1 /* 2 * Copyright (c) 2003, 2006, Oracle and/or its affiliates. All rights reserved. 3 */ 4 5 /* 6 * Copyright 2005 The Apache Software Foundation. 7 * 8 * Licensed under the Apache License, Version 2.0 (the "License"); 9 * you may not use this file except in compliance with the License. 10 * You may obtain a copy of the License at 11 * 12 * http://www.apache.org/licenses/LICENSE-2.0 13 * 14 * Unless required by applicable law or agreed to in writing, software 15 * distributed under the License is distributed on an "AS IS" BASIS, 16 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 17 * See the License for the specific language governing permissions and 18 * limitations under the License. 19 */ 20 21 22 package com.sun.org.apache.xerces.internal.impl; 23 24 import com.sun.xml.internal.stream.XMLBufferListener; 25 import com.sun.xml.internal.stream.XMLEntityStorage; 26 import com.sun.xml.internal.stream.XMLInputFactoryImpl; 27 import com.sun.xml.internal.stream.dtd.DTDGrammarUtil; 28 29 import java.io.EOFException; 30 import java.io.IOException; 31 import javax.xml.stream.XMLInputFactory; 32 import javax.xml.stream.events.XMLEvent; 33 import com.sun.org.apache.xerces.internal.impl.msg.XMLMessageFormatter; 34 import com.sun.org.apache.xerces.internal.util.AugmentationsImpl; 35 import com.sun.org.apache.xerces.internal.util.XMLAttributesIteratorImpl; 36 import com.sun.org.apache.xerces.internal.util.XMLChar; 37 import com.sun.org.apache.xerces.internal.util.XMLStringBuffer; 38 import com.sun.org.apache.xerces.internal.util.XMLSymbols; 39 import com.sun.org.apache.xerces.internal.xni.QName; 40 import com.sun.org.apache.xerces.internal.xni.XMLAttributes; 41 import com.sun.org.apache.xerces.internal.xni.XMLDocumentHandler; 42 import com.sun.org.apache.xerces.internal.xni.XMLResourceIdentifier; 43 import com.sun.org.apache.xerces.internal.xni.XMLString; 44 import com.sun.org.apache.xerces.internal.xni.XNIException; 45 import com.sun.org.apache.xerces.internal.xni.parser.XMLComponent; 46 import com.sun.org.apache.xerces.internal.xni.parser.XMLComponentManager; 47 import com.sun.org.apache.xerces.internal.xni.parser.XMLConfigurationException; 48 import com.sun.org.apache.xerces.internal.xni.parser.XMLDocumentScanner; 49 import com.sun.org.apache.xerces.internal.xni.parser.XMLInputSource; 50 import com.sun.org.apache.xerces.internal.xni.Augmentations; 51 import com.sun.org.apache.xerces.internal.impl.Constants; 52 import com.sun.org.apache.xerces.internal.impl.XMLEntityHandler; 53 import com.sun.org.apache.xerces.internal.util.SecurityManager; 54 import com.sun.org.apache.xerces.internal.util.NamespaceSupport; 55 import com.sun.org.apache.xerces.internal.xni.NamespaceContext; 56 import javax.xml.stream.XMLStreamConstants; 57 import javax.xml.stream.events.XMLEvent; 58 59 /** 60 * 61 * This class is responsible for scanning the structure and content 62 * of document fragments. 63 * 64 * This class has been modified as per the new design which is more suited to 65 * efficiently build pull parser. Lot of improvements have been done and 66 * the code has been added to support stax functionality/features. 67 * 68 * @author Neeraj Bajaj SUN Microsystems 69 * @author K.Venugopal SUN Microsystems 70 * @author Glenn Marcy, IBM 71 * @author Andy Clark, IBM 72 * @author Arnaud Le Hors, IBM 73 * @author Eric Ye, IBM 74 * @author Sunitha Reddy, SUN Microsystems 75 * @version $Id: XMLDocumentFragmentScannerImpl.java,v 1.19 2010-11-02 19:54:55 joehw Exp $ 76 * 77 */ 78 public class XMLDocumentFragmentScannerImpl 79 extends XMLScanner 80 implements XMLDocumentScanner, XMLComponent, XMLEntityHandler, XMLBufferListener { 81 82 // 83 // Constants 84 // 85 86 protected int fElementAttributeLimit; 87 88 /** External subset resolver. **/ 89 protected ExternalSubsetResolver fExternalSubsetResolver; 90 91 // scanner states 92 93 //XXX this should be divided into more states. 94 /** Scanner state: start of markup. */ 95 protected static final int SCANNER_STATE_START_OF_MARKUP = 21; 96 97 /** Scanner state: content. */ 98 protected static final int SCANNER_STATE_CONTENT = 22; 99 100 /** Scanner state: processing instruction. */ 101 protected static final int SCANNER_STATE_PI = 23; 102 103 /** Scanner state: DOCTYPE. */ 104 protected static final int SCANNER_STATE_DOCTYPE = 24; 105 106 /** Scanner state: XML Declaration */ 107 protected static final int SCANNER_STATE_XML_DECL = 25; 108 109 /** Scanner state: root element. */ 110 protected static final int SCANNER_STATE_ROOT_ELEMENT = 26; 111 112 /** Scanner state: comment. */ 113 protected static final int SCANNER_STATE_COMMENT = 27; 114 115 /** Scanner state: reference. */ 116 protected static final int SCANNER_STATE_REFERENCE = 28; 117 118 // <book type="hard"> reading attribute name 'type' 119 protected static final int SCANNER_STATE_ATTRIBUTE = 29; 120 121 // <book type="hard"> //reading attribute value. 122 protected static final int SCANNER_STATE_ATTRIBUTE_VALUE = 30; 123 124 /** Scanner state: trailing misc. USED BY DOCUMENT_SCANNER_IMPL*/ 125 //protected static final int SCANNER_STATE_TRAILING_MISC = 32; 126 127 /** Scanner state: end of input. */ 128 protected static final int SCANNER_STATE_END_OF_INPUT = 33; 129 130 /** Scanner state: terminated. */ 131 protected static final int SCANNER_STATE_TERMINATED = 34; 132 133 /** Scanner state: CDATA section. */ 134 protected static final int SCANNER_STATE_CDATA = 35; 135 136 /** Scanner state: Text declaration. */ 137 protected static final int SCANNER_STATE_TEXT_DECL = 36; 138 139 /** Scanner state: Text declaration. */ 140 protected static final int SCANNER_STATE_CHARACTER_DATA = 37; 141 142 //<book type="hard">foo</book> 143 protected static final int SCANNER_STATE_START_ELEMENT_TAG = 38; 144 145 //<book type="hard">foo</book> reading </book> 146 protected static final int SCANNER_STATE_END_ELEMENT_TAG = 39; 147 148 protected static final int SCANNER_STATE_CHAR_REFERENCE = 40; 149 protected static final int SCANNER_STATE_BUILT_IN_REFS = 41; 150 151 // feature identifiers 152 153 154 /** Feature identifier: notify built-in refereces. */ 155 protected static final String NOTIFY_BUILTIN_REFS = 156 Constants.XERCES_FEATURE_PREFIX + Constants.NOTIFY_BUILTIN_REFS_FEATURE; 157 158 /** Property identifier: entity resolver. */ 159 protected static final String ENTITY_RESOLVER = 160 Constants.XERCES_PROPERTY_PREFIX + Constants.ENTITY_RESOLVER_PROPERTY; 161 162 // recognized features and properties 163 164 /** Recognized features. */ 165 private static final String[] RECOGNIZED_FEATURES = { 166 NAMESPACES, 167 VALIDATION, 168 NOTIFY_BUILTIN_REFS, 169 NOTIFY_CHAR_REFS, 170 Constants.STAX_REPORT_CDATA_EVENT 171 }; 172 173 /** Feature defaults. */ 174 private static final Boolean[] FEATURE_DEFAULTS = { 175 Boolean.TRUE, 176 null, 177 Boolean.FALSE, 178 Boolean.FALSE, 179 Boolean.TRUE 180 }; 181 182 /** Recognized properties. */ 183 private static final String[] RECOGNIZED_PROPERTIES = { 184 SYMBOL_TABLE, 185 ERROR_REPORTER, 186 ENTITY_MANAGER, 187 }; 188 189 /** Property defaults. */ 190 private static final Object[] PROPERTY_DEFAULTS = { 191 null, 192 null, 193 null, 194 }; 195 196 protected static final char [] cdata = {'[','C','D','A','T','A','['}; 197 protected static final char [] xmlDecl = {'<','?','x','m','l'}; 198 protected static final char [] endTag = {'<','/'}; 199 // debugging 200 201 /** Debug scanner state. */ 202 private static final boolean DEBUG_SCANNER_STATE = false; 203 204 /** Debug driver. */ 205 private static final boolean DEBUG_DISPATCHER = false; 206 207 /** Debug content driver scanning. */ 208 protected static final boolean DEBUG_START_END_ELEMENT = false; 209 210 211 /** Debug driver next */ 212 protected static final boolean DEBUG_NEXT = false ; 213 214 /** Debug driver next */ 215 protected static final boolean DEBUG = false; 216 protected static final boolean DEBUG_COALESCE = false; 217 // 218 // Data 219 // 220 221 // protected data 222 223 /** Document handler. */ 224 protected XMLDocumentHandler fDocumentHandler; 225 protected int fScannerLastState ; 226 227 /** Entity Storage */ 228 protected XMLEntityStorage fEntityStore; 229 230 /** Entity stack. */ 231 protected int[] fEntityStack = new int[4]; 232 233 /** Markup depth. */ 234 protected int fMarkupDepth; 235 236 //is the element empty 237 protected boolean fEmptyElement ; 238 239 //track if we are reading attributes, this is usefule while 240 //there is a callback 241 protected boolean fReadingAttributes = false; 242 243 /** Scanner state. */ 244 protected int fScannerState; 245 246 /** SubScanner state: inside scanContent method. */ 247 protected boolean fInScanContent = false; 248 protected boolean fLastSectionWasCData = false; 249 protected boolean fLastSectionWasEntityReference = false; 250 protected boolean fLastSectionWasCharacterData = false; 251 252 /** has external dtd */ 253 protected boolean fHasExternalDTD; 254 255 /** Standalone. */ 256 protected boolean fStandaloneSet; 257 protected boolean fStandalone; 258 protected String fVersion; 259 260 // element information 261 262 /** Current element. */ 263 protected QName fCurrentElement; 264 265 /** Element stack. */ 266 protected ElementStack fElementStack = new ElementStack(); 267 protected ElementStack2 fElementStack2 = new ElementStack2(); 268 269 // other info 270 271 /** Document system identifier. 272 * REVISIT: So what's this used for? - NG 273 * protected String fDocumentSystemId; 274 ******/ 275 276 protected String fPITarget ; 277 278 //xxx do we need to create an extra XMLString object... look for using fTempString for collecting all the data values 279 protected XMLString fPIData = new XMLString(); 280 281 // features 282 283 284 /** Notify built-in references. */ 285 protected boolean fNotifyBuiltInRefs = false; 286 287 //STAX related properties 288 //defaultValues. 289 protected boolean fSupportDTD = true; 290 protected boolean fReplaceEntityReferences = true; 291 protected boolean fSupportExternalEntities = false; 292 protected boolean fReportCdataEvent = false ; 293 protected boolean fIsCoalesce = false ; 294 protected String fDeclaredEncoding = null; 295 /** Xerces Feature: Disallow doctype declaration. */ 296 protected boolean fDisallowDoctype = false; 297 298 // drivers 299 300 /** Active driver. */ 301 protected Driver fDriver; 302 303 /** Content driver. */ 304 protected Driver fContentDriver = createContentDriver(); 305 306 // temporary variables 307 308 /** Element QName. */ 309 protected QName fElementQName = new QName(); 310 311 /** Attribute QName. */ 312 protected QName fAttributeQName = new QName(); 313 314 /** 315 * CHANGED: Using XMLAttributesIteratorImpl instead of XMLAttributesImpl. This class 316 * implements Iterator interface so we can directly give Attributes in the form of 317 * iterator. 318 */ 319 protected XMLAttributesIteratorImpl fAttributes = new XMLAttributesIteratorImpl(); 320 321 322 /** String. */ 323 protected XMLString fTempString = new XMLString(); 324 325 /** String. */ 326 protected XMLString fTempString2 = new XMLString(); 327 328 /** Array of 3 strings. */ 329 private String[] fStrings = new String[3]; 330 331 /** Making the buffer accesible to derived class -- String buffer. */ 332 protected XMLStringBuffer fStringBuffer = new XMLStringBuffer(); 333 334 /** Making the buffer accesible to derived class -- String buffer. */ 335 protected XMLStringBuffer fStringBuffer2 = new XMLStringBuffer(); 336 337 /** stores character data. */ 338 /** Making the buffer accesible to derived class -- stores PI data */ 339 protected XMLStringBuffer fContentBuffer = new XMLStringBuffer(); 340 341 /** Single character array. */ 342 private final char[] fSingleChar = new char[1]; 343 private String fCurrentEntityName = null; 344 345 // New members 346 protected boolean fScanToEnd = false; 347 348 protected DTDGrammarUtil dtdGrammarUtil= null; 349 350 protected boolean fAddDefaultAttr = false; 351 352 protected boolean foundBuiltInRefs = false; 353 354 protected SecurityManager fSecurityManager = null; 355 356 //skip element algorithm 357 static final short MAX_DEPTH_LIMIT = 5 ; 358 static final short ELEMENT_ARRAY_LENGTH = 200 ; 359 static final short MAX_POINTER_AT_A_DEPTH = 4 ; 360 static final boolean DEBUG_SKIP_ALGORITHM = false; 361 //create a elemnet array of length equal to ELEMENT_ARRAY_LENGTH 362 String [] fElementArray = new String[ELEMENT_ARRAY_LENGTH] ; 363 //pointer location where last element was skipped 364 short fLastPointerLocation = 0 ; 365 short fElementPointer = 0 ; 366 //2D array to store pointer info 367 short [] [] fPointerInfo = new short[MAX_DEPTH_LIMIT] [MAX_POINTER_AT_A_DEPTH] ; 368 protected String fElementRawname ; 369 protected boolean fShouldSkip = false; 370 protected boolean fAdd = false ; 371 protected boolean fSkip = false; 372 373 /** Reusable Augmentations. */ 374 private Augmentations fTempAugmentations = null; 375 // 376 // Constructors 377 // 378 379 /** Default constructor. */ 380 public XMLDocumentFragmentScannerImpl() { 381 } // <init>() 382 383 // 384 // XMLDocumentScanner methods 385 // 386 387 /** 388 * Sets the input source. 389 * 390 * @param inputSource The input source. 391 * 392 * @throws IOException Thrown on i/o error. 393 */ 394 public void setInputSource(XMLInputSource inputSource) throws IOException { 395 fEntityManager.setEntityHandler(this); 396 fEntityManager.startEntity("$fragment$", inputSource, false, true); 397 // fDocumentSystemId = fEntityManager.expandSystemId(inputSource.getSystemId()); 398 } // setInputSource(XMLInputSource) 399 400 /** 401 * Scans a document. 402 * 403 * @param complete True if the scanner should scan the document 404 * completely, pushing all events to the registered 405 * document handler. A value of false indicates that 406 * that the scanner should only scan the next portion 407 * of the document and return. A scanner instance is 408 * permitted to completely scan a document if it does 409 * not support this "pull" scanning model. 410 * 411 * @return True if there is more to scan, false otherwise. 412 */ 413 /* public boolean scanDocument(boolean complete) 414 throws IOException, XNIException { 415 416 // keep dispatching "events" 417 fEntityManager.setEntityHandler(this); 418 419 return true; 420 421 } // scanDocument(boolean):boolean 422 */ 423 424 public boolean scanDocument(boolean complete) 425 throws IOException, XNIException { 426 427 // keep dispatching "events" 428 fEntityManager.setEntityHandler(this); 429 //System.out.println(" get Document Handler in NSDocumentHandler " + fDocumentHandler ); 430 431 int event = next(); 432 do { 433 switch (event) { 434 case XMLStreamConstants.START_DOCUMENT : 435 //fDocumentHandler.startDocument(fEntityManager.getEntityScanner(),fEntityManager.getEntityScanner().getVersion(),fNamespaceContext,null);// not able to get 436 break; 437 case XMLStreamConstants.START_ELEMENT : 438 //System.out.println(" in scann element"); 439 //fDocumentHandler.startElement(getElementQName(),fAttributes,null); 440 break; 441 case XMLStreamConstants.CHARACTERS : 442 fDocumentHandler.characters(getCharacterData(),null); 443 break; 444 case XMLStreamConstants.SPACE: 445 //check if getCharacterData() is the right function to retrieve ignorableWhitespace information. 446 //System.out.println("in the space"); 447 //fDocumentHandler.ignorableWhitespace(getCharacterData(), null); 448 break; 449 case XMLStreamConstants.ENTITY_REFERENCE : 450 //entity reference callback are given in startEntity 451 break; 452 case XMLStreamConstants.PROCESSING_INSTRUCTION : 453 fDocumentHandler.processingInstruction(getPITarget(),getPIData(),null); 454 break; 455 case XMLStreamConstants.COMMENT : 456 //System.out.println(" in COMMENT of the XMLNSDocumentScannerImpl"); 457 fDocumentHandler.comment(getCharacterData(),null); 458 break; 459 case XMLStreamConstants.DTD : 460 //all DTD related callbacks are handled in DTDScanner. 461 //1. Stax doesn't define DTD states as it does for XML Document. 462 //therefore we don't need to take care of anything here. So Just break; 463 break; 464 case XMLStreamConstants.CDATA: 465 fDocumentHandler.startCDATA(null); 466 //xxx: check if CDATA values comes from getCharacterData() function 467 fDocumentHandler.characters(getCharacterData(),null); 468 fDocumentHandler.endCDATA(null); 469 //System.out.println(" in CDATA of the XMLNSDocumentScannerImpl"); 470 break; 471 case XMLStreamConstants.NOTATION_DECLARATION : 472 break; 473 case XMLStreamConstants.ENTITY_DECLARATION : 474 break; 475 case XMLStreamConstants.NAMESPACE : 476 break; 477 case XMLStreamConstants.ATTRIBUTE : 478 break; 479 case XMLStreamConstants.END_ELEMENT : 480 //do not give callback here. 481 //this callback is given in scanEndElement function. 482 //fDocumentHandler.endElement(getElementQName(),null); 483 break; 484 default : 485 throw new InternalError("processing event: " + event); 486 487 } 488 //System.out.println("here in before calling next"); 489 event = next(); 490 //System.out.println("here in after calling next"); 491 } while (event!=XMLStreamConstants.END_DOCUMENT && complete); 492 493 if(event == XMLStreamConstants.END_DOCUMENT) { 494 fDocumentHandler.endDocument(null); 495 return false; 496 } 497 498 return true; 499 500 } // scanDocument(boolean):boolean 501 502 503 504 public com.sun.org.apache.xerces.internal.xni.QName getElementQName(){ 505 if(fScannerLastState == XMLEvent.END_ELEMENT){ 506 fElementQName.setValues(fElementStack.getLastPoppedElement()); 507 } 508 return fElementQName ; 509 } 510 511 /** return the next state on the input 512 * @return int 513 */ 514 515 public int next() throws IOException, XNIException { 516 return fDriver.next(); 517 } 518 519 // 520 // XMLComponent methods 521 // 522 523 /** 524 * Resets the component. The component can query the component manager 525 * about any features and properties that affect the operation of the 526 * component. 527 * 528 * @param componentManager The component manager. 529 * 530 * @throws SAXException Thrown by component on initialization error. 531 * For example, if a feature or property is 532 * required for the operation of the component, the 533 * component manager may throw a 534 * SAXNotRecognizedException or a 535 * SAXNotSupportedException. 536 */ 537 538 public void reset(XMLComponentManager componentManager) 539 throws XMLConfigurationException { 540 541 super.reset(componentManager); 542 543 // other settings 544 // fDocumentSystemId = null; 545 546 // sax features 547 //fAttributes.setNamespaces(fNamespaces); 548 549 // xerces features 550 fReportCdataEvent = componentManager.getFeature(Constants.STAX_REPORT_CDATA_EVENT, true); 551 552 fSecurityManager = (SecurityManager)componentManager.getProperty(Constants.SECURITY_MANAGER, null); 553 fElementAttributeLimit = (fSecurityManager != null)?fSecurityManager.getElementAttrLimit():0; 554 555 fNotifyBuiltInRefs = componentManager.getFeature(NOTIFY_BUILTIN_REFS, false); 556 557 Object resolver = componentManager.getProperty(ENTITY_RESOLVER, null); 558 fExternalSubsetResolver = (resolver instanceof ExternalSubsetResolver) ? 559 (ExternalSubsetResolver) resolver : null; 560 561 // initialize vars 562 fMarkupDepth = 0; 563 fCurrentElement = null; 564 fElementStack.clear(); 565 fHasExternalDTD = false; 566 fStandaloneSet = false; 567 fStandalone = false; 568 fInScanContent = false; 569 //skipping algorithm 570 fShouldSkip = false; 571 fAdd = false; 572 fSkip = false; 573 574 //attribute 575 fReadingAttributes = false; 576 //xxx: external entities are supported in Xerces 577 // it would be good to define feature for this case 578 fSupportExternalEntities = true; 579 fReplaceEntityReferences = true; 580 fIsCoalesce = false; 581 582 // setup Driver 583 setScannerState(SCANNER_STATE_CONTENT); 584 setDriver(fContentDriver); 585 fEntityStore = fEntityManager.getEntityStore(); 586 587 dtdGrammarUtil = null; 588 589 590 //fEntityManager.test(); 591 } // reset(XMLComponentManager) 592 593 594 public void reset(PropertyManager propertyManager){ 595 596 super.reset(propertyManager); 597 598 // other settings 599 // fDocumentSystemId = null; 600 fNamespaces = ((Boolean)propertyManager.getProperty(XMLInputFactory.IS_NAMESPACE_AWARE)).booleanValue(); 601 fNotifyBuiltInRefs = false ; 602 603 // initialize vars 604 fMarkupDepth = 0; 605 fCurrentElement = null; 606 fShouldSkip = false; 607 fAdd = false; 608 fSkip = false; 609 fElementStack.clear(); 610 //fElementStack2.clear(); 611 fHasExternalDTD = false; 612 fStandaloneSet = false; 613 fStandalone = false; 614 //fReplaceEntityReferences = true; 615 //fSupportExternalEntities = true; 616 Boolean bo = (Boolean)propertyManager.getProperty(XMLInputFactoryImpl.IS_REPLACING_ENTITY_REFERENCES); 617 fReplaceEntityReferences = bo.booleanValue(); 618 bo = (Boolean)propertyManager.getProperty(XMLInputFactoryImpl.IS_SUPPORTING_EXTERNAL_ENTITIES); 619 fSupportExternalEntities = bo.booleanValue(); 620 Boolean cdata = (Boolean)propertyManager.getProperty(Constants.ZEPHYR_PROPERTY_PREFIX + Constants.STAX_REPORT_CDATA_EVENT) ; 621 if(cdata != null) 622 fReportCdataEvent = cdata.booleanValue() ; 623 Boolean coalesce = (Boolean)propertyManager.getProperty(XMLInputFactory.IS_COALESCING) ; 624 if(coalesce != null) 625 fIsCoalesce = coalesce.booleanValue(); 626 fReportCdataEvent = fIsCoalesce ? false : (fReportCdataEvent && true) ; 627 //if fIsCoalesce is set to true, set the value of fReplaceEntityReferences to true, 628 //if fIsCoalesce is set to false, take the value of fReplaceEntityReferences as set by application 629 fReplaceEntityReferences = fIsCoalesce ? true : fReplaceEntityReferences; 630 // setup Driver 631 //we dont need to do this -- nb. 632 //setScannerState(SCANNER_STATE_CONTENT); 633 //setDriver(fContentDriver); 634 fEntityStore = fEntityManager.getEntityStore(); 635 //fEntityManager.test(); 636 637 dtdGrammarUtil = null; 638 639 } // reset(XMLComponentManager) 640 641 /** 642 * Returns a list of feature identifiers that are recognized by 643 * this component. This method may return null if no features 644 * are recognized by this component. 645 */ 646 public String[] getRecognizedFeatures() { 647 return (String[])(RECOGNIZED_FEATURES.clone()); 648 } // getRecognizedFeatures():String[] 649 650 /** 651 * Sets the state of a feature. This method is called by the component 652 * manager any time after reset when a feature changes state. 653 * <p> 654 * <strong>Note:</strong> Components should silently ignore features 655 * that do not affect the operation of the component. 656 * 657 * @param featureId The feature identifier. 658 * @param state The state of the feature. 659 * 660 * @throws SAXNotRecognizedException The component should not throw 661 * this exception. 662 * @throws SAXNotSupportedException The component should not throw 663 * this exception. 664 */ 665 public void setFeature(String featureId, boolean state) 666 throws XMLConfigurationException { 667 668 super.setFeature(featureId, state); 669 670 // Xerces properties 671 if (featureId.startsWith(Constants.XERCES_FEATURE_PREFIX)) { 672 String feature = featureId.substring(Constants.XERCES_FEATURE_PREFIX.length()); 673 if (feature.equals(Constants.NOTIFY_BUILTIN_REFS_FEATURE)) { 674 fNotifyBuiltInRefs = state; 675 } 676 } 677 678 } // setFeature(String,boolean) 679 680 /** 681 * Returns a list of property identifiers that are recognized by 682 * this component. This method may return null if no properties 683 * are recognized by this component. 684 */ 685 public String[] getRecognizedProperties() { 686 return (String[])(RECOGNIZED_PROPERTIES.clone()); 687 } // getRecognizedProperties():String[] 688 689 /** 690 * Sets the value of a property. This method is called by the component 691 * manager any time after reset when a property changes value. 692 * <p> 693 * <strong>Note:</strong> Components should silently ignore properties 694 * that do not affect the operation of the component. 695 * 696 * @param propertyId The property identifier. 697 * @param value The value of the property. 698 * 699 * @throws SAXNotRecognizedException The component should not throw 700 * this exception. 701 * @throws SAXNotSupportedException The component should not throw 702 * this exception. 703 */ 704 public void setProperty(String propertyId, Object value) 705 throws XMLConfigurationException { 706 707 super.setProperty(propertyId, value); 708 709 // Xerces properties 710 if (propertyId.startsWith(Constants.XERCES_PROPERTY_PREFIX)) { 711 final int suffixLength = propertyId.length() - Constants.XERCES_PROPERTY_PREFIX.length(); 712 if (suffixLength == Constants.ENTITY_MANAGER_PROPERTY.length() && 713 propertyId.endsWith(Constants.ENTITY_MANAGER_PROPERTY)) { 714 fEntityManager = (XMLEntityManager)value; 715 return; 716 } 717 if (suffixLength == Constants.ENTITY_RESOLVER_PROPERTY.length() && 718 propertyId.endsWith(Constants.ENTITY_RESOLVER_PROPERTY)) { 719 fExternalSubsetResolver = (value instanceof ExternalSubsetResolver) ? 720 (ExternalSubsetResolver) value : null; 721 return; 722 } 723 } 724 725 726 // Xerces properties 727 if (propertyId.startsWith(Constants.XERCES_PROPERTY_PREFIX)) { 728 String property = propertyId.substring(Constants.XERCES_PROPERTY_PREFIX.length()); 729 if (property.equals(Constants.ENTITY_MANAGER_PROPERTY)) { 730 fEntityManager = (XMLEntityManager)value; 731 } 732 return; 733 } 734 735 } // setProperty(String,Object) 736 737 /** 738 * Returns the default state for a feature, or null if this 739 * component does not want to report a default value for this 740 * feature. 741 * 742 * @param featureId The feature identifier. 743 * 744 * @since Xerces 2.2.0 745 */ 746 public Boolean getFeatureDefault(String featureId) { 747 for (int i = 0; i < RECOGNIZED_FEATURES.length; i++) { 748 if (RECOGNIZED_FEATURES[i].equals(featureId)) { 749 return FEATURE_DEFAULTS[i]; 750 } 751 } 752 return null; 753 } // getFeatureDefault(String):Boolean 754 755 /** 756 * Returns the default state for a property, or null if this 757 * component does not want to report a default value for this 758 * property. 759 * 760 * @param propertyId The property identifier. 761 * 762 * @since Xerces 2.2.0 763 */ 764 public Object getPropertyDefault(String propertyId) { 765 for (int i = 0; i < RECOGNIZED_PROPERTIES.length; i++) { 766 if (RECOGNIZED_PROPERTIES[i].equals(propertyId)) { 767 return PROPERTY_DEFAULTS[i]; 768 } 769 } 770 return null; 771 } // getPropertyDefault(String):Object 772 773 // 774 // XMLDocumentSource methods 775 // 776 777 /** 778 * setDocumentHandler 779 * 780 * @param documentHandler 781 */ 782 public void setDocumentHandler(XMLDocumentHandler documentHandler) { 783 fDocumentHandler = documentHandler; 784 //System.out.println(" In Set DOCUMENT HANDLER" + fDocumentHandler + " scanner =" + this); 785 } // setDocumentHandler(XMLDocumentHandler) 786 787 788 /** Returns the document handler */ 789 public XMLDocumentHandler getDocumentHandler(){ 790 return fDocumentHandler; 791 } 792 793 // 794 // XMLEntityHandler methods 795 // 796 797 /** 798 * This method notifies of the start of an entity. The DTD has the 799 * pseudo-name of "[dtd]" parameter entity names start with '%'; and 800 * general entities are just specified by their name. 801 * 802 * @param name The name of the entity. 803 * @param identifier The resource identifier. 804 * @param encoding The auto-detected IANA encoding name of the entity 805 * stream. This value will be null in those situations 806 * where the entity encoding is not auto-detected (e.g. 807 * internal entities or a document entity that is 808 * parsed from a java.io.Reader). 809 * 810 * @throws XNIException Thrown by handler to signal an error. 811 */ 812 public void startEntity(String name, 813 XMLResourceIdentifier identifier, 814 String encoding, Augmentations augs) throws XNIException { 815 816 // keep track of this entity before fEntityDepth is increased 817 if (fEntityDepth == fEntityStack.length) { 818 int[] entityarray = new int[fEntityStack.length * 2]; 819 System.arraycopy(fEntityStack, 0, entityarray, 0, fEntityStack.length); 820 fEntityStack = entityarray; 821 } 822 fEntityStack[fEntityDepth] = fMarkupDepth; 823 824 super.startEntity(name, identifier, encoding, augs); 825 826 // WFC: entity declared in external subset in standalone doc 827 if(fStandalone && fEntityStore.isEntityDeclInExternalSubset(name)) { 828 reportFatalError("MSG_REFERENCE_TO_EXTERNALLY_DECLARED_ENTITY_WHEN_STANDALONE", 829 new Object[]{name}); 830 } 831 832 /** we are not calling the handlers yet.. */ 833 // call handler 834 if (fDocumentHandler != null && !fScanningAttribute) { 835 if (!name.equals("[xml]")) { 836 fDocumentHandler.startGeneralEntity(name, identifier, encoding, null); 837 } 838 } 839 840 } // startEntity(String,XMLResourceIdentifier,String) 841 842 /** 843 * This method notifies the end of an entity. The DTD has the pseudo-name 844 * of "[dtd]" parameter entity names start with '%'; and general entities 845 * are just specified by their name. 846 * 847 * @param name The name of the entity. 848 * 849 * @throws XNIException Thrown by handler to signal an error. 850 */ 851 public void endEntity(String name, Augmentations augs) throws IOException, XNIException { 852 853 /** 854 * // flush possible pending output buffer - see scanContent 855 * if (fInScanContent && fStringBuffer.length != 0 856 * && fDocumentHandler != null) { 857 * fDocumentHandler.characters(fStringBuffer, null); 858 * fStringBuffer.length = 0; // make sure we know it's been flushed 859 * } 860 */ 861 super.endEntity(name, augs); 862 863 // make sure markup is properly balanced 864 if (fMarkupDepth != fEntityStack[fEntityDepth]) { 865 reportFatalError("MarkupEntityMismatch", null); 866 } 867 868 /**/ 869 // call handler 870 if (fDocumentHandler != null && !fScanningAttribute) { 871 if (!name.equals("[xml]")) { 872 fDocumentHandler.endGeneralEntity(name, null); 873 } 874 } 875 876 877 } // endEntity(String) 878 879 // 880 // Protected methods 881 // 882 883 // Driver factory methods 884 885 /** Creates a content Driver. */ 886 protected Driver createContentDriver() { 887 return new FragmentContentDriver(); 888 } // createContentDriver():Driver 889 890 // scanning methods 891 892 /** 893 * Scans an XML or text declaration. 894 * <p> 895 * <pre> 896 * [23] XMLDecl ::= '<?xml' VersionInfo EncodingDecl? SDDecl? S? '?>' 897 * [24] VersionInfo ::= S 'version' Eq (' VersionNum ' | " VersionNum ") 898 * [80] EncodingDecl ::= S 'encoding' Eq ('"' EncName '"' | "'" EncName "'" ) 899 * [81] EncName ::= [A-Za-z] ([A-Za-z0-9._] | '-')* 900 * [32] SDDecl ::= S 'standalone' Eq (("'" ('yes' | 'no') "'") 901 * | ('"' ('yes' | 'no') '"')) 902 * 903 * [77] TextDecl ::= '<?xml' VersionInfo? EncodingDecl S? '?>' 904 * </pre> 905 * 906 * @param scanningTextDecl True if a text declaration is to 907 * be scanned instead of an XML 908 * declaration. 909 */ 910 protected void scanXMLDeclOrTextDecl(boolean scanningTextDecl) 911 throws IOException, XNIException { 912 913 // scan decl 914 super.scanXMLDeclOrTextDecl(scanningTextDecl, fStrings); 915 fMarkupDepth--; 916 917 // pseudo-attribute values 918 String version = fStrings[0]; 919 String encoding = fStrings[1]; 920 String standalone = fStrings[2]; 921 fDeclaredEncoding = encoding; 922 // set standalone 923 fStandaloneSet = standalone != null; 924 fStandalone = fStandaloneSet && standalone.equals("yes"); 925 ///xxx see where its used.. this is not used anywhere. it may be useful for entity to store this information 926 //but this information is only related with Document Entity. 927 fEntityManager.setStandalone(fStandalone); 928 929 930 // call handler 931 if (fDocumentHandler != null) { 932 if (scanningTextDecl) { 933 fDocumentHandler.textDecl(version, encoding, null); 934 } else { 935 fDocumentHandler.xmlDecl(version, encoding, standalone, null); 936 } 937 } 938 939 if(version != null){ 940 fEntityScanner.setVersion(version); 941 fEntityScanner.setXMLVersion(version); 942 } 943 // set encoding on reader, only if encoding was not specified by the application explicitly 944 if (encoding != null && !fEntityScanner.getCurrentEntity().isEncodingExternallySpecified()) { 945 fEntityScanner.setEncoding(encoding); 946 } 947 948 } // scanXMLDeclOrTextDecl(boolean) 949 950 public String getPITarget(){ 951 return fPITarget ; 952 } 953 954 public XMLStringBuffer getPIData(){ 955 return fContentBuffer ; 956 } 957 958 //XXX: why not this function behave as per the state of the parser? 959 public XMLString getCharacterData(){ 960 if(fUsebuffer){ 961 return fContentBuffer ; 962 }else{ 963 return fTempString; 964 } 965 966 } 967 968 969 /** 970 * Scans a processing data. This is needed to handle the situation 971 * where a document starts with a processing instruction whose 972 * target name <em>starts with</em> "xml". (e.g. xmlfoo) 973 * 974 * @param target The PI target 975 * @param data The XMLStringBuffer to fill in with the data 976 */ 977 protected void scanPIData(String target, XMLStringBuffer data) 978 throws IOException, XNIException { 979 980 super.scanPIData(target, data); 981 982 //set the PI target and values 983 fPITarget = target ; 984 985 fMarkupDepth--; 986 987 } // scanPIData(String) 988 989 /** 990 * Scans a comment. 991 * <p> 992 * <pre> 993 * [15] Comment ::= '<!--' ((Char - '-') | ('-' (Char - '-')))* '-->' 994 * </pre> 995 * <p> 996 * <strong>Note:</strong> Called after scanning past '<!--' 997 */ 998 protected void scanComment() throws IOException, XNIException { 999 fContentBuffer.clear(); 1000 scanComment(fContentBuffer); 1001 //getTextCharacters can also be called for reading comments 1002 fUsebuffer = true; 1003 fMarkupDepth--; 1004 1005 } // scanComment() 1006 1007 //xxx value returned by this function may not remain valid if another event is scanned. 1008 public String getComment(){ 1009 return fContentBuffer.toString(); 1010 } 1011 1012 void addElement(String rawname){ 1013 if(fElementPointer < ELEMENT_ARRAY_LENGTH){ 1014 //storing element raw name in a linear list of array 1015 fElementArray[fElementPointer] = rawname ; 1016 //storing elemnetPointer for particular element depth 1017 1018 if(DEBUG_SKIP_ALGORITHM){ 1019 StringBuffer sb = new StringBuffer() ; 1020 sb.append(" Storing element information ") ; 1021 sb.append(" fElementPointer = " + fElementPointer) ; 1022 sb.append(" fElementRawname = " + fElementQName.rawname) ; 1023 sb.append(" fElementStack.fDepth = " + fElementStack.fDepth); 1024 System.out.println(sb.toString()) ; 1025 } 1026 1027 //store pointer information only when element depth is less MAX_DEPTH_LIMIT 1028 if(fElementStack.fDepth < MAX_DEPTH_LIMIT){ 1029 short column = storePointerForADepth(fElementPointer); 1030 if(column > 0){ 1031 short pointer = getElementPointer((short)fElementStack.fDepth, (short)(column - 1) ); 1032 //identity comparison shouldn't take much time and we can rely on this 1033 //since its guaranteed to have same object id for same string. 1034 if(rawname == fElementArray[pointer]){ 1035 fShouldSkip = true ; 1036 fLastPointerLocation = pointer ; 1037 //reset the things and return. 1038 resetPointer((short)fElementStack.fDepth , column) ; 1039 fElementArray[fElementPointer] = null ; 1040 return ; 1041 }else{ 1042 fShouldSkip = false ; 1043 } 1044 } 1045 } 1046 fElementPointer++ ; 1047 } 1048 } 1049 1050 1051 void resetPointer(short depth, short column){ 1052 fPointerInfo[depth] [column] = (short)0; 1053 } 1054 1055 //returns column information at which pointer was stored. 1056 short storePointerForADepth(short elementPointer){ 1057 short depth = (short) fElementStack.fDepth ; 1058 1059 //Stores element pointer locations at particular depth , only 4 pointer locations 1060 //are stored at particular depth for now. 1061 for(short i = 0 ; i < MAX_POINTER_AT_A_DEPTH ; i++){ 1062 1063 if(canStore(depth, i)){ 1064 fPointerInfo[depth][i] = elementPointer ; 1065 if(DEBUG_SKIP_ALGORITHM){ 1066 StringBuffer sb = new StringBuffer() ; 1067 sb.append(" Pointer information ") ; 1068 sb.append(" fElementPointer = " + fElementPointer) ; 1069 sb.append(" fElementStack.fDepth = " + fElementStack.fDepth); 1070 sb.append(" column = " + i ) ; 1071 System.out.println(sb.toString()) ; 1072 } 1073 return i; 1074 } 1075 //else 1076 //pointer was not stored because we reached the limit 1077 } 1078 return -1 ; 1079 } 1080 1081 boolean canStore(short depth, short column){ 1082 //colum = 0 , means first element at particular depth 1083 //column = 1, means second element at particular depth 1084 // calle should make sure that it doesn't call for value outside allowed co-ordinates 1085 return fPointerInfo[depth][column] == 0 ? true : false ; 1086 } 1087 1088 1089 short getElementPointer(short depth, short column){ 1090 //colum = 0 , means first element at particular depth 1091 //column = 1, means second element at particular depth 1092 // calle should make sure that it doesn't call for value outside allowed co-ordinates 1093 return fPointerInfo[depth][column] ; 1094 } 1095 1096 //this function assumes that string passed is not null and skips 1097 //the following string from the buffer this makes sure 1098 boolean skipFromTheBuffer(String rawname) throws IOException{ 1099 if(fEntityScanner.skipString(rawname)){ 1100 char c = (char)fEntityScanner.peekChar() ; 1101 //If the start element was completely skipped we should encounter either ' '(space), 1102 //or '/' (in case of empty element) or '>' 1103 if( c == ' ' || c == '/' || c == '>'){ 1104 fElementRawname = rawname ; 1105 return true ; 1106 } else{ 1107 return false; 1108 } 1109 } else 1110 return false ; 1111 } 1112 1113 boolean skipQElement(String rawname) throws IOException{ 1114 1115 final int c = fEntityScanner.getChar(rawname.length()); 1116 //if this character is still valid element name -- this means string can't match 1117 if(XMLChar.isName(c)){ 1118 return false; 1119 }else{ 1120 return fEntityScanner.skipString(rawname); 1121 } 1122 } 1123 1124 protected boolean skipElement() throws IOException { 1125 1126 if(!fShouldSkip) return false ; 1127 1128 if(fLastPointerLocation != 0){ 1129 //Look at the next element stored in the array list.. we might just get a match. 1130 String rawname = fElementArray[fLastPointerLocation + 1] ; 1131 if(rawname != null && skipFromTheBuffer(rawname)){ 1132 fLastPointerLocation++ ; 1133 if(DEBUG_SKIP_ALGORITHM){ 1134 System.out.println("Element " + fElementRawname + " was SKIPPED at pointer location = " + fLastPointerLocation); 1135 } 1136 return true ; 1137 } else{ 1138 //reset it back to zero... we haven't got the correct subset yet. 1139 fLastPointerLocation = 0 ; 1140 1141 } 1142 } 1143 //xxx: we can put some logic here as from what column it should start looking 1144 //for now we always start at 0 1145 //fallback to tolerant algorithm, it would look for differnt element stored at different 1146 //depth and get us the pointer location. 1147 return fShouldSkip && skipElement((short)0); 1148 1149 } 1150 1151 //start of the column at which it should try searching 1152 boolean skipElement(short column) throws IOException { 1153 short depth = (short)fElementStack.fDepth ; 1154 1155 if(depth > MAX_DEPTH_LIMIT){ 1156 return fShouldSkip = false ; 1157 } 1158 for(short i = column ; i < MAX_POINTER_AT_A_DEPTH ; i++){ 1159 short pointer = getElementPointer(depth , i ) ; 1160 1161 if(pointer == 0){ 1162 return fShouldSkip = false ; 1163 } 1164 1165 if(fElementArray[pointer] != null && skipFromTheBuffer(fElementArray[pointer])){ 1166 if(DEBUG_SKIP_ALGORITHM){ 1167 System.out.println(); 1168 System.out.println("Element " + fElementRawname + " was SKIPPED at depth = " + fElementStack.fDepth + " column = " + column ); 1169 System.out.println(); 1170 } 1171 fLastPointerLocation = pointer ; 1172 return fShouldSkip = true ; 1173 } 1174 } 1175 return fShouldSkip = false ; 1176 } 1177 1178 /** 1179 * Scans a start element. This method will handle the binding of 1180 * namespace information and notifying the handler of the start 1181 * of the element. 1182 * <p> 1183 * <pre> 1184 * [44] EmptyElemTag ::= '<' Name (S Attribute)* S? '/>' 1185 * [40] STag ::= '<' Name (S Attribute)* S? '>' 1186 * </pre> 1187 * <p> 1188 * <strong>Note:</strong> This method assumes that the leading 1189 * '<' character has been consumed. 1190 * <p> 1191 * <strong>Note:</strong> This method uses the fElementQName and 1192 * fAttributes variables. The contents of these variables will be 1193 * destroyed. The caller should copy important information out of 1194 * these variables before calling this method. 1195 * NB: Content in fAttributes is valid only till the state of the parser is XMLEvent.START_ELEMENT 1196 * 1197 * @return True if element is empty. (i.e. It matches 1198 * production [44]. 1199 */ 1200 // fElementQName will have the details of element just read.. 1201 // fAttributes will have the details of all the attributes. 1202 protected boolean scanStartElement() 1203 throws IOException, XNIException { 1204 1205 if (DEBUG_START_END_ELEMENT) System.out.println( this.getClass().toString() + ">>> scanStartElement()"); 1206 //when skipping is true and no more elements should be added 1207 if(fSkip && !fAdd){ 1208 //get the stored element -- if everything goes right this should match the 1209 //token in the buffer 1210 1211 QName name = fElementStack.getNext(); 1212 1213 if(DEBUG_SKIP_ALGORITHM){ 1214 System.out.println("Trying to skip String = " + name.rawname); 1215 } 1216 1217 //Be conservative -- if skipping fails -- stop. 1218 fSkip = fEntityScanner.skipString(name.rawname); 1219 1220 if(fSkip){ 1221 if(DEBUG_SKIP_ALGORITHM){ 1222 System.out.println("Element SUCESSFULLY skipped = " + name.rawname); 1223 } 1224 fElementStack.push(); 1225 fElementQName = name; 1226 }else{ 1227 //if skipping fails reposition the stack or fallback to normal way of processing 1228 fElementStack.reposition(); 1229 if(DEBUG_SKIP_ALGORITHM){ 1230 System.out.println("Element was NOT skipped, REPOSITIONING stack" ); 1231 } 1232 } 1233 } 1234 1235 //we are still at the stage of adding elements 1236 //the elements were not matched or 1237 //fSkip is not set to true 1238 if(!fSkip || fAdd){ 1239 //get the next element from the stack 1240 fElementQName = fElementStack.nextElement(); 1241 // name 1242 if (fNamespaces) { 1243 fEntityScanner.scanQName(fElementQName); 1244 } else { 1245 String name = fEntityScanner.scanName(); 1246 fElementQName.setValues(null, name, name, null); 1247 } 1248 1249 if(DEBUG)System.out.println("Element scanned in start element is " + fElementQName.toString()); 1250 if(DEBUG_SKIP_ALGORITHM){ 1251 if(fAdd){ 1252 System.out.println("Elements are being ADDED -- elemet added is = " + fElementQName.rawname + " at count = " + fElementStack.fCount); 1253 } 1254 } 1255 1256 } 1257 1258 //when the elements are being added , we need to check if we are set for skipping the elements 1259 if(fAdd){ 1260 //this sets the value of fAdd variable 1261 fElementStack.matchElement(fElementQName); 1262 } 1263 1264 1265 //xxx: We dont need another pointer, fCurrentElement, we can use fElementQName 1266 fCurrentElement = fElementQName; 1267 1268 String rawname = fElementQName.rawname; 1269 1270 fEmptyElement = false; 1271 1272 fAttributes.removeAllAttributes(); 1273 1274 if(!seekCloseOfStartTag()){ 1275 fReadingAttributes = true; 1276 fAttributeCacheUsedCount =0; 1277 fStringBufferIndex =0; 1278 fAddDefaultAttr = true; 1279 do { 1280 scanAttribute(fAttributes); 1281 if (fSecurityManager != null && fAttributes.getLength() > fElementAttributeLimit){ 1282 fErrorReporter.reportError(XMLMessageFormatter.XML_DOMAIN, 1283 "ElementAttributeLimit", 1284 new Object[]{rawname, new Integer(fAttributes.getLength()) }, 1285 XMLErrorReporter.SEVERITY_FATAL_ERROR ); 1286 } 1287 1288 } while (!seekCloseOfStartTag()); 1289 fReadingAttributes=false; 1290 } 1291 1292 if (fEmptyElement) { 1293 //decrease the markup depth.. 1294 fMarkupDepth--; 1295 1296 // check that this element was opened in the same entity 1297 if (fMarkupDepth < fEntityStack[fEntityDepth - 1]) { 1298 reportFatalError("ElementEntityMismatch", 1299 new Object[]{fCurrentElement.rawname}); 1300 } 1301 // call handler 1302 if (fDocumentHandler != null) { 1303 fDocumentHandler.emptyElement(fElementQName, fAttributes, null); 1304 } 1305 1306 //We should not be popping out the context here in endELement becaause the namespace context is still 1307 //valid when parser is at the endElement state. 1308 //if (fNamespaces) { 1309 // fNamespaceContext.popContext(); 1310 //} 1311 1312 //pop the element off the stack.. 1313 fElementStack.popElement(); 1314 1315 } else { 1316 1317 if(dtdGrammarUtil != null) 1318 dtdGrammarUtil.startElement(fElementQName, fAttributes); 1319 if(fDocumentHandler != null){ 1320 //complete element and attributes are traversed in this function so we can send a callback 1321 //here. 1322 //<strong>we shouldn't be sending callback in scanDocument()</strong> 1323 fDocumentHandler.startElement(fElementQName, fAttributes, null); 1324 } 1325 } 1326 1327 1328 if (DEBUG_START_END_ELEMENT) System.out.println(this.getClass().toString() + "<<< scanStartElement(): "+fEmptyElement); 1329 return fEmptyElement; 1330 1331 } // scanStartElement():boolean 1332 1333 /** 1334 * Looks for the close of start tag, i.e. if it finds '>' or '/>' 1335 * Characters are consumed. 1336 */ 1337 protected boolean seekCloseOfStartTag() throws IOException, XNIException { 1338 // spaces 1339 boolean sawSpace = fEntityScanner.skipSpaces(); 1340 1341 // end tag? 1342 final int c = fEntityScanner.peekChar(); 1343 if (c == '>') { 1344 fEntityScanner.scanChar(); 1345 return true; 1346 } else if (c == '/') { 1347 fEntityScanner.scanChar(); 1348 if (!fEntityScanner.skipChar('>')) { 1349 reportFatalError("ElementUnterminated", 1350 new Object[]{fElementQName.rawname}); 1351 } 1352 fEmptyElement = true; 1353 return true; 1354 } else if (!isValidNameStartChar(c) || !sawSpace) { 1355 reportFatalError("ElementUnterminated", new Object[]{fElementQName.rawname}); 1356 } 1357 1358 return false; 1359 } 1360 1361 public boolean hasAttributes(){ 1362 return fAttributes.getLength() > 0 ? true : false ; 1363 } 1364 1365 1366 /** 1367 * Scans an attribute. 1368 * <p> 1369 * <pre> 1370 * [41] Attribute ::= Name Eq AttValue 1371 * </pre> 1372 * <p> 1373 * <strong>Note:</strong> This method assumes that the next 1374 * character on the stream is the first character of the attribute 1375 * name. 1376 * <p> 1377 * <strong>Note:</strong> This method uses the fAttributeQName and 1378 * fQName variables. The contents of these variables will be 1379 * destroyed. 1380 * 1381 * @param attributes The attributes list for the scanned attribute. 1382 */ 1383 1384 /** 1385 * protected void scanAttribute(AttributeIteratorImpl attributes) 1386 * throws IOException, XNIException { 1387 * if (DEBUG_START_END_ELEMENT) System.out.println(">>> scanAttribute()"); 1388 * 1389 * 1390 * // name 1391 * if (fNamespaces) { 1392 * fEntityScanner.scanQName(fAttributeQName); 1393 * } 1394 * else { 1395 * String name = fEntityScanner.scanName(); 1396 * fAttributeQName.setValues(null, name, name, null); 1397 * } 1398 * 1399 * // equals 1400 * fEntityScanner.skipSpaces(); 1401 * if (!fEntityScanner.skipChar('=')) { 1402 * reportFatalError("EqRequiredInAttribute", 1403 * new Object[]{fAttributeQName.rawname}); 1404 * } 1405 * fEntityScanner.skipSpaces(); 1406 * 1407 * 1408 * // content 1409 * int oldLen = attributes.getLength(); 1410 */ 1411 /**xxx there is one check of duplicate attribute that has been removed. 1412 * attributes.addAttribute(fAttributeQName, XMLSymbols.fCDATASymbol, null); 1413 * 1414 * // WFC: Unique Att Spec 1415 * if (oldLen == attributes.getLength()) { 1416 * reportFatalError("AttributeNotUnique", 1417 * new Object[]{fCurrentElement.rawname, 1418 * fAttributeQName.rawname}); 1419 * } 1420 */ 1421 1422 /* 1423 //REVISIT: one more case needs to be included: external PE and standalone is no 1424 boolean isVC = fHasExternalDTD && !fStandalone; 1425 scanAttributeValue(fTempString, fTempString2, 1426 fAttributeQName.rawname, attributes, 1427 oldLen, isVC); 1428 1429 //attributes.setValue(oldLen, fTempString.toString()); 1430 //attributes.setNonNormalizedValue(oldLen, fTempString2.toString()); 1431 //attributes.setSpecified(oldLen, true); 1432 1433 AttributeImpl attribute = new AttributeImpl(fAttributeQName.prefix,fAttributeQName.localpart,fAttributeQName.uri,fTempString.toString(),fTempString2.toString(),XMLSymbols.fCDATASymbol,true); 1434 fAttributes.addAttribute(attribute); 1435 if (DEBUG_START_END_ELEMENT) System.out.println("<<< scanAttribute()"); 1436 } // scanAttribute(XMLAttributes) 1437 1438 */ 1439 1440 /** return the attribute iterator implementation */ 1441 public XMLAttributesIteratorImpl getAttributeIterator(){ 1442 if(dtdGrammarUtil != null && fAddDefaultAttr){ 1443 dtdGrammarUtil.addDTDDefaultAttrs(fElementQName,fAttributes); 1444 fAddDefaultAttr = false; 1445 } 1446 return fAttributes; 1447 } 1448 1449 /** return if standalone is set */ 1450 public boolean standaloneSet(){ 1451 return fStandaloneSet; 1452 } 1453 /** return if the doucment is standalone */ 1454 public boolean isStandAlone(){ 1455 return fStandalone ; 1456 } 1457 /** 1458 * Scans an attribute name value pair. 1459 * <p> 1460 * <pre> 1461 * [41] Attribute ::= Name Eq AttValue 1462 * </pre> 1463 * <p> 1464 * <strong>Note:</strong> This method assumes that the next 1465 * character on the stream is the first character of the attribute 1466 * name. 1467 * <p> 1468 * <strong>Note:</strong> This method uses the fAttributeQName and 1469 * fQName variables. The contents of these variables will be 1470 * destroyed. 1471 * 1472 * @param attributes The attributes list for the scanned attribute. 1473 */ 1474 1475 protected void scanAttribute(XMLAttributes attributes) 1476 throws IOException, XNIException { 1477 if (DEBUG_START_END_ELEMENT) System.out.println(this.getClass().toString() +">>> scanAttribute()"); 1478 1479 // name 1480 if (fNamespaces) { 1481 fEntityScanner.scanQName(fAttributeQName); 1482 } else { 1483 String name = fEntityScanner.scanName(); 1484 fAttributeQName.setValues(null, name, name, null); 1485 } 1486 1487 // equals 1488 fEntityScanner.skipSpaces(); 1489 if (!fEntityScanner.skipChar('=')) { 1490 reportFatalError("EqRequiredInAttribute", 1491 new Object[] {fCurrentElement.rawname, fAttributeQName.rawname}); 1492 } 1493 fEntityScanner.skipSpaces(); 1494 1495 int attIndex = 0 ; 1496 //REVISIT: one more case needs to be included: external PE and standalone is no 1497 boolean isVC = fHasExternalDTD && !fStandalone; 1498 //fTempString would store attribute value 1499 ///fTempString2 would store attribute non-normalized value 1500 1501 //this function doesn't use 'attIndex'. We are adding the attribute later 1502 //after we have figured out that current attribute is not namespace declaration 1503 //since scanAttributeValue doesn't use attIndex parameter therefore we 1504 //can safely add the attribute later.. 1505 XMLString tmpStr = getString(); 1506 1507 scanAttributeValue(tmpStr, fTempString2, 1508 fAttributeQName.rawname, attributes, 1509 attIndex, isVC); 1510 1511 // content 1512 int oldLen = attributes.getLength(); 1513 //if the attribute name already exists.. new value is replaced with old value 1514 attIndex = attributes.addAttribute(fAttributeQName, XMLSymbols.fCDATASymbol, null); 1515 1516 // WFC: Unique Att Spec 1517 //attributes count will be same if the current attribute name already exists for this element name. 1518 //this means there are two duplicate attributes. 1519 if (oldLen == attributes.getLength()) { 1520 reportFatalError("AttributeNotUnique", 1521 new Object[]{fCurrentElement.rawname, 1522 fAttributeQName.rawname}); 1523 } 1524 1525 //tmpString contains attribute value 1526 //we are passing null as the attribute value 1527 attributes.setValue(attIndex, null, tmpStr); 1528 1529 ///xxx: nonNormalizedValue is not being set as it is not required by SAX & DOM 1530 //attributes.setNonNormalizedValue(oldLen, fTempString2.toString()); 1531 attributes.setSpecified(attIndex, true); 1532 1533 if (DEBUG_START_END_ELEMENT) System.out.println(this.getClass().toString() +"<<< scanAttribute()"); 1534 1535 } // scanAttribute(XMLAttributes) 1536 1537 /** 1538 * Scans element content. 1539 * 1540 * @return Returns the next character on the stream. 1541 */ 1542 //CHANGED: 1543 //EARLIER: scanContent() 1544 //NOW: scanContent(XMLStringBuffer) 1545 //It makes things easy if this functions takes XMLStringBuffer as parameter.. 1546 //this function appends the data to the buffer. 1547 protected int scanContent(XMLStringBuffer content) throws IOException, XNIException { 1548 //set the fTempString length to 0 before passing it on to scanContent 1549 //scanContent sets the correct co-ordinates as per the content read 1550 fTempString.length = 0; 1551 int c = fEntityScanner.scanContent(fTempString); 1552 content.append(fTempString); 1553 fTempString.length = 0; 1554 if (c == '\r') { 1555 // happens when there is the character reference 1556 //xxx: We know the next chracter.. we should just skip it and add ']' directlry 1557 fEntityScanner.scanChar(); 1558 content.append((char)c); 1559 c = -1; 1560 } else if (c == ']') { 1561 //fStringBuffer.clear(); 1562 //xxx: We know the next chracter.. we should just skip it and add ']' directlry 1563 content.append((char)fEntityScanner.scanChar()); 1564 // remember where we are in case we get an endEntity before we 1565 // could flush the buffer out - this happens when we're parsing an 1566 // entity which ends with a ] 1567 fInScanContent = true; 1568 // 1569 // We work on a single character basis to handle cases such as: 1570 // ']]]>' which we might otherwise miss. 1571 // 1572 if (fEntityScanner.skipChar(']')) { 1573 content.append(']'); 1574 while (fEntityScanner.skipChar(']')) { 1575 content.append(']'); 1576 } 1577 if (fEntityScanner.skipChar('>')) { 1578 reportFatalError("CDEndInContent", null); 1579 } 1580 } 1581 fInScanContent = false; 1582 c = -1; 1583 } 1584 if (fDocumentHandler != null && content.length > 0) { 1585 //fDocumentHandler.characters(content, null); 1586 } 1587 return c; 1588 1589 } // scanContent():int 1590 1591 1592 /** 1593 * Scans a CDATA section. 1594 * <p> 1595 * <strong>Note:</strong> This method uses the fTempString and 1596 * fStringBuffer variables. 1597 * 1598 * @param complete True if the CDATA section is to be scanned 1599 * completely. 1600 * 1601 * @return True if CDATA is completely scanned. 1602 */ 1603 //CHANGED: 1604 protected boolean scanCDATASection(XMLStringBuffer contentBuffer, boolean complete) 1605 throws IOException, XNIException { 1606 1607 // call handler 1608 if (fDocumentHandler != null) { 1609 //fDocumentHandler.startCDATA(null); 1610 } 1611 1612 while (true) { 1613 //scanData will fill the contentBuffer 1614 if (!fEntityScanner.scanData("]]>", contentBuffer)) { 1615 break ; 1616 /** We dont need all this code if we pass ']]>' as delimeter.. 1617 * int brackets = 2; 1618 * while (fEntityScanner.skipChar(']')) { 1619 * brackets++; 1620 * } 1621 * 1622 * //When we find more than 2 square brackets 1623 * if (fDocumentHandler != null && brackets > 2) { 1624 * //we dont need to clear the buffer.. 1625 * //contentBuffer.clear(); 1626 * for (int i = 2; i < brackets; i++) { 1627 * contentBuffer.append(']'); 1628 * } 1629 * fDocumentHandler.characters(contentBuffer, null); 1630 * } 1631 * 1632 * if (fEntityScanner.skipChar('>')) { 1633 * break; 1634 * } 1635 * if (fDocumentHandler != null) { 1636 * //we dont need to clear the buffer now.. 1637 * //contentBuffer.clear(); 1638 * contentBuffer.append("]]"); 1639 * fDocumentHandler.characters(contentBuffer, null); 1640 * } 1641 **/ 1642 } else { 1643 int c = fEntityScanner.peekChar(); 1644 if (c != -1 && isInvalidLiteral(c)) { 1645 if (XMLChar.isHighSurrogate(c)) { 1646 //contentBuffer.clear(); 1647 //scan surrogates if any.... 1648 scanSurrogates(contentBuffer); 1649 } else { 1650 reportFatalError("InvalidCharInCDSect", 1651 new Object[]{Integer.toString(c,16)}); 1652 fEntityScanner.scanChar(); 1653 } 1654 } 1655 //by this time we have also read surrogate contents if any... 1656 if (fDocumentHandler != null) { 1657 //fDocumentHandler.characters(contentBuffer, null); 1658 } 1659 } 1660 } 1661 fMarkupDepth--; 1662 1663 if (fDocumentHandler != null && contentBuffer.length > 0) { 1664 //fDocumentHandler.characters(contentBuffer, null); 1665 } 1666 1667 // call handler 1668 if (fDocumentHandler != null) { 1669 //fDocumentHandler.endCDATA(null); 1670 } 1671 1672 return true; 1673 1674 } // scanCDATASection(XMLStringBuffer, boolean):boolean 1675 1676 /** 1677 * Scans an end element. 1678 * <p> 1679 * <pre> 1680 * [42] ETag ::= '</' Name S? '>' 1681 * </pre> 1682 * <p> 1683 * <strong>Note:</strong> This method uses the fElementQName variable. 1684 * The contents of this variable will be destroyed. The caller should 1685 * copy the needed information out of this variable before calling 1686 * this method. 1687 * 1688 * @return The element depth. 1689 */ 1690 protected int scanEndElement() throws IOException, XNIException { 1691 if (DEBUG_START_END_ELEMENT) System.out.println(this.getClass().toString() +">>> scanEndElement()"); 1692 1693 // pop context 1694 QName endElementName = fElementStack.popElement(); 1695 1696 String rawname = endElementName.rawname; 1697 if(DEBUG)System.out.println("endElementName = " + endElementName.toString()); 1698 // Take advantage of the fact that next string _should_ be "fElementQName.rawName", 1699 //In scanners most of the time is consumed on checks done for XML characters, we can 1700 // optimize on it and avoid the checks done for endElement, 1701 //we will also avoid symbol table lookup - neeraj.bajaj@sun.com 1702 1703 // this should work both for namespace processing true or false... 1704 1705 //REVISIT: if the string is not the same as expected.. we need to do better error handling.. 1706 //We can skip this for now... In any case if the string doesn't match -- document is not well formed. 1707 1708 if (!fEntityScanner.skipString(endElementName.rawname)) { 1709 reportFatalError("ETagRequired", new Object[]{rawname}); 1710 } 1711 1712 // end 1713 fEntityScanner.skipSpaces(); 1714 if (!fEntityScanner.skipChar('>')) { 1715 reportFatalError("ETagUnterminated", 1716 new Object[]{rawname}); 1717 } 1718 fMarkupDepth--; 1719 1720 //we have increased the depth for two markup "<" characters 1721 fMarkupDepth--; 1722 1723 // check that this element was opened in the same entity 1724 if (fMarkupDepth < fEntityStack[fEntityDepth - 1]) { 1725 reportFatalError("ElementEntityMismatch", 1726 new Object[]{rawname}); 1727 } 1728 1729 //We should not be popping out the context here in endELement becaause the namespace context is still 1730 //valid when parser is at the endElement state. 1731 1732 //if (fNamespaces) { 1733 // fNamespaceContext.popContext(); 1734 //} 1735 1736 // call handler 1737 if (fDocumentHandler != null ) { 1738 //end element is scanned in this function so we can send a callback 1739 //here. 1740 //<strong>we shouldn't be sending callback in scanDocument()</strong> 1741 1742 fDocumentHandler.endElement(endElementName, null); 1743 } 1744 if(dtdGrammarUtil != null) 1745 dtdGrammarUtil.endElement(endElementName); 1746 1747 return fMarkupDepth; 1748 1749 } // scanEndElement():int 1750 1751 /** 1752 * Scans a character reference. 1753 * <p> 1754 * <pre> 1755 * [66] CharRef ::= '&#' [0-9]+ ';' | '&#x' [0-9a-fA-F]+ ';' 1756 * </pre> 1757 */ 1758 protected void scanCharReference() 1759 throws IOException, XNIException { 1760 1761 fStringBuffer2.clear(); 1762 int ch = scanCharReferenceValue(fStringBuffer2, null); 1763 fMarkupDepth--; 1764 if (ch != -1) { 1765 // call handler 1766 1767 if (fDocumentHandler != null) { 1768 if (fNotifyCharRefs) { 1769 fDocumentHandler.startGeneralEntity(fCharRefLiteral, null, null, null); 1770 } 1771 Augmentations augs = null; 1772 if (fValidation && ch <= 0x20) { 1773 if (fTempAugmentations != null) { 1774 fTempAugmentations.removeAllItems(); 1775 } 1776 else { 1777 fTempAugmentations = new AugmentationsImpl(); 1778 } 1779 augs = fTempAugmentations; 1780 augs.putItem(Constants.CHAR_REF_PROBABLE_WS, Boolean.TRUE); 1781 } 1782 //xxx: How do we deal with this - how to return charReferenceValues 1783 //now this is being commented because this is taken care in scanDocument() 1784 //fDocumentHandler.characters(fStringBuffer2, null); 1785 if (fNotifyCharRefs) { 1786 fDocumentHandler.endGeneralEntity(fCharRefLiteral, null); 1787 } 1788 } 1789 } 1790 1791 } // scanCharReference() 1792 1793 1794 /** 1795 * Scans an entity reference. 1796 * 1797 * @return returns true if the new entity is started. If it was built-in entity 1798 * 'false' is returned. 1799 * @throws IOException Thrown if i/o error occurs. 1800 * @throws XNIException Thrown if handler throws exception upon 1801 * notification. 1802 */ 1803 protected void scanEntityReference(XMLStringBuffer content) throws IOException, XNIException { 1804 String name = fEntityScanner.scanName(); 1805 if (name == null) { 1806 reportFatalError("NameRequiredInReference", null); 1807 return; 1808 } 1809 if (!fEntityScanner.skipChar(';')) { 1810 reportFatalError("SemicolonRequiredInReference", new Object []{name}); 1811 } 1812 if (fEntityStore.isUnparsedEntity(name)) { 1813 reportFatalError("ReferenceToUnparsedEntity", new Object[]{name}); 1814 } 1815 fMarkupDepth--; 1816 fCurrentEntityName = name; 1817 1818 // handle built-in entities 1819 if (name == fAmpSymbol) { 1820 handleCharacter('&', fAmpSymbol, content); 1821 fScannerState = SCANNER_STATE_BUILT_IN_REFS; 1822 return ; 1823 } else if (name == fLtSymbol) { 1824 handleCharacter('<', fLtSymbol, content); 1825 fScannerState = SCANNER_STATE_BUILT_IN_REFS; 1826 return ; 1827 } else if (name == fGtSymbol) { 1828 handleCharacter('>', fGtSymbol, content); 1829 fScannerState = SCANNER_STATE_BUILT_IN_REFS; 1830 return ; 1831 } else if (name == fQuotSymbol) { 1832 handleCharacter('"', fQuotSymbol, content); 1833 fScannerState = SCANNER_STATE_BUILT_IN_REFS; 1834 return ; 1835 } else if (name == fAposSymbol) { 1836 handleCharacter('\'', fAposSymbol, content); 1837 fScannerState = SCANNER_STATE_BUILT_IN_REFS; 1838 return ; 1839 } 1840 1841 //1. if the entity is external and support to external entities is not required 1842 // 2. or entities should not be replaced 1843 //3. or if it is built in entity reference. 1844 if((fEntityStore.isExternalEntity(name) && !fSupportExternalEntities) || (!fEntityStore.isExternalEntity(name) && !fReplaceEntityReferences) || foundBuiltInRefs){ 1845 fScannerState = SCANNER_STATE_REFERENCE; 1846 return ; 1847 } 1848 // start general entity 1849 if (!fEntityStore.isDeclaredEntity(name)) { 1850 //SUPPORT_DTD=false && ReplaceEntityReferences should throw exception 1851 if (!fSupportDTD && fReplaceEntityReferences) { 1852 reportFatalError("EntityNotDeclared", new Object[]{name}); 1853 return; 1854 } 1855 //REVISIT: one more case needs to be included: external PE and standalone is no 1856 if ( fHasExternalDTD && !fStandalone) { 1857 if (fValidation) 1858 fErrorReporter.reportError(fEntityScanner, XMLMessageFormatter.XML_DOMAIN,"EntityNotDeclared", 1859 new Object[]{name}, XMLErrorReporter.SEVERITY_ERROR); 1860 } else 1861 reportFatalError("EntityNotDeclared", new Object[]{name}); 1862 } 1863 //we are starting the entity even if the entity was not declared 1864 //if that was the case it its taken care in XMLEntityManager.startEntity() 1865 //we immediately call the endEntity. Application gets to know if there was 1866 //any entity that was not declared. 1867 fEntityManager.startEntity(name, false); 1868 //set the scaner state to content.. parser will automatically revive itself at any point of time. 1869 //setScannerState(SCANNER_STATE_CONTENT); 1870 //return true ; 1871 } // scanEntityReference() 1872 1873 // utility methods 1874 1875 /** 1876 * Calls document handler with a single character resulting from 1877 * built-in entity resolution. 1878 * 1879 * @param c 1880 * @param entity built-in name 1881 * @param XMLStringBuffer append the character to buffer 1882 * 1883 * we really dont need to call this function -- this function is only required when 1884 * we integrate with rest of Xerces2. SO maintaining the current behavior and still 1885 * calling this function to hanlde built-in entity reference. 1886 * 1887 */ 1888 private void handleCharacter(char c, String entity, XMLStringBuffer content) throws XNIException { 1889 foundBuiltInRefs = true; 1890 content.append(c); 1891 if (fDocumentHandler != null) { 1892 fSingleChar[0] = c; 1893 if (fNotifyBuiltInRefs) { 1894 fDocumentHandler.startGeneralEntity(entity, null, null, null); 1895 } 1896 fTempString.setValues(fSingleChar, 0, 1); 1897 //fDocumentHandler.characters(fTempString, null); 1898 1899 if (fNotifyBuiltInRefs) { 1900 fDocumentHandler.endGeneralEntity(entity, null); 1901 } 1902 } 1903 } // handleCharacter(char) 1904 1905 // helper methods 1906 1907 /** 1908 * Sets the scanner state. 1909 * 1910 * @param state The new scanner state. 1911 */ 1912 protected final void setScannerState(int state) { 1913 1914 fScannerState = state; 1915 if (DEBUG_SCANNER_STATE) { 1916 System.out.print("### setScannerState: "); 1917 //System.out.print(fScannerState); 1918 System.out.print(getScannerStateName(state)); 1919 System.out.println(); 1920 } 1921 1922 } // setScannerState(int) 1923 1924 1925 /** 1926 * Sets the Driver. 1927 * 1928 * @param Driver The new Driver. 1929 */ 1930 protected final void setDriver(Driver driver) { 1931 fDriver = driver; 1932 if (DEBUG_DISPATCHER) { 1933 System.out.print("%%% setDriver: "); 1934 System.out.print(getDriverName(driver)); 1935 System.out.println(); 1936 } 1937 } 1938 1939 // 1940 // Private methods 1941 // 1942 1943 /** Returns the scanner state name. */ 1944 protected String getScannerStateName(int state) { 1945 1946 switch (state) { 1947 case SCANNER_STATE_DOCTYPE: return "SCANNER_STATE_DOCTYPE"; 1948 case SCANNER_STATE_ROOT_ELEMENT: return "SCANNER_STATE_ROOT_ELEMENT"; 1949 case SCANNER_STATE_START_OF_MARKUP: return "SCANNER_STATE_START_OF_MARKUP"; 1950 case SCANNER_STATE_COMMENT: return "SCANNER_STATE_COMMENT"; 1951 case SCANNER_STATE_PI: return "SCANNER_STATE_PI"; 1952 case SCANNER_STATE_CONTENT: return "SCANNER_STATE_CONTENT"; 1953 case SCANNER_STATE_REFERENCE: return "SCANNER_STATE_REFERENCE"; 1954 case SCANNER_STATE_END_OF_INPUT: return "SCANNER_STATE_END_OF_INPUT"; 1955 case SCANNER_STATE_TERMINATED: return "SCANNER_STATE_TERMINATED"; 1956 case SCANNER_STATE_CDATA: return "SCANNER_STATE_CDATA"; 1957 case SCANNER_STATE_TEXT_DECL: return "SCANNER_STATE_TEXT_DECL"; 1958 case SCANNER_STATE_ATTRIBUTE: return "SCANNER_STATE_ATTRIBUTE"; 1959 case SCANNER_STATE_ATTRIBUTE_VALUE: return "SCANNER_STATE_ATTRIBUTE_VALUE"; 1960 case SCANNER_STATE_START_ELEMENT_TAG: return "SCANNER_STATE_START_ELEMENT_TAG"; 1961 case SCANNER_STATE_END_ELEMENT_TAG: return "SCANNER_STATE_END_ELEMENT_TAG"; 1962 case SCANNER_STATE_CHARACTER_DATA: return "SCANNER_STATE_CHARACTER_DATA" ; 1963 } 1964 1965 return "??? ("+state+')'; 1966 1967 } // getScannerStateName(int):String 1968 public String getEntityName(){ 1969 //return the cached name 1970 return fCurrentEntityName; 1971 } 1972 1973 /** Returns the driver name. */ 1974 public String getDriverName(Driver driver) { 1975 1976 if (DEBUG_DISPATCHER) { 1977 if (driver != null) { 1978 String name = driver.getClass().getName(); 1979 int index = name.lastIndexOf('.'); 1980 if (index != -1) { 1981 name = name.substring(index + 1); 1982 index = name.lastIndexOf('$'); 1983 if (index != -1) { 1984 name = name.substring(index + 1); 1985 } 1986 } 1987 return name; 1988 } 1989 } 1990 return "null"; 1991 1992 } // getDriverName():String 1993 1994 // 1995 // Classes 1996 // 1997 1998 /** 1999 * @author Neeraj Bajaj, Sun Microsystems. 2000 */ 2001 protected static final class Element { 2002 2003 // 2004 // Data 2005 // 2006 2007 /** Symbol. */ 2008 public QName qname; 2009 2010 //raw name stored as characters 2011 public char[] fRawname; 2012 2013 /** The next Element entry. */ 2014 public Element next; 2015 2016 // 2017 // Constructors 2018 // 2019 2020 /** 2021 * Constructs a new Element from the given QName and next Element 2022 * reference. 2023 */ 2024 public Element(QName qname, Element next) { 2025 this.qname.setValues(qname); 2026 this.fRawname = qname.rawname.toCharArray(); 2027 this.next = next; 2028 } 2029 2030 } // class Element 2031 2032 /** 2033 * Element stack. 2034 * 2035 * @author Neeraj Bajaj, Sun Microsystems. 2036 */ 2037 protected class ElementStack2 { 2038 2039 // 2040 // Data 2041 // 2042 2043 /** The stack data. */ 2044 protected QName [] fQName = new QName[20]; 2045 2046 //Element depth 2047 protected int fDepth; 2048 //total number of elements 2049 protected int fCount; 2050 //current position 2051 protected int fPosition; 2052 //Mark refers to the position 2053 protected int fMark; 2054 2055 protected int fLastDepth ; 2056 2057 // 2058 // Constructors 2059 // 2060 2061 /** Default constructor. */ 2062 public ElementStack2() { 2063 for (int i = 0; i < fQName.length; i++) { 2064 fQName[i] = new QName(); 2065 } 2066 fMark = fPosition = 1; 2067 } // <init>() 2068 2069 public void resize(){ 2070 /** 2071 * int length = fElements.length; 2072 * Element [] temp = new Element[length * 2]; 2073 * System.arraycopy(fElements, 0, temp, 0, length); 2074 * fElements = temp; 2075 */ 2076 //resize QNames 2077 int oldLength = fQName.length; 2078 QName [] tmp = new QName[oldLength * 2]; 2079 System.arraycopy(fQName, 0, tmp, 0, oldLength); 2080 fQName = tmp; 2081 2082 for (int i = oldLength; i < fQName.length; i++) { 2083 fQName[i] = new QName(); 2084 } 2085 2086 } 2087 2088 2089 // 2090 // Public methods 2091 // 2092 2093 /** Check if the element scanned during the start element 2094 *matches the stored element. 2095 * 2096 *@return true if the match suceeds. 2097 */ 2098 public boolean matchElement(QName element) { 2099 //last depth is the depth when last elemnt was pushed 2100 //if last depth is greater than current depth 2101 if(DEBUG_SKIP_ALGORITHM){ 2102 System.out.println("fLastDepth = " + fLastDepth); 2103 System.out.println("fDepth = " + fDepth); 2104 } 2105 boolean match = false; 2106 if(fLastDepth > fDepth && fDepth <= 2){ 2107 if(DEBUG_SKIP_ALGORITHM){ 2108 System.out.println("Checking if the elements match " + element.rawname + " , " + fQName[fDepth].rawname); 2109 } 2110 if(element.rawname == fQName[fDepth].rawname){ 2111 fAdd = false; 2112 //mark this position 2113 //decrease the depth by 1 as arrays are 0 based 2114 fMark = fDepth - 1; 2115 //we found the match and from next element skipping will start, add 1 2116 fPosition = fMark + 1 ; 2117 match = true; 2118 //Once we get match decrease the count -- this was increased by nextElement() 2119 --fCount; 2120 if(DEBUG_SKIP_ALGORITHM){ 2121 System.out.println("fAdd FALSE -- NOW ELEMENT SHOULD NOT BE ADDED"); 2122 System.out.println("fMark = " + fMark); 2123 System.out.println("fPosition = " + fPosition); 2124 System.out.println("fDepth = " + fDepth); 2125 System.out.println("fCount = " + fCount); 2126 } 2127 }else{ 2128 fAdd = true; 2129 if(DEBUG_SKIP_ALGORITHM)System.out.println("fAdd is " + fAdd); 2130 } 2131 } 2132 //store the last depth 2133 fLastDepth = fDepth++; 2134 return match; 2135 } // pushElement(QName):QName 2136 2137 /** 2138 * This function doesn't increase depth. The function in this function is 2139 *broken down into two functions for efficiency. <@see>matchElement</see>. 2140 * This function just returns the pointer to the object and its values are set. 2141 * 2142 *@return QName reference to the next element in the list 2143 */ 2144 public QName nextElement() { 2145 2146 //if number of elements becomes equal to the length of array -- stop the skipping 2147 if (fCount == fQName.length) { 2148 fShouldSkip = false; 2149 fAdd = false; 2150 if(DEBUG_SKIP_ALGORITHM)System.out.println("SKIPPING STOPPED, fShouldSkip = " + fShouldSkip); 2151 //xxx: this is not correct, we are returning the last element 2152 //this wont make any difference since flag has been set to 'false' 2153 return fQName[--fCount]; 2154 } 2155 if(DEBUG_SKIP_ALGORITHM){ 2156 System.out.println("fCount = " + fCount); 2157 } 2158 return fQName[fCount++]; 2159 2160 } 2161 2162 /** Note that this function is considerably different than nextElement() 2163 * This function just returns the previously stored elements 2164 */ 2165 public QName getNext(){ 2166 //when position reaches number of elements in the list.. 2167 //set the position back to mark, making it a circular linked list. 2168 if(fPosition == fCount){ 2169 fPosition = fMark; 2170 } 2171 return fQName[fPosition++]; 2172 } 2173 2174 /** returns the current depth 2175 */ 2176 public int popElement(){ 2177 return fDepth--; 2178 } 2179 2180 2181 /** Clears the stack without throwing away existing QName objects. */ 2182 public void clear() { 2183 fLastDepth = 0; 2184 fDepth = 0; 2185 fCount = 0 ; 2186 fPosition = fMark = 1; 2187 } // clear() 2188 2189 } // class ElementStack 2190 2191 /** 2192 * Element stack. This stack operates without synchronization, error 2193 * checking, and it re-uses objects instead of throwing popped items 2194 * away. 2195 * 2196 * @author Andy Clark, IBM 2197 */ 2198 protected class ElementStack { 2199 2200 // 2201 // Data 2202 // 2203 2204 /** The stack data. */ 2205 protected QName[] fElements; 2206 protected int [] fInt = new int[20]; 2207 2208 2209 //Element depth 2210 protected int fDepth; 2211 //total number of elements 2212 protected int fCount; 2213 //current position 2214 protected int fPosition; 2215 //Mark refers to the position 2216 protected int fMark; 2217 2218 protected int fLastDepth ; 2219 2220 // 2221 // Constructors 2222 // 2223 2224 /** Default constructor. */ 2225 public ElementStack() { 2226 fElements = new QName[20]; 2227 for (int i = 0; i < fElements.length; i++) { 2228 fElements[i] = new QName(); 2229 } 2230 } // <init>() 2231 2232 // 2233 // Public methods 2234 // 2235 2236 /** 2237 * Pushes an element on the stack. 2238 * <p> 2239 * <strong>Note:</strong> The QName values are copied into the 2240 * stack. In other words, the caller does <em>not</em> orphan 2241 * the element to the stack. Also, the QName object returned 2242 * is <em>not</em> orphaned to the caller. It should be 2243 * considered read-only. 2244 * 2245 * @param element The element to push onto the stack. 2246 * 2247 * @return Returns the actual QName object that stores the 2248 */ 2249 //XXX: THIS FUNCTION IS NOT USED 2250 public QName pushElement(QName element) { 2251 if (fDepth == fElements.length) { 2252 QName[] array = new QName[fElements.length * 2]; 2253 System.arraycopy(fElements, 0, array, 0, fDepth); 2254 fElements = array; 2255 for (int i = fDepth; i < fElements.length; i++) { 2256 fElements[i] = new QName(); 2257 } 2258 } 2259 fElements[fDepth].setValues(element); 2260 return fElements[fDepth++]; 2261 } // pushElement(QName):QName 2262 2263 2264 /** Note that this function is considerably different than nextElement() 2265 * This function just returns the previously stored elements 2266 */ 2267 public QName getNext(){ 2268 //when position reaches number of elements in the list.. 2269 //set the position back to mark, making it a circular linked list. 2270 if(fPosition == fCount){ 2271 fPosition = fMark; 2272 } 2273 //store the position of last opened tag at particular depth 2274 //fInt[++fDepth] = fPosition; 2275 if(DEBUG_SKIP_ALGORITHM){ 2276 System.out.println("Element at fPosition = " + fPosition + " is " + fElements[fPosition].rawname); 2277 } 2278 //return fElements[fPosition++]; 2279 return fElements[fPosition]; 2280 } 2281 2282 /** This function should be called only when element was skipped sucessfully. 2283 * 1. Increase the depth - because element was sucessfully skipped. 2284 *2. Store the position of the element token in array "last opened tag" at depth. 2285 *3. increase the position counter so as to point to the next element in the array 2286 */ 2287 public void push(){ 2288 2289 fInt[++fDepth] = fPosition++; 2290 } 2291 2292 /** Check if the element scanned during the start element 2293 *matches the stored element. 2294 * 2295 *@return true if the match suceeds. 2296 */ 2297 public boolean matchElement(QName element) { 2298 //last depth is the depth when last elemnt was pushed 2299 //if last depth is greater than current depth 2300 //if(DEBUG_SKIP_ALGORITHM){ 2301 // System.out.println("Check if the element " + element.rawname + " matches"); 2302 // System.out.println("fLastDepth = " + fLastDepth); 2303 // System.out.println("fDepth = " + fDepth); 2304 //} 2305 boolean match = false; 2306 if(fLastDepth > fDepth && fDepth <= 3){ 2307 if(DEBUG_SKIP_ALGORITHM){ 2308 System.out.println("----------ENTERED THE LOOP WHERE WE CHECK FOR MATCHING OF ELMENT-----"); 2309 System.out.println("Depth = " + fDepth + " Checking if INCOMING element " + element.rawname + " match STORED ELEMENT " + fElements[fDepth - 1].rawname); 2310 } 2311 if(element.rawname == fElements[fDepth - 1].rawname){ 2312 fAdd = false; 2313 //mark this position 2314 //decrease the depth by 1 as arrays are 0 based 2315 fMark = fDepth - 1; 2316 //we found the match 2317 fPosition = fMark; 2318 match = true; 2319 //Once we get match decrease the count -- this was increased by nextElement() 2320 --fCount; 2321 if(DEBUG_SKIP_ALGORITHM){ 2322 System.out.println("NOW ELEMENT SHOULD NOT BE ADDED, fAdd is set to false"); 2323 System.out.println("fMark = " + fMark); 2324 System.out.println("fPosition = " + fPosition); 2325 System.out.println("fDepth = " + fDepth); 2326 System.out.println("fCount = " + fCount); 2327 System.out.println("---------MATCH SUCEEDED-----------------"); 2328 System.out.println(""); 2329 } 2330 }else{ 2331 fAdd = true; 2332 if(DEBUG_SKIP_ALGORITHM)System.out.println("fAdd is " + fAdd); 2333 } 2334 } 2335 //store the position for the current depth 2336 //when we are adding the elements, when skipping 2337 //starts even then this should be tracked ie. when 2338 //calling getNext() 2339 if(match){ 2340 //from next element skipping will start, add 1 2341 fInt[fDepth] = fPosition++; 2342 } else{ 2343 if(DEBUG_SKIP_ALGORITHM){ 2344 System.out.println("At depth = " + fDepth + "array position is = " + (fCount - 1)); 2345 } 2346 //sicne fInt[fDepth] contains pointer to the element array which are 0 based. 2347 fInt[fDepth] = fCount - 1; 2348 } 2349 2350 //if number of elements becomes equal to the length of array -- stop the skipping 2351 //xxx: should we do "fCount == fInt.length" 2352 if (fCount == fElements.length) { 2353 fSkip = false; 2354 fAdd = false; 2355 //reposition the stack -- it seems to be too complex document and there is no symmerty in structure 2356 reposition(); 2357 if(DEBUG_SKIP_ALGORITHM){ 2358 System.out.println("ALL THE ELMENTS IN ARRAY HAVE BEEN FILLED"); 2359 System.out.println("REPOSITIONING THE STACK"); 2360 System.out.println("-----------SKIPPING STOPPED----------"); 2361 System.out.println(""); 2362 } 2363 return false; 2364 } 2365 if(DEBUG_SKIP_ALGORITHM){ 2366 if(match){ 2367 System.out.println("Storing fPosition = " + fInt[fDepth] + " at fDepth = " + fDepth); 2368 }else{ 2369 System.out.println("Storing fCount = " + fInt[fDepth] + " at fDepth = " + fDepth); 2370 } 2371 } 2372 //store the last depth 2373 fLastDepth = fDepth; 2374 return match; 2375 } // matchElement(QName):QName 2376 2377 2378 /** 2379 * Returns the next element on the stack. 2380 * 2381 * @return Returns the actual QName object. Callee should 2382 * use this object to store the details of next element encountered. 2383 */ 2384 public QName nextElement() { 2385 if(fSkip){ 2386 fDepth++; 2387 //boundary checks are done in matchElement() 2388 return fElements[fCount++]; 2389 } else if (fDepth == fElements.length) { 2390 QName[] array = new QName[fElements.length * 2]; 2391 System.arraycopy(fElements, 0, array, 0, fDepth); 2392 fElements = array; 2393 for (int i = fDepth; i < fElements.length; i++) { 2394 fElements[i] = new QName(); 2395 } 2396 } 2397 2398 return fElements[fDepth++]; 2399 2400 } // pushElement(QName):QName 2401 2402 2403 /** 2404 * Pops an element off of the stack by setting the values of 2405 * the specified QName. 2406 * <p> 2407 * <strong>Note:</strong> The object returned is <em>not</em> 2408 * orphaned to the caller. Therefore, the caller should consider 2409 * the object to be read-only. 2410 */ 2411 public QName popElement() { 2412 //return the same object that was pushed -- this would avoid 2413 //setting the values for every end element. 2414 //STRONG: this object is read only -- this object reference shouldn't be stored. 2415 if(fSkip || fAdd ){ 2416 if(DEBUG_SKIP_ALGORITHM){ 2417 System.out.println("POPPING Element, at position " + fInt[fDepth] + " element at that count is = " + fElements[fInt[fDepth]].rawname); 2418 System.out.println(""); 2419 } 2420 return fElements[fInt[fDepth--]]; 2421 } else{ 2422 if(DEBUG_SKIP_ALGORITHM){ 2423 System.out.println("Retrieveing element at depth = " + fDepth + " is " + fElements[fDepth].rawname ); 2424 } 2425 return fElements[--fDepth] ; 2426 } 2427 //element.setValues(fElements[--fDepth]); 2428 } // popElement(QName) 2429 2430 /** Reposition the stack. fInt [] contains all the opened tags at particular depth. 2431 * Transfer all the opened tags starting from depth '2' to the current depth and reposition them 2432 *as per the depth. 2433 */ 2434 public void reposition(){ 2435 for( int i = 2 ; i <= fDepth ; i++){ 2436 fElements[i-1] = fElements[fInt[i]]; 2437 } 2438 if(DEBUG_SKIP_ALGORITHM){ 2439 for( int i = 0 ; i < fDepth ; i++){ 2440 System.out.println("fElements[" + i + "]" + " = " + fElements[i].rawname); 2441 } 2442 } 2443 } 2444 2445 /** Clears the stack without throwing away existing QName objects. */ 2446 public void clear() { 2447 fDepth = 0; 2448 fLastDepth = 0; 2449 fCount = 0 ; 2450 fPosition = fMark = 1; 2451 2452 } // clear() 2453 2454 /** 2455 * This function is as a result of optimization done for endElement -- 2456 * we dont need to set the value for every end element encouterd. 2457 * For Well formedness checks we can have the same QName object that was pushed. 2458 * the values will be set only if application need to know about the endElement 2459 * -- neeraj.bajaj@sun.com 2460 */ 2461 2462 public QName getLastPoppedElement(){ 2463 return fElements[fDepth]; 2464 } 2465 } // class ElementStack 2466 2467 /** 2468 * Drives the parser to the next state/event on the input. Parser is guaranteed 2469 * to stop at the next state/event. 2470 * 2471 * Internally XML document is divided into several states. Each state represents 2472 * a sections of XML document. When this functions returns normally, it has read 2473 * the section of XML document and returns the state corresponding to section of 2474 * document which has been read. For optimizations, a particular driver 2475 * can read ahead of the section of document (state returned) just read and 2476 * can maintain a different internal state. 2477 * 2478 * 2479 * @author Neeraj Bajaj, Sun Microsystems 2480 */ 2481 protected interface Driver { 2482 2483 2484 /** 2485 * Drives the parser to the next state/event on the input. Parser is guaranteed 2486 * to stop at the next state/event. 2487 * 2488 * Internally XML document is divided into several states. Each state represents 2489 * a sections of XML document. When this functions returns normally, it has read 2490 * the section of XML document and returns the state corresponding to section of 2491 * document which has been read. For optimizations, a particular driver 2492 * can read ahead of the section of document (state returned) just read and 2493 * can maintain a different internal state. 2494 * 2495 * @return state representing the section of document just read. 2496 * 2497 * @throws IOException Thrown on i/o error. 2498 * @throws XNIException Thrown on parse error. 2499 */ 2500 2501 public int next() throws IOException, XNIException; 2502 2503 } // interface Driver 2504 2505 /** 2506 * Driver to handle content scanning. This driver is capable of reading 2507 * the fragment of XML document. When it has finished reading fragment 2508 * of XML documents, it can pass the job of reading to another driver. 2509 * 2510 * This class has been modified as per the new design which is more suited to 2511 * efficiently build pull parser. Lot of performance improvements have been done and 2512 * the code has been added to support stax functionality/features. 2513 * 2514 * @author Neeraj Bajaj, Sun Microsystems 2515 * 2516 * 2517 * @author Andy Clark, IBM 2518 * @author Eric Ye, IBM 2519 */ 2520 protected class FragmentContentDriver 2521 implements Driver { 2522 2523 // 2524 // Driver methods 2525 // 2526 private boolean fContinueDispatching = true; 2527 private boolean fScanningForMarkup = true; 2528 2529 /** 2530 * decides the appropriate state of the parser 2531 */ 2532 private void startOfMarkup() throws IOException { 2533 fMarkupDepth++; 2534 final int ch = fEntityScanner.peekChar(); 2535 2536 switch(ch){ 2537 case '?' :{ 2538 setScannerState(SCANNER_STATE_PI); 2539 fEntityScanner.skipChar(ch); 2540 break; 2541 } 2542 case '!' :{ 2543 fEntityScanner.skipChar(ch); 2544 if (fEntityScanner.skipChar('-')) { 2545 if (!fEntityScanner.skipChar('-')) { 2546 reportFatalError("InvalidCommentStart", 2547 null); 2548 } 2549 setScannerState(SCANNER_STATE_COMMENT); 2550 } else if (fEntityScanner.skipString(cdata)) { 2551 setScannerState(SCANNER_STATE_CDATA ); 2552 } else if (!scanForDoctypeHook()) { 2553 reportFatalError("MarkupNotRecognizedInContent", 2554 null); 2555 } 2556 break; 2557 } 2558 case '/' :{ 2559 setScannerState(SCANNER_STATE_END_ELEMENT_TAG); 2560 fEntityScanner.skipChar(ch); 2561 break; 2562 } 2563 default :{ 2564 if (isValidNameStartChar(ch)) { 2565 setScannerState(SCANNER_STATE_START_ELEMENT_TAG); 2566 } else { 2567 reportFatalError("MarkupNotRecognizedInContent", 2568 null); 2569 } 2570 } 2571 } 2572 2573 }//startOfMarkup 2574 2575 private void startOfContent() throws IOException { 2576 if (fEntityScanner.skipChar('<')) { 2577 setScannerState(SCANNER_STATE_START_OF_MARKUP); 2578 } else if (fEntityScanner.skipChar('&')) { 2579 setScannerState(SCANNER_STATE_REFERENCE) ; //XMLEvent.ENTITY_REFERENCE ); //SCANNER_STATE_REFERENCE 2580 } else { 2581 //element content is there.. 2582 setScannerState(SCANNER_STATE_CHARACTER_DATA); 2583 } 2584 }//startOfContent 2585 2586 2587 /** 2588 * 2589 * SCANNER_STATE_CONTENT and SCANNER_STATE_START_OF_MARKUP are two super states of the parser. 2590 * At any point of time when in doubt over the current state of the parser, the state should be 2591 * set to SCANNER_STATE_CONTENT. Parser will automatically revive itself and will set state of 2592 * the parser to one of its sub state. 2593 * sub states are defined in the parser on the basis of different XML component like 2594 * SCANNER_STATE_ENTITY_REFERENCE , SCANNER_STATE_START_ELEMENT, SCANNER_STATE_CDATA etc.. 2595 * These sub states help the parser to have fine control over the parsing. These are the 2596 * different milepost, parser stops at each sub state (milepost). Based on this state it is 2597 * decided if paresr needs to stop at next milepost ?? 2598 * 2599 */ 2600 public void decideSubState() throws IOException { 2601 while( fScannerState == SCANNER_STATE_CONTENT || fScannerState == SCANNER_STATE_START_OF_MARKUP){ 2602 2603 switch (fScannerState) { 2604 2605 case SCANNER_STATE_CONTENT: { 2606 startOfContent() ; 2607 break; 2608 } 2609 2610 case SCANNER_STATE_START_OF_MARKUP: { 2611 startOfMarkup() ; 2612 break; 2613 } 2614 } 2615 } 2616 }//decideSubState 2617 2618 /** 2619 * Drives the parser to the next state/event on the input. Parser is guaranteed 2620 * to stop at the next state/event. Internally XML document 2621 * is divided into several states. Each state represents a sections of XML 2622 * document. When this functions returns normally, it has read the section 2623 * of XML document and returns the state corresponding to section of 2624 * document which has been read. For optimizations, a particular driver 2625 * can read ahead of the section of document (state returned) just read and 2626 * can maintain a different internal state. 2627 * 2628 * State returned corresponds to Stax states. 2629 * 2630 * @return state representing the section of document just read. 2631 * 2632 * @throws IOException Thrown on i/o error. 2633 * @throws XNIException Thrown on parse error. 2634 */ 2635 2636 public int next() throws IOException, XNIException { 2637 while (true) { 2638 try { 2639 if(DEBUG_NEXT){ 2640 System.out.println("NOW IN FragmentContentDriver"); 2641 System.out.println("Entering the FragmentContentDriver with = " + getScannerStateName(fScannerState)); 2642 } 2643 2644 //decide the actual sub state of the scanner.For more information refer to the javadoc of 2645 //decideSubState. 2646 2647 switch (fScannerState) { 2648 case SCANNER_STATE_CONTENT: { 2649 final int ch = fEntityScanner.peekChar(); 2650 if (ch == '<') { 2651 fEntityScanner.scanChar(); 2652 setScannerState(SCANNER_STATE_START_OF_MARKUP); 2653 } else if (ch == '&') { 2654 fEntityScanner.scanChar(); 2655 setScannerState(SCANNER_STATE_REFERENCE) ; //XMLEvent.ENTITY_REFERENCE ); //SCANNER_STATE_REFERENCE 2656 break; 2657 } else { 2658 //element content is there.. 2659 setScannerState(SCANNER_STATE_CHARACTER_DATA); 2660 break; 2661 } 2662 } 2663 2664 case SCANNER_STATE_START_OF_MARKUP: { 2665 startOfMarkup(); 2666 break; 2667 }//case: SCANNER_STATE_START_OF_MARKUP 2668 2669 }//end of switch 2670 //decideSubState() ; 2671 2672 //do some special handling if isCoalesce is set to true. 2673 if(fIsCoalesce){ 2674 fUsebuffer = true ; 2675 //if the last section was character data 2676 if(fLastSectionWasCharacterData){ 2677 2678 //if we dont encounter any CDATA or ENITY REFERENCE and current state is also not SCANNER_STATE_CHARACTER_DATA 2679 //return the last scanned charactrer data. 2680 if((fScannerState != SCANNER_STATE_CDATA) && (fScannerState != SCANNER_STATE_REFERENCE) 2681 && (fScannerState != SCANNER_STATE_CHARACTER_DATA)){ 2682 fLastSectionWasCharacterData = false; 2683 return XMLEvent.CHARACTERS; 2684 } 2685 }//if last section was CDATA or ENTITY REFERENCE 2686 //xxx: there might be another entity reference or CDATA after this 2687 //<foo>blah blah &<<![CDATA[[aa]]>blah blah</foo> 2688 else if((fLastSectionWasCData || fLastSectionWasEntityReference)){ 2689 //and current state is not SCANNER_STATE_CHARACTER_DATA 2690 //or SCANNER_STATE_CDATA or SCANNER_STATE_REFERENCE 2691 //this means there is nothing more to be coalesced. 2692 //return the CHARACTERS event. 2693 if((fScannerState != SCANNER_STATE_CDATA) && (fScannerState != SCANNER_STATE_REFERENCE) 2694 && (fScannerState != SCANNER_STATE_CHARACTER_DATA)){ 2695 2696 fLastSectionWasCData = false; 2697 fLastSectionWasEntityReference = false; 2698 return XMLEvent.CHARACTERS; 2699 } 2700 } 2701 } 2702 2703 2704 if(DEBUG_NEXT){ 2705 System.out.println("Actual scanner state set by decideSubState is = " + getScannerStateName(fScannerState)); 2706 } 2707 2708 switch(fScannerState){ 2709 2710 case XMLEvent.START_DOCUMENT : 2711 return XMLEvent.START_DOCUMENT; 2712 2713 case SCANNER_STATE_START_ELEMENT_TAG :{ 2714 2715 //xxx this function returns true when element is empty.. can be linked to end element event. 2716 //returns true if the element is empty 2717 fEmptyElement = scanStartElement() ; 2718 //if the element is empty the next event is "end element" 2719 if(fEmptyElement){ 2720 setScannerState(SCANNER_STATE_END_ELEMENT_TAG); 2721 }else{ 2722 //set the next possible state 2723 setScannerState(SCANNER_STATE_CONTENT); 2724 } 2725 return XMLEvent.START_ELEMENT ; 2726 } 2727 2728 case SCANNER_STATE_CHARACTER_DATA: { 2729 if(DEBUG_COALESCE){ 2730 System.out.println("fLastSectionWasCData = " + fLastSectionWasCData); 2731 System.out.println("fIsCoalesce = " + fIsCoalesce); 2732 } 2733 //if last section was either entity reference or cdata or character data we should be using buffer 2734 fUsebuffer = fLastSectionWasEntityReference || fLastSectionWasCData || fLastSectionWasCharacterData ; 2735 2736 //When coalesce is set to true and last state was REFERENCE or CDATA or CHARACTER_DATA, buffer should not be cleared. 2737 if( fIsCoalesce && (fLastSectionWasEntityReference || fLastSectionWasCData || fLastSectionWasCharacterData) ){ 2738 fLastSectionWasEntityReference = false; 2739 fLastSectionWasCData = false; 2740 fLastSectionWasCharacterData = true ; 2741 fUsebuffer = true; 2742 }else{ 2743 //clear the buffer 2744 fContentBuffer.clear(); 2745 } 2746 2747 //set the fTempString length to 0 before passing it on to scanContent 2748 //scanContent sets the correct co-ordinates as per the content read 2749 fTempString.length = 0; 2750 int c = fEntityScanner.scanContent(fTempString); 2751 if(DEBUG){ 2752 System.out.println("fTempString = " + fTempString); 2753 } 2754 if(fEntityScanner.skipChar('<')){ 2755 //check if we have reached end of element 2756 if(fEntityScanner.skipChar('/')){ 2757 //increase the mark up depth 2758 fMarkupDepth++; 2759 fLastSectionWasCharacterData = false; 2760 setScannerState(SCANNER_STATE_END_ELEMENT_TAG); 2761 //check if its start of new element 2762 }else if(XMLChar.isNameStart(fEntityScanner.peekChar())){ 2763 fMarkupDepth++; 2764 fLastSectionWasCharacterData = false; 2765 setScannerState(SCANNER_STATE_START_ELEMENT_TAG); 2766 }else{ 2767 setScannerState(SCANNER_STATE_START_OF_MARKUP); 2768 //there can be cdata ahead if coalesce is true we should call again 2769 if(fIsCoalesce){ 2770 fUsebuffer = true; 2771 fLastSectionWasCharacterData = true; 2772 fContentBuffer.append(fTempString); 2773 fTempString.length = 0; 2774 continue; 2775 } 2776 } 2777 //in case last section was either entity reference or cdata or character data -- we should be using buffer 2778 if(fUsebuffer){ 2779 fContentBuffer.append(fTempString); 2780 fTempString.length = 0; 2781 } 2782 if(DEBUG){ 2783 System.out.println("NOT USING THE BUFFER, STRING = " + fTempString.toString()); 2784 } 2785 if(dtdGrammarUtil!= null && dtdGrammarUtil.isIgnorableWhiteSpace(fContentBuffer)){ 2786 if(DEBUG)System.out.println("Return SPACE EVENT"); 2787 return XMLEvent.SPACE; 2788 }else 2789 return XMLEvent.CHARACTERS; 2790 2791 } else{ 2792 fUsebuffer = true ; 2793 if(DEBUG){ 2794 System.out.println("fContentBuffer = " + fContentBuffer); 2795 System.out.println("fTempString = " + fTempString); 2796 } 2797 fContentBuffer.append(fTempString); 2798 fTempString.length = 0; 2799 } 2800 if (c == '\r') { 2801 if(DEBUG){ 2802 System.out.println("'\r' character found"); 2803 } 2804 // happens when there is the character reference 2805 //xxx: We know the next chracter.. we should just skip it and add ']' directlry 2806 fEntityScanner.scanChar(); 2807 fUsebuffer = true; 2808 fContentBuffer.append((char)c); 2809 c = -1 ; 2810 } else if (c == ']') { 2811 //fStringBuffer.clear(); 2812 //xxx: We know the next chracter.. we should just skip it and add ']' directlry 2813 fUsebuffer = true; 2814 fContentBuffer.append((char)fEntityScanner.scanChar()); 2815 // remember where we are in case we get an endEntity before we 2816 // could flush the buffer out - this happens when we're parsing an 2817 // entity which ends with a ] 2818 fInScanContent = true; 2819 2820 // We work on a single character basis to handle cases such as: 2821 // ']]]>' which we might otherwise miss. 2822 // 2823 if (fEntityScanner.skipChar(']')) { 2824 fContentBuffer.append(']'); 2825 while (fEntityScanner.skipChar(']')) { 2826 fContentBuffer.append(']'); 2827 } 2828 if (fEntityScanner.skipChar('>')) { 2829 reportFatalError("CDEndInContent", null); 2830 } 2831 } 2832 c = -1 ; 2833 fInScanContent = false; 2834 } 2835 2836 do{ 2837 //xxx: we should be using only one buffer.. 2838 // we need not to grow the buffer only when isCoalesce() is not true; 2839 2840 if (c == '<') { 2841 fEntityScanner.scanChar(); 2842 setScannerState(SCANNER_STATE_START_OF_MARKUP); 2843 break; 2844 }//xxx what should be the behavior if entity reference is present in the content ? 2845 else if (c == '&') { 2846 fEntityScanner.scanChar(); 2847 setScannerState(SCANNER_STATE_REFERENCE); 2848 break; 2849 }///xxx since this part is also characters, it should be merged... 2850 else if (c != -1 && isInvalidLiteral(c)) { 2851 if (XMLChar.isHighSurrogate(c)) { 2852 // special case: surrogates 2853 scanSurrogates(fContentBuffer) ; 2854 setScannerState(SCANNER_STATE_CONTENT); 2855 } else { 2856 reportFatalError("InvalidCharInContent", 2857 new Object[] { 2858 Integer.toString(c, 16)}); 2859 fEntityScanner.scanChar(); 2860 } 2861 break; 2862 } 2863 //xxx: scanContent also gives character callback. 2864 c = scanContent(fContentBuffer) ; 2865 //we should not be iterating again if fIsCoalesce is not set to true 2866 2867 if(!fIsCoalesce){ 2868 setScannerState(SCANNER_STATE_CONTENT); 2869 break; 2870 } 2871 2872 }while(true); 2873 2874 //if (fDocumentHandler != null) { 2875 // fDocumentHandler.characters(fContentBuffer, null); 2876 //} 2877 if(DEBUG)System.out.println("USING THE BUFFER, STRING START=" + fContentBuffer.toString() +"=END"); 2878 //if fIsCoalesce is true there might be more data so call fDriver.next() 2879 if(fIsCoalesce){ 2880 fLastSectionWasCharacterData = true ; 2881 continue; 2882 }else{ 2883 if(dtdGrammarUtil!= null && dtdGrammarUtil.isIgnorableWhiteSpace(fContentBuffer)){ 2884 if(DEBUG)System.out.println("Return SPACE EVENT"); 2885 return XMLEvent.SPACE; 2886 } else 2887 return XMLEvent.CHARACTERS ; 2888 } 2889 } 2890 2891 case SCANNER_STATE_END_ELEMENT_TAG :{ 2892 if(fEmptyElement){ 2893 //set it back to false. 2894 fEmptyElement = false; 2895 setScannerState(SCANNER_STATE_CONTENT); 2896 //check the case when there is comment after single element document 2897 //<foo/> and some comment after this 2898 return (fMarkupDepth == 0 && elementDepthIsZeroHook() ) ? XMLEvent.END_ELEMENT : XMLEvent.END_ELEMENT ; 2899 2900 } else if(scanEndElement() == 0) { 2901 //It is last element of the document 2902 if (elementDepthIsZeroHook()) { 2903 //if element depth is zero , it indicates the end of the document 2904 //the state shouldn't be set, because it is set by elementDepthIsZeroHook() function 2905 //xxx understand this point once again.. 2906 return XMLEvent.END_ELEMENT ; 2907 } 2908 2909 } 2910 setScannerState(SCANNER_STATE_CONTENT); 2911 return XMLEvent.END_ELEMENT ; 2912 } 2913 2914 case SCANNER_STATE_COMMENT: { //SCANNER_STATE_COMMENT: 2915 scanComment(); 2916 setScannerState(SCANNER_STATE_CONTENT); 2917 return XMLEvent.COMMENT; 2918 //break; 2919 } 2920 case SCANNER_STATE_PI:{ //SCANNER_STATE_PI: { 2921 //clear the buffer first 2922 fContentBuffer.clear() ; 2923 //xxx: which buffer should be passed. Ideally we shouldn't have 2924 //more than two buffers -- 2925 //xxx: where should we add the switch for buffering. 2926 scanPI(fContentBuffer); 2927 setScannerState(SCANNER_STATE_CONTENT); 2928 return XMLEvent.PROCESSING_INSTRUCTION; 2929 //break; 2930 } 2931 case SCANNER_STATE_CDATA :{ //SCANNER_STATE_CDATA: { 2932 //xxx: What if CDATA is the first event 2933 //<foo><![CDATA[hello<><>]]>append</foo> 2934 2935 //we should not clear the buffer only when the last state was either SCANNER_STATE_REFERENCE or 2936 //SCANNER_STATE_CHARACTER_DATA or SCANNER_STATE_REFERENCE 2937 if(fIsCoalesce && ( fLastSectionWasEntityReference || fLastSectionWasCData || fLastSectionWasCharacterData)){ 2938 fLastSectionWasCData = true ; 2939 fLastSectionWasEntityReference = false; 2940 fLastSectionWasCharacterData = false; 2941 }//if we dont need to coalesce clear the buffer 2942 else{ 2943 fContentBuffer.clear(); 2944 } 2945 fUsebuffer = true; 2946 //CDATA section is completely read in all the case. 2947 scanCDATASection(fContentBuffer , true); 2948 setScannerState(SCANNER_STATE_CONTENT); 2949 //1. if fIsCoalesce is set to true we set the variable fLastSectionWasCData to true 2950 //and just call fDispatche.next(). Since we have set the scanner state to 2951 //SCANNER_STATE_CONTENT (super state) parser will automatically recover and 2952 //behave appropriately. When isCoalesce is set to true we dont need to reportCDATA event 2953 //2. Check if application has set for reporting CDATA event 2954 //3. if the application has neither set the fIsCoalesce to true nor fReportCdataEvent 2955 //return the cdata event as characters. 2956 if(fIsCoalesce){ 2957 fLastSectionWasCData = true ; 2958 //there might be more data to coalesce. 2959 continue; 2960 }else if(fReportCdataEvent){ 2961 return XMLEvent.CDATA; 2962 } else{ 2963 return XMLEvent.CHARACTERS; 2964 } 2965 } 2966 2967 case SCANNER_STATE_REFERENCE :{ 2968 fMarkupDepth++; 2969 foundBuiltInRefs = false; 2970 2971 //we should not clear the buffer only when the last state was either CDATA or 2972 //SCANNER_STATE_CHARACTER_DATA or SCANNER_STATE_REFERENCE 2973 if(fIsCoalesce && ( fLastSectionWasEntityReference || fLastSectionWasCData || fLastSectionWasCharacterData)){ 2974 //fLastSectionWasEntityReference or fLastSectionWasCData are only 2975 //used when fIsCoalesce is set to true. 2976 fLastSectionWasEntityReference = true ; 2977 fLastSectionWasCData = false; 2978 fLastSectionWasCharacterData = false; 2979 }//if we dont need to coalesce clear the buffer 2980 else{ 2981 fContentBuffer.clear(); 2982 } 2983 fUsebuffer = true ; 2984 //take care of character reference 2985 if (fEntityScanner.skipChar('#')) { 2986 scanCharReferenceValue(fContentBuffer, null); 2987 fMarkupDepth--; 2988 if(!fIsCoalesce){ 2989 setScannerState(SCANNER_STATE_CONTENT); 2990 return XMLEvent.CHARACTERS; 2991 } 2992 } else { 2993 // this function also starts new entity 2994 scanEntityReference(fContentBuffer); 2995 //if there was built-in entity reference & coalesce is not true 2996 //return CHARACTERS 2997 if(fScannerState == SCANNER_STATE_BUILT_IN_REFS && !fIsCoalesce){ 2998 setScannerState(SCANNER_STATE_CONTENT); 2999 return XMLEvent.CHARACTERS; 3000 } 3001 3002 //if there was a text declaration, call next() it will be taken care. 3003 if(fScannerState == SCANNER_STATE_TEXT_DECL){ 3004 fLastSectionWasEntityReference = true ; 3005 continue; 3006 } 3007 3008 if(fScannerState == SCANNER_STATE_REFERENCE){ 3009 setScannerState(SCANNER_STATE_CONTENT); 3010 if (fReplaceEntityReferences && fEntityStore.isDeclaredEntity(fCurrentEntityName)) { 3011 // Skip the entity reference, we don't care 3012 continue; 3013 } 3014 return XMLEvent.ENTITY_REFERENCE; 3015 } 3016 } 3017 //Wether it was character reference, entity reference or built-in entity 3018 //set the next possible state to SCANNER_STATE_CONTENT 3019 setScannerState(SCANNER_STATE_CONTENT); 3020 fLastSectionWasEntityReference = true ; 3021 continue; 3022 } 3023 3024 case SCANNER_STATE_TEXT_DECL: { 3025 // scan text decl 3026 if (fEntityScanner.skipString("<?xml")) { 3027 fMarkupDepth++; 3028 // NOTE: special case where entity starts with a PI 3029 // whose name starts with "xml" (e.g. "xmlfoo") 3030 if (isValidNameChar(fEntityScanner.peekChar())) { 3031 fStringBuffer.clear(); 3032 fStringBuffer.append("xml"); 3033 3034 if (fNamespaces) { 3035 while (isValidNCName(fEntityScanner.peekChar())) { 3036 fStringBuffer.append((char)fEntityScanner.scanChar()); 3037 } 3038 } else { 3039 while (isValidNameChar(fEntityScanner.peekChar())) { 3040 fStringBuffer.append((char)fEntityScanner.scanChar()); 3041 } 3042 } 3043 String target = fSymbolTable.addSymbol(fStringBuffer.ch, fStringBuffer.offset, fStringBuffer.length); 3044 fContentBuffer.clear(); 3045 scanPIData(target, fContentBuffer); 3046 } 3047 3048 // standard text declaration 3049 else { 3050 //xxx: this function gives callback 3051 scanXMLDeclOrTextDecl(true); 3052 } 3053 } 3054 // now that we've straightened out the readers, we can read in chunks: 3055 fEntityManager.fCurrentEntity.mayReadChunks = true; 3056 setScannerState(SCANNER_STATE_CONTENT); 3057 //xxx: we don't return any state, so how do we get to know about TEXT declarations. 3058 //it seems we have to careful when to allow function issue a callback 3059 //and when to allow adapter issue a callback. 3060 continue; 3061 } 3062 3063 3064 case SCANNER_STATE_ROOT_ELEMENT: { 3065 if (scanRootElementHook()) { 3066 fEmptyElement = true; 3067 //rest would be taken care by fTrailingMiscDriver set by scanRootElementHook 3068 return XMLEvent.START_ELEMENT; 3069 } 3070 setScannerState(SCANNER_STATE_CONTENT); 3071 return XMLEvent.START_ELEMENT ; 3072 } 3073 case SCANNER_STATE_CHAR_REFERENCE : { 3074 fContentBuffer.clear(); 3075 scanCharReferenceValue(fContentBuffer, null); 3076 fMarkupDepth--; 3077 setScannerState(SCANNER_STATE_CONTENT); 3078 return XMLEvent.CHARACTERS; 3079 } 3080 default: 3081 throw new XNIException("Scanner State " + fScannerState + " not Recognized "); 3082 3083 }//switch 3084 } 3085 // premature end of file 3086 catch (EOFException e) { 3087 endOfFileHook(e); 3088 return -1; 3089 } 3090 } //while loop 3091 }//next 3092 3093 3094 // 3095 // Protected methods 3096 // 3097 3098 // hooks 3099 3100 // NOTE: These hook methods are added so that the full document 3101 // scanner can share the majority of code with this class. 3102 3103 /** 3104 * Scan for DOCTYPE hook. This method is a hook for subclasses 3105 * to add code to handle scanning for a the "DOCTYPE" string 3106 * after the string "<!" has been scanned. 3107 * 3108 * @return True if the "DOCTYPE" was scanned; false if "DOCTYPE" 3109 * was not scanned. 3110 */ 3111 protected boolean scanForDoctypeHook() 3112 throws IOException, XNIException { 3113 return false; 3114 } // scanForDoctypeHook():boolean 3115 3116 /** 3117 * Element depth iz zero. This methos is a hook for subclasses 3118 * to add code to handle when the element depth hits zero. When 3119 * scanning a document fragment, an element depth of zero is 3120 * normal. However, when scanning a full XML document, the 3121 * scanner must handle the trailing miscellanous section of 3122 * the document after the end of the document's root element. 3123 * 3124 * @return True if the caller should stop and return true which 3125 * allows the scanner to switch to a new scanning 3126 * driver. A return value of false indicates that 3127 * the content driver should continue as normal. 3128 */ 3129 protected boolean elementDepthIsZeroHook() 3130 throws IOException, XNIException { 3131 return false; 3132 } // elementDepthIsZeroHook():boolean 3133 3134 /** 3135 * Scan for root element hook. This method is a hook for 3136 * subclasses to add code that handles scanning for the root 3137 * element. When scanning a document fragment, there is no 3138 * "root" element. However, when scanning a full XML document, 3139 * the scanner must handle the root element specially. 3140 * 3141 * @return True if the caller should stop and return true which 3142 * allows the scanner to switch to a new scanning 3143 * driver. A return value of false indicates that 3144 * the content driver should continue as normal. 3145 */ 3146 protected boolean scanRootElementHook() 3147 throws IOException, XNIException { 3148 return false; 3149 } // scanRootElementHook():boolean 3150 3151 /** 3152 * End of file hook. This method is a hook for subclasses to 3153 * add code that handles the end of file. The end of file in 3154 * a document fragment is OK if the markup depth is zero. 3155 * However, when scanning a full XML document, an end of file 3156 * is always premature. 3157 */ 3158 protected void endOfFileHook(EOFException e) 3159 throws IOException, XNIException { 3160 3161 // NOTE: An end of file is only only an error if we were 3162 // in the middle of scanning some markup. -Ac 3163 if (fMarkupDepth != 0) { 3164 reportFatalError("PrematureEOF", null); 3165 } 3166 3167 } // endOfFileHook() 3168 3169 } // class FragmentContentDriver 3170 3171 static void pr(String str) { 3172 System.out.println(str) ; 3173 } 3174 3175 protected boolean fUsebuffer ; 3176 3177 /** this function gets an XMLString (which is used to store the attribute value) from the special pool 3178 * maintained for attributes. 3179 * fAttributeCacheUsedCount tracks the number of attributes that has been consumed from the pool. 3180 * if all the attributes has been consumed, it adds a new XMLString inthe pool and returns the same 3181 * XMLString. 3182 * 3183 * @return XMLString XMLString used to store an attribute value. 3184 */ 3185 3186 protected XMLString getString(){ 3187 if(fAttributeCacheUsedCount < initialCacheCount || fAttributeCacheUsedCount < attributeValueCache.size()){ 3188 return (XMLString)attributeValueCache.get(fAttributeCacheUsedCount++); 3189 } else{ 3190 XMLString str = new XMLString(); 3191 fAttributeCacheUsedCount++; 3192 attributeValueCache.add(str); 3193 return str; 3194 } 3195 } 3196 3197 /** 3198 * Implements XMLBufferListener interface. 3199 */ 3200 3201 public void refresh(){ 3202 refresh(0); 3203 } 3204 3205 /** 3206 * receives callbacks from {@link XMLEntityReader } when buffer 3207 * is being changed. 3208 * @param refreshPosition 3209 */ 3210 public void refresh(int refreshPosition){ 3211 //If you are reading attributes and you got a callback 3212 //cache available attributes. 3213 if(fReadingAttributes){ 3214 fAttributes.refresh(); 3215 } 3216 if(fScannerState == SCANNER_STATE_CHARACTER_DATA){ 3217 //since fTempString directly matches to the underlying main buffer 3218 //store the data into buffer 3219 fContentBuffer.append(fTempString); 3220 //clear the XMLString so that data can't be added again. 3221 fTempString.length = 0; 3222 fUsebuffer = true; 3223 } 3224 } 3225 3226 } // class XMLDocumentFragmentScannerImpl