1 /*
2 * Copyright (c) 1997, 2013, Oracle and/or its affiliates. All rights reserved.
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 *
5 * This code is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU General Public License version 2 only, as
7 * published by the Free Software Foundation. Oracle designates this
8 * particular file as subject to the "Classpath" exception as provided
9 * by Oracle in the LICENSE file that accompanied this code.
10 *
11 * This code is distributed in the hope that it will be useful, but WITHOUT
12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 * version 2 for more details (a copy is included in the LICENSE file that
15 * accompanied this code).
16 *
17 * You should have received a copy of the GNU General Public License version
18 * 2 along with this work; if not, write to the Free Software Foundation,
19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
20 *
21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
22 * or visit www.oracle.com if you need additional information or have any
23 * questions.
24 */
25
26 package com.sun.tools.internal.xjc.reader.internalizer;
27
28 import java.io.IOException;
29 import java.io.OutputStream;
30 import java.io.OutputStreamWriter;
31 import java.util.ArrayList;
32 import java.util.Collections;
33 import java.util.HashMap;
34 import java.util.HashSet;
35 import java.util.List;
36 import java.util.Map;
37 import java.util.Set;
38
39 import static javax.xml.XMLConstants.W3C_XML_SCHEMA_NS_URI;
40 import javax.xml.parsers.DocumentBuilder;
41 import javax.xml.parsers.DocumentBuilderFactory;
42 import javax.xml.parsers.ParserConfigurationException;
43 import javax.xml.parsers.SAXParserFactory;
44 import javax.xml.stream.XMLStreamException;
45 import javax.xml.stream.XMLStreamReader;
46 import javax.xml.transform.Source;
47 import javax.xml.transform.Transformer;
48 import javax.xml.transform.TransformerException;
49 import javax.xml.transform.TransformerFactory;
50 import javax.xml.transform.dom.DOMSource;
51 import javax.xml.transform.sax.SAXResult;
52 import javax.xml.transform.sax.SAXSource;
53 import javax.xml.validation.SchemaFactory;
54
55 import com.sun.istack.internal.NotNull;
56 import com.sun.istack.internal.XMLStreamReaderToContentHandler;
57 import com.sun.tools.internal.xjc.ErrorReceiver;
58 import com.sun.tools.internal.xjc.Options;
59 import com.sun.tools.internal.xjc.reader.Const;
60 import com.sun.tools.internal.xjc.util.ErrorReceiverFilter;
61 import com.sun.xml.internal.bind.marshaller.DataWriter;
62 import com.sun.xml.internal.bind.v2.util.XmlFactory;
63 import com.sun.xml.internal.xsom.parser.JAXPParser;
64 import com.sun.xml.internal.xsom.parser.XMLParser;
65
66 import org.w3c.dom.Document;
67 import org.w3c.dom.Element;
68 import org.xml.sax.ContentHandler;
69 import org.xml.sax.EntityResolver;
70 import org.xml.sax.InputSource;
71 import org.xml.sax.SAXException;
72 import org.xml.sax.SAXParseException;
73 import org.xml.sax.XMLReader;
74 import org.xml.sax.helpers.XMLFilterImpl;
75
76
77 /**
78 * Builds a DOM forest and maintains association from
79 * system IDs to DOM trees.
80 *
81 * <p>
82 * A forest is a transitive reflexive closure of referenced documents.
83 * IOW, if a document is in a forest, all the documents referenced from
84 * it is in a forest, too. To support this semantics, {@link DOMForest}
85 * uses {@link InternalizationLogic} to find referenced documents.
86 *
87 * <p>
88 * Some documents are marked as "root"s, meaning those documents were
89 * put into a forest explicitly, not because it is referenced from another
90 * document. (However, a root document can be referenced from other
91 * documents, too.)
92 *
93 * @author
94 * Kohsuke Kawaguchi (kohsuke.kawaguchi@sun.com)
95 */
96 public final class DOMForest {
97 /** actual data storage map<SystemId,Document>. */
98 private final Map<String,Document> core = new HashMap<String,Document>();
99
100 /**
101 * To correctly feed documents to a schema parser, we need to remember
102 * which documents (of the forest) were given as the root
103 * documents, and which of them are read as included/imported
104 * documents.
105 *
106 * <p>
107 * Set of system ids as strings.
108 */
109 private final Set<String> rootDocuments = new HashSet<String>();
110
111 /** Stores location information for all the trees in this forest. */
112 public final LocatorTable locatorTable = new LocatorTable();
113
114 /** Stores all the outer-most <jaxb:bindings> customizations. */
115 public final Set<Element> outerMostBindings = new HashSet<Element>();
116
117 /** Used to resolve references to other schema documents. */
118 private EntityResolver entityResolver = null;
119
120 /** Errors encountered during the parsing will be sent to this object. */
121 private ErrorReceiver errorReceiver = null;
122
123 /** Schema language dependent part of the processing. */
124 protected final InternalizationLogic logic;
125
126 private final SAXParserFactory parserFactory;
127 private final DocumentBuilder documentBuilder;
128
129 private final Options options;
130
131 public DOMForest(
132 SAXParserFactory parserFactory, DocumentBuilder documentBuilder,
133 InternalizationLogic logic ) {
134
135 this.parserFactory = parserFactory;
136 this.documentBuilder = documentBuilder;
137 this.logic = logic;
138 this.options = null;
139 }
140
141 public DOMForest( InternalizationLogic logic, Options opt ) {
142
143 if (opt == null) throw new AssertionError("Options object null");
144 this.options = opt;
145
146 try {
147 DocumentBuilderFactory dbf = XmlFactory.createDocumentBuilderFactory(opt.disableXmlSecurity);
148 this.documentBuilder = dbf.newDocumentBuilder();
149 this.parserFactory = XmlFactory.createParserFactory(opt.disableXmlSecurity);
150 } catch( ParserConfigurationException e ) {
151 throw new AssertionError(e);
152 }
153
154 this.logic = logic;
155 }
156
157 /**
158 * Gets the DOM tree associated with the specified system ID,
159 * or null if none is found.
160 */
161 public Document get( String systemId ) {
162 Document doc = core.get(systemId);
163
164 if( doc==null && systemId.startsWith("file:/") && !systemId.startsWith("file://") ) {
165 // As of JDK1.4, java.net.URL.toExternal method returns URLs like
166 // "file:/abc/def/ghi" which is an incorrect file protocol URL according to RFC1738.
167 // Some other correctly functioning parts return the correct URLs ("file:///abc/def/ghi"),
168 // and this descripancy breaks DOM look up by system ID.
169
170 // this extra check solves this problem.
171 doc = core.get( "file://"+systemId.substring(5) );
172 }
173
174 if( doc==null && systemId.startsWith("file:") ) {
175 // on Windows, filenames are case insensitive.
176 // perform case-insensitive search for improved user experience
177 String systemPath = getPath(systemId);
178 for (String key : core.keySet()) {
179 if(key.startsWith("file:") && getPath(key).equalsIgnoreCase(systemPath)) {
180 doc = core.get(key);
181 break;
182 }
183 }
184 }
185
186 return doc;
187 }
188
189 /**
190 * Strips off the leading 'file:///' portion from an URL.
191 */
192 private String getPath(String key) {
193 key = key.substring(5); // skip 'file:'
194 while(key.length()>0 && key.charAt(0)=='/') {
195 key = key.substring(1);
196 }
197 return key;
198 }
199
200 /**
201 * Returns a read-only set of root document system IDs.
202 */
203 public Set<String> getRootDocuments() {
204 return Collections.unmodifiableSet(rootDocuments);
205 }
206
207 /**
208 * Picks one document at random and returns it.
209 */
210 public Document getOneDocument() {
211 for (Document dom : core.values()) {
212 if (!dom.getDocumentElement().getNamespaceURI().equals(Const.JAXB_NSURI))
213 return dom;
214 }
215 // we should have caught this error very early on
216 throw new AssertionError();
217 }
218
219 /**
220 * Checks the correctness of the XML Schema documents and return true
221 * if it's OK.
222 *
223 * <p>
224 * This method performs a weaker version of the tests where error messages
225 * are provided without line number information. So whenever possible
226 * use {@link SchemaConstraintChecker}.
227 *
228 * @see SchemaConstraintChecker
229 */
230 public boolean checkSchemaCorrectness(ErrorReceiver errorHandler) {
231 try {
232 boolean disableXmlSecurity = false;
233 if (options != null) {
234 disableXmlSecurity = options.disableXmlSecurity;
235 }
236 SchemaFactory sf = XmlFactory.createSchemaFactory(W3C_XML_SCHEMA_NS_URI, disableXmlSecurity);
237 ErrorReceiverFilter filter = new ErrorReceiverFilter(errorHandler);
238 sf.setErrorHandler(filter);
239 Set<String> roots = getRootDocuments();
240 Source[] sources = new Source[roots.size()];
241 int i=0;
242 for (String root : roots) {
243 sources[i++] = new DOMSource(get(root),root);
244 }
245 sf.newSchema(sources);
246 return !filter.hadError();
247 } catch (SAXException e) {
248 // the errors should have been reported
249 return false;
250 }
251 }
252
253 /**
254 * Gets the system ID from which the given DOM is parsed.
255 * <p>
256 * Poor-man's base URI.
257 */
258 public String getSystemId( Document dom ) {
259 for (Map.Entry<String,Document> e : core.entrySet()) {
260 if (e.getValue() == dom)
261 return e.getKey();
262 }
263 return null;
264 }
265
266 public Document parse( InputSource source, boolean root ) throws SAXException {
267 if( source.getSystemId()==null )
268 throw new IllegalArgumentException();
269
270 return parse( source.getSystemId(), source, root );
271 }
272
273 /**
274 * Parses an XML at the given location (
275 * and XMLs referenced by it) into DOM trees
276 * and stores them to this forest.
277 *
278 * @return the parsed DOM document object.
279 */
280 public Document parse( String systemId, boolean root ) throws SAXException, IOException {
281
282 systemId = Options.normalizeSystemId(systemId);
283
284 if( core.containsKey(systemId) )
285 // this document has already been parsed. Just ignore.
286 return core.get(systemId);
287
288 InputSource is=null;
289
290 // allow entity resolver to find the actual byte stream.
291 if( entityResolver!=null )
292 is = entityResolver.resolveEntity(null,systemId);
293 if( is==null )
294 is = new InputSource(systemId);
295
296 // but we still use the original system Id as the key.
297 return parse( systemId, is, root );
298 }
299
300 /**
301 * Returns a {@link ContentHandler} to feed SAX events into.
302 *
303 * <p>
304 * The client of this class can feed SAX events into the handler
305 * to parse a document into this DOM forest.
306 *
307 * This version requires that the DOM object to be created and registered
308 * to the map beforehand.
309 */
310 private ContentHandler getParserHandler( Document dom ) {
311 ContentHandler handler = new DOMBuilder(dom,locatorTable,outerMostBindings);
312 handler = new WhitespaceStripper(handler,errorReceiver,entityResolver);
313 handler = new VersionChecker(handler,errorReceiver,entityResolver);
314
315 // insert the reference finder so that
316 // included/imported schemas will be also parsed
317 XMLFilterImpl f = logic.createExternalReferenceFinder(this);
318 f.setContentHandler(handler);
319
320 if(errorReceiver!=null)
321 f.setErrorHandler(errorReceiver);
322 if(entityResolver!=null)
323 f.setEntityResolver(entityResolver);
324
325 return f;
326 }
327
328 public interface Handler extends ContentHandler {
329 /**
330 * Gets the DOM that was built.
331 */
332 public Document getDocument();
333 }
334
335 private static abstract class HandlerImpl extends XMLFilterImpl implements Handler {
336 }
337
338 /**
339 * Returns a {@link ContentHandler} to feed SAX events into.
340 *
341 * <p>
342 * The client of this class can feed SAX events into the handler
343 * to parse a document into this DOM forest.
344 */
345 public Handler getParserHandler( String systemId, boolean root ) {
346 final Document dom = documentBuilder.newDocument();
347 core.put( systemId, dom );
348 if(root)
349 rootDocuments.add(systemId);
350
351 ContentHandler handler = getParserHandler(dom);
352
353 // we will register the DOM to the map once the system ID becomes available.
354 // but the SAX allows the event source to not to provide that information,
355 // so be prepared for such case.
356 HandlerImpl x = new HandlerImpl() {
357 public Document getDocument() {
358 return dom;
359 }
360 };
361 x.setContentHandler(handler);
362
363 return x;
364 }
365
366 /**
367 * Parses the given document and add it to the DOM forest.
368 *
369 * @return
370 * null if there was a parse error. otherwise non-null.
371 */
372 public Document parse( String systemId, InputSource inputSource, boolean root ) throws SAXException {
373 Document dom = documentBuilder.newDocument();
374
375 systemId = Options.normalizeSystemId(systemId);
376
377 // put into the map before growing a tree, to
378 // prevent recursive reference from causing infinite loop.
379 core.put( systemId, dom );
380 if(root)
381 rootDocuments.add(systemId);
382
383 try {
384 XMLReader reader = parserFactory.newSAXParser().getXMLReader();
385 reader.setContentHandler(getParserHandler(dom));
386 if(errorReceiver!=null)
387 reader.setErrorHandler(errorReceiver);
388 if(entityResolver!=null)
389 reader.setEntityResolver(entityResolver);
390 reader.parse(inputSource);
391 } catch( ParserConfigurationException e ) {
392 // in practice, this exception won't happen.
393 errorReceiver.error(e.getMessage(),e);
394 core.remove(systemId);
395 rootDocuments.remove(systemId);
396 return null;
397 } catch( IOException e ) {
398 errorReceiver.error(Messages.format(Messages.DOMFOREST_INPUTSOURCE_IOEXCEPTION, systemId, e.toString()),e);
399 core.remove(systemId);
400 rootDocuments.remove(systemId);
401 return null;
402 }
403
404 return dom;
405 }
406
407 public Document parse( String systemId, XMLStreamReader parser, boolean root ) throws XMLStreamException {
408 Document dom = documentBuilder.newDocument();
409
410 systemId = Options.normalizeSystemId(systemId);
411
412 if(root)
413 rootDocuments.add(systemId);
414
415 if(systemId==null)
416 throw new IllegalArgumentException("system id cannot be null");
417 core.put( systemId, dom );
418
419 new XMLStreamReaderToContentHandler(parser,getParserHandler(dom),false,false).bridge();
420
421 return dom;
422 }
423
424 /**
425 * Performs internalization.
426 *
427 * This method should be called only once, only after all the
428 * schemas are parsed.
429 *
430 * @return
431 * the returned bindings need to be applied after schema
432 * components are built.
433 */
434 public SCDBasedBindingSet transform(boolean enableSCD) {
435 return Internalizer.transform(this, enableSCD, options.disableXmlSecurity);
436 }
437
438 /**
439 * Performs the schema correctness check by using JAXP 1.3.
440 *
441 * <p>
442 * This is "weak", because {@link SchemaFactory#newSchema(Source[])}
443 * doesn't handle inclusions very correctly (it ends up parsing it
444 * from its original source, not in this tree), and because
445 * it doesn't handle two documents for the same namespace very
446 * well.
447 *
448 * <p>
449 * We should eventually fix JAXP (and Xerces), but meanwhile
450 * this weaker and potentially wrong correctness check is still
451 * better than nothing when used inside JAX-WS (JAXB CLI and Ant
452 * does a better job of checking this.)
453 *
454 * <p>
455 * To receive errors, use {@link SchemaFactory#setErrorHandler(ErrorHandler)}.
456 */
457 public void weakSchemaCorrectnessCheck(SchemaFactory sf) {
458 List<SAXSource> sources = new ArrayList<SAXSource>();
459 for( String systemId : getRootDocuments() ) {
460 Document dom = get(systemId);
461 if (dom.getDocumentElement().getNamespaceURI().equals(Const.JAXB_NSURI))
462 continue; // this isn't a schema. we have to do a negative check because if we see completely unrelated ns, we want to report that as an error
463
464 SAXSource ss = createSAXSource(systemId);
465 try {
466 ss.getXMLReader().setFeature("http://xml.org/sax/features/namespace-prefixes",true);
467 } catch (SAXException e) {
468 throw new AssertionError(e); // Xerces wants this. See 6395322.
469 }
470 sources.add(ss);
471 }
472
473 try {
474 sf.newSchema(sources.toArray(new SAXSource[0]));
475 } catch (SAXException e) {
476 // error should have been reported.
477 } catch (RuntimeException re) {
478 // JAXP RI isn't very trustworthy when it comes to schema error check,
479 // and we know some cases where it just dies with NPE. So handle it gracefully.
480 // this masks a bug in the JAXP RI, but we need a release that we have to make.
481 try {
482 sf.getErrorHandler().warning(
483 new SAXParseException(Messages.format(
484 Messages.ERR_GENERAL_SCHEMA_CORRECTNESS_ERROR,re.getMessage()),
485 null,null,-1,-1,re));
486 } catch (SAXException e) {
487 // ignore
488 }
489 }
490 }
491
492 /**
493 * Creates a {@link SAXSource} that, when parsed, reads from this {@link DOMForest}
494 * (instead of parsing the original source identified by the system ID.)
495 */
496 public @NotNull SAXSource createSAXSource(String systemId) {
497 ContentHandlerNamespacePrefixAdapter reader = new ContentHandlerNamespacePrefixAdapter(new XMLFilterImpl() {
498 // XMLReader that uses XMLParser to parse. We need to use XMLFilter to indrect
499 // handlers, since SAX allows handlers to be changed while parsing.
500 @Override
501 public void parse(InputSource input) throws SAXException, IOException {
502 createParser().parse(input, this, this, this);
503 }
504
505 @Override
506 public void parse(String systemId) throws SAXException, IOException {
507 parse(new InputSource(systemId));
508 }
509 });
510
511 return new SAXSource(reader,new InputSource(systemId));
512 }
513
514 /**
515 * Creates {@link XMLParser} for XSOM which reads documents from
516 * this DOMForest rather than doing a fresh parse.
517 *
518 * The net effect is that XSOM will read transformed XML Schemas
519 * instead of the original documents.
520 */
521 public XMLParser createParser() {
522 return new DOMForestParser(this, new JAXPParser(XmlFactory.createParserFactory(options.disableXmlSecurity)));
523 }
524
525 public EntityResolver getEntityResolver() {
526 return entityResolver;
527 }
528
529 public void setEntityResolver(EntityResolver entityResolver) {
530 this.entityResolver = entityResolver;
531 }
532
533 public ErrorReceiver getErrorHandler() {
534 return errorReceiver;
535 }
536
537 public void setErrorHandler(ErrorReceiver errorHandler) {
538 this.errorReceiver = errorHandler;
539 }
540
541 /**
542 * Gets all the parsed documents.
543 */
544 public Document[] listDocuments() {
545 return core.values().toArray(new Document[core.size()]);
546 }
547
548 /**
549 * Gets all the system IDs of the documents.
550 */
551 public String[] listSystemIDs() {
552 return core.keySet().toArray(new String[core.keySet().size()]);
553 }
554
555 /**
556 * Dumps the contents of the forest to the specified stream.
557 *
558 * This is a debug method. As such, error handling is sloppy.
559 */
560 @SuppressWarnings("CallToThreadDumpStack")
561 public void dump( OutputStream out ) throws IOException {
562 try {
563 // create identity transformer
564 boolean disableXmlSecurity = false;
565 if (options != null) {
566 disableXmlSecurity = options.disableXmlSecurity;
567 }
568 TransformerFactory tf = XmlFactory.createTransformerFactory(disableXmlSecurity);
569 Transformer it = tf.newTransformer();
570
571 for (Map.Entry<String, Document> e : core.entrySet()) {
572 out.write( ("---<< "+e.getKey()+'\n').getBytes() );
573
574 DataWriter dw = new DataWriter(new OutputStreamWriter(out),null);
575 dw.setIndentStep(" ");
576 it.transform( new DOMSource(e.getValue()),
577 new SAXResult(dw));
578
579 out.write( "\n\n\n".getBytes() );
580 }
581 } catch( TransformerException e ) {
582 e.printStackTrace();
583 }
584 }
585 }