportOptions;
ohair@286: import com.sun.tools.internal.ws.wsdl.document.schema.SchemaConstants;
ohair@286: import com.sun.tools.internal.xjc.reader.internalizer.LocatorTable;
ohair@286: import com.sun.xml.internal.bind.marshaller.DataWriter;
ohair@286: import org.w3c.dom.Document;
ohair@286: import org.w3c.dom.Element;
ohair@286: import org.w3c.dom.NodeList;
ohair@286: import org.xml.sax.ContentHandler;
ohair@286: import org.xml.sax.*;
ohair@286: import org.xml.sax.helpers.XMLFilterImpl;
ohair@286: 
ohair@286: import javax.xml.parsers.DocumentBuilder;
ohair@286: import javax.xml.parsers.DocumentBuilderFactory;
ohair@286: import javax.xml.parsers.ParserConfigurationException;
ohair@286: import javax.xml.parsers.SAXParserFactory;
ohair@286: import javax.xml.transform.Transformer;
ohair@286: import javax.xml.transform.TransformerException;
ohair@286: import javax.xml.transform.TransformerFactory;
ohair@286: import javax.xml.transform.dom.DOMSource;
ohair@286: import javax.xml.transform.sax.SAXResult;
ohair@286: import java.io.IOException;
ohair@286: import java.io.InputStream;
ohair@286: import java.io.OutputStream;
ohair@286: import java.io.OutputStreamWriter;
ohair@286: import java.net.*;
ohair@286: import java.util.*;
ohair@286: 
ohair@286: /**
ohair@286:  * @author Vivek Pandey
ohair@286:  */
ohair@286: public class DOMForest {
ohair@286:     /**
ohair@286:      * To correctly feed documents to a schema parser, we need to remember
ohair@286:      * which documents (of the forest) were given as the root
ohair@286:      * documents, and which of them are read as included/imported
ohair@286:      * documents.
ohair@286:      * <p/>
ohair@286:      * <p/>
ohair@286:      * Set of system ids as strings.
ohair@286:      */
ohair@286:     protected final Set<String> rootDocuments = new HashSet<String>();
ohair@286: 
ohair@286:     /**
ohair@286:      * Contains wsdl:import(s)
ohair@286:      */
ohair@286:     protected final Set<String> externalReferences = new HashSet<String>();
ohair@286: 
ohair@286:     /**
ohair@286:      * actual data storage map&lt;SystemId,Document>.
ohair@286:      */
ohair@286:     protected final Map<String, Document> core = new HashMap<String, Document>();
ohair@286:     protected final ErrorReceiver errorReceiver;
ohair@286: 
ohair@286:     private final DocumentBuilder documentBuilder;
ohair@286:     private final SAXParserFactory parserFactory;
ohair@286: 
ohair@286:     /**
ohair@286:      * inlined schema elements inside wsdl:type section
ohair@286:      */
ohair@286:     protected final List<Element> inlinedSchemaElements = new ArrayList<Element>();
ohair@286: 
ohair@286: 
ohair@286:     /**
ohair@286:      * Stores location information for all the trees in this forest.
ohair@286:      */
ohair@286:     public final LocatorTable locatorTable = new LocatorTable();
ohair@286: 
ohair@286:     protected final EntityResolver entityResolver;
ohair@286:     /**
ohair@286:      * Stores all the outer-most &lt;jaxb:bindings> customizations.
ohair@286:      */
ohair@286:     public final Set<Element> outerMostBindings = new HashSet<Element>();
ohair@286: 
ohair@286:     /**
ohair@286:      * Schema language dependent part of the processing.
ohair@286:      */
ohair@286:     protected final InternalizationLogic logic;
ohair@286:     protected final WsimportOptions options;
ohair@286: 
ohair@286:     public DOMForest(InternalizationLogic logic, @NotNull EntityResolver entityResolver, WsimportOptions options, ErrorReceiver errReceiver) {
ohair@286:         this.options = options;
ohair@286:         this.entityResolver = entityResolver;
ohair@286:         this.errorReceiver = errReceiver;
ohair@286:         this.logic = logic;
ohair@286:         try {
alanb@368:             // secure xml processing can be switched off if input requires it
mkos@408:             boolean secureProcessingEnabled = options == null || !options.disableXmlSecurity;
alanb@368:             DocumentBuilderFactory dbf = XmlUtil.newDocumentBuilderFactory(secureProcessingEnabled);
ohair@286:             dbf.setNamespaceAware(true);
ohair@286:             this.documentBuilder = dbf.newDocumentBuilder();
ohair@286: 
alanb@368:             this.parserFactory = XmlUtil.newSAXParserFactory(secureProcessingEnabled);
ohair@286:             this.parserFactory.setNamespaceAware(true);
ohair@286:         } catch (ParserConfigurationException e) {
ohair@286:             throw new AssertionError(e);
ohair@286:         }
ohair@286:     }
ohair@286: 
ohair@286:     public List<Element> getInlinedSchemaElement() {
ohair@286:         return inlinedSchemaElements;
ohair@286:     }
ohair@286: 
ohair@286:     public @NotNull Document parse(InputSource source, boolean root) throws SAXException, IOException {
ohair@286:         if (source.getSystemId() == null)
ohair@286:             throw new IllegalArgumentException();
ohair@286:         return parse(source.getSystemId(), source, root);
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * Parses an XML at the given location (
ohair@286:      * and XMLs referenced by it) into DOM trees
ohair@286:      * and stores them to this forest.
ohair@286:      *
ohair@286:      * @return the parsed DOM document object.
ohair@286:      */
ohair@286:     public Document parse(String systemId, boolean root) throws SAXException, IOException{
ohair@286: 
ohair@286:         systemId = normalizeSystemId(systemId);
ohair@286: 
ohair@286:         InputSource is = null;
ohair@286: 
ohair@286:         // allow entity resolver to find the actual byte stream.
ohair@286:         is = entityResolver.resolveEntity(null, systemId);
ohair@286:         if (is == null)
ohair@286:             is = new InputSource(systemId);
ohair@286:         else {
ohair@286:             resolvedCache.put(systemId, is.getSystemId());
ohair@286:             systemId=is.getSystemId();
ohair@286:         }
ohair@286: 
ohair@286:         if (core.containsKey(systemId)) {
ohair@286:             // this document has already been parsed. Just ignore.
ohair@286:             return core.get(systemId);
ohair@286:         }
ohair@286: 
ohair@286:         if(!root)
ohair@286:             addExternalReferences(systemId);
ohair@286: 
ohair@286:         // but we still use the original system Id as the key.
ohair@286:         return parse(systemId, is, root);
ohair@286:     }
ohair@286:     protected Map<String,String> resolvedCache = new HashMap<String,String>();
ohair@286: 
ohair@286:     public Map<String,String> getReferencedEntityMap() {
ohair@286:         return resolvedCache;
ohair@286:     }
ohair@286:     /**
ohair@286:      * Parses the given document and add it to the DOM forest.
ohair@286:      *
ohair@286:      * @return null if there was a parse error. otherwise non-null.
ohair@286:      */
ohair@286:     private @NotNull Document parse(String systemId, InputSource inputSource, boolean root) throws SAXException, IOException{
ohair@286:         Document dom = documentBuilder.newDocument();
ohair@286: 
ohair@286:         systemId = normalizeSystemId(systemId);
ohair@286: 
ohair@286:         // put into the map before growing a tree, to
ohair@286:         // prevent recursive reference from causing infinite loop.
ohair@286:         core.put(systemId, dom);
ohair@286: 
ohair@286:         dom.setDocumentURI(systemId);
ohair@286:         if (root)
ohair@286:             rootDocuments.add(systemId);
ohair@286: 
ohair@286:         try {
ohair@286:             XMLReader reader = createReader(dom);
ohair@286: 
ohair@286:             InputStream is = null;
ohair@286:             if(inputSource.getByteStream() == null){
ohair@286:                 inputSource = entityResolver.resolveEntity(null, systemId);
ohair@286:             }
ohair@286:             reader.parse(inputSource);
ohair@286:             Element doc = dom.getDocumentElement();
ohair@286:             if (doc == null) {
ohair@286:                 return null;
ohair@286:             }
ohair@286:             NodeList schemas = doc.getElementsByTagNameNS(SchemaConstants.NS_XSD, "schema");
ohair@286:             for (int i = 0; i < schemas.getLength(); i++) {
ohair@286:                 inlinedSchemaElements.add((Element) schemas.item(i));
ohair@286:             }
ohair@286:         } catch (ParserConfigurationException e) {
ohair@286:             errorReceiver.error(e);
ohair@286:             throw new SAXException(e.getMessage());
ohair@286:         }
ohair@286:         resolvedCache.put(systemId, dom.getDocumentURI());
ohair@286:         return dom;
ohair@286:     }
ohair@286: 
ohair@286:     public void addExternalReferences(String ref) {
ohair@286:         if (!externalReferences.contains(ref))
ohair@286:             externalReferences.add(ref);
ohair@286:     }
ohair@286: 
ohair@286: 
ohair@286:     public Set<String> getExternalReferences() {
ohair@286:         return externalReferences;
ohair@286:     }
ohair@286: 
ohair@286: 
ohair@286: 
ohair@286:     public interface Handler extends ContentHandler {
ohair@286:         /**
ohair@286:          * Gets the DOM that was built.
ohair@286:          */
ohair@286:         public Document getDocument();
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:          * Returns a {@link org.xml.sax.XMLReader} to parse a document into this DOM forest.
ohair@286:          * <p/>
ohair@286:          * This version requires that the DOM object to be created and registered
ohair@286:          * to the map beforehand.
ohair@286:          */
ohair@286:     private XMLReader createReader(Document dom) throws SAXException, ParserConfigurationException {
ohair@286:         XMLReader reader = parserFactory.newSAXParser().getXMLReader();
ohair@286:         DOMBuilder dombuilder = new DOMBuilder(dom, locatorTable, outerMostBindings);
ohair@286:         try {
ohair@286:             reader.setProperty("http://xml.org/sax/properties/lexical-handler", dombuilder);
ohair@286:         } catch(SAXException e) {
ohair@286:             errorReceiver.debug(e.getMessage());
ohair@286:         }
ohair@286: 
ohair@286:         ContentHandler handler = new WhitespaceStripper(dombuilder, errorReceiver, entityResolver);
ohair@286:         handler = new VersionChecker(handler, errorReceiver, entityResolver);
ohair@286: 
ohair@286:         // insert the reference finder so that
ohair@286:         // included/imported schemas will be also parsed
ohair@286:         XMLFilterImpl f = logic.createExternalReferenceFinder(this);
ohair@286:         f.setContentHandler(handler);
ohair@286:         if (errorReceiver != null)
ohair@286:             f.setErrorHandler(errorReceiver);
ohair@286:         f.setEntityResolver(entityResolver);
ohair@286: 
ohair@286:         reader.setContentHandler(f);
ohair@286:         if (errorReceiver != null)
ohair@286:             reader.setErrorHandler(errorReceiver);
ohair@286:         reader.setEntityResolver(entityResolver);
ohair@286:         return reader;
ohair@286:     }
ohair@286: 
ohair@286:     private String normalizeSystemId(String systemId) {
ohair@286:         try {
ohair@286:             systemId = new URI(systemId).normalize().toString();
ohair@286:         } catch (URISyntaxException e) {
ohair@286:             // leave the system ID untouched. In my experience URI is often too strict
ohair@286:         }
ohair@286:         return systemId;
ohair@286:     }
ohair@286: 
ohair@286:     boolean isExtensionMode() {
ohair@286:         return options.isExtensionMode();
ohair@286:     }
ohair@286: 
ohair@286: 
ohair@286:     /**
ohair@286:      * Gets the DOM tree associated with the specified system ID,
ohair@286:      * or null if none is found.
ohair@286:      */
ohair@286:     public Document get(String systemId) {
ohair@286:         Document doc = core.get(systemId);
ohair@286: 
ohair@286:         if (doc == null && systemId.startsWith("file:/") && !systemId.startsWith("file://")) {
ohair@286:             // As of JDK1.4, java.net.URL.toExternal method returns URLs like
ohair@286:             // "file:/abc/def/ghi" which is an incorrect file protocol URL according to RFC1738.
ohair@286:             // Some other correctly functioning parts return the correct URLs ("file:///abc/def/ghi"),
ohair@286:             // and this descripancy breaks DOM look up by system ID.
ohair@286: 
ohair@286:             // this extra check solves this problem.
ohair@286:             doc = core.get("file://" + systemId.substring(5));
ohair@286:         }
ohair@286: 
ohair@286:         if (doc == null && systemId.startsWith("file:")) {
ohair@286:             // on Windows, filenames are case insensitive.
ohair@286:             // perform case-insensitive search for improved user experience
ohair@286:             String systemPath = getPath(systemId);
ohair@286:             for (String key : core.keySet()) {
ohair@286:                 if (key.startsWith("file:") && getPath(key).equalsIgnoreCase(systemPath)) {
ohair@286:                     doc = core.get(key);
ohair@286:                     break;
ohair@286:                 }
ohair@286:             }
ohair@286:         }
ohair@286: 
ohair@286:         return doc;
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * Strips off the leading 'file:///' portion from an URL.
ohair@286:      */
ohair@286:     private String getPath(String key) {
ohair@286:         key = key.substring(5); // skip 'file:'
ohair@286:         while (key.length() > 0 && key.charAt(0) == '/')
ohair@286:             key = key.substring(1);
ohair@286:         return key;
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * Gets all the system IDs of the documents.
ohair@286:      */
ohair@286:     public String[] listSystemIDs() {
ohair@286:         return core.keySet().toArray(new String[core.keySet().size()]);
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * Gets the system ID from which the given DOM is parsed.
ohair@286:      * <p/>
ohair@286:      * Poor-man's base URI.
ohair@286:      */
ohair@286:     public String getSystemId(Document dom) {
ohair@286:         for (Map.Entry<String, Document> e : core.entrySet()) {
ohair@286:             if (e.getValue() == dom)
ohair@286:                 return e.getKey();
ohair@286:         }
ohair@286:         return null;
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * Gets the first one (which is more or less random) in {@link #rootDocuments}.
ohair@286:      */
ohair@286:     public String getFirstRootDocument() {
ohair@286:         if(rootDocuments.isEmpty()) return null;
ohair@286:         return rootDocuments.iterator().next();
ohair@286:     }
ohair@286: 
ohair@286:     public Set<String> getRootDocuments() {
ohair@286:         return rootDocuments;
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * Dumps the contents of the forest to the specified stream.
ohair@286:      * <p/>
ohair@286:      * This is a debug method. As such, error handling is sloppy.
ohair@286:      */
ohair@286:     public void dump(OutputStream out) throws IOException {
ohair@286:         try {
ohair@286:             // create identity transformer
alanb@368:             // secure xml processing can be switched off if input requires it
mkos@408:             boolean secureProcessingEnabled = options == null || !options.disableXmlSecurity;
alanb@368:             TransformerFactory tf = XmlUtil.newTransformerFactory(secureProcessingEnabled);
alanb@368:             Transformer it = tf.newTransformer();
ohair@286: 
ohair@286:             for (Map.Entry<String, Document> e : core.entrySet()) {
ohair@286:                 out.write(("---<< " + e.getKey() + '\n').getBytes());
ohair@286: 
ohair@286:                 DataWriter dw = new DataWriter(new OutputStreamWriter(out), null);
ohair@286:                 dw.setIndentStep("  ");
ohair@286:                 it.transform(new DOMSource(e.getValue()),
ohair@286:                         new SAXResult(dw));
ohair@286: 
ohair@286:                 out.write("\n\n\n".getBytes());
ohair@286:             }
ohair@286:         } catch (TransformerException e) {
ohair@286:             e.printStackTrace();
ohair@286:         }
ohair@286:     }
ohair@286: 
ohair@286: }