1 /* 2 * Copyright (c) 2006, 2018, Oracle and/or its affiliates. All rights reserved. 3 */ 4 /* 5 * Licensed to the Apache Software Foundation (ASF) under one or more 6 * contributor license agreements. See the NOTICE file distributed with 7 * this work for additional information regarding copyright ownership. 8 * The ASF licenses this file to You under the Apache License, Version 2.0 9 * (the "License"); you may not use this file except in compliance with 10 * the License. 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 package com.sun.org.apache.xml.internal.serializer; 22 23 import com.sun.org.apache.xml.internal.serializer.utils.MsgKey; 24 import com.sun.org.apache.xml.internal.serializer.utils.Utils; 25 import com.sun.org.apache.xml.internal.serializer.utils.WrappedRuntimeException; 26 import java.io.IOException; 27 import java.io.OutputStream; 28 import java.io.OutputStreamWriter; 29 import java.io.UnsupportedEncodingException; 30 import java.io.Writer; 31 import java.util.ArrayList; 32 import java.util.Arrays; 33 import java.util.EmptyStackException; 34 import java.util.Enumeration; 35 import java.util.Iterator; 36 import java.util.List; 37 import java.util.Properties; 38 import java.util.Set; 39 import java.util.StringTokenizer; 40 import javax.xml.transform.ErrorListener; 41 import javax.xml.transform.OutputKeys; 42 import javax.xml.transform.Transformer; 43 import javax.xml.transform.TransformerException; 44 import org.w3c.dom.Node; 45 import org.xml.sax.Attributes; 46 import org.xml.sax.ContentHandler; 47 import org.xml.sax.SAXException; 48 49 /** 50 * This abstract class is a base class for other stream 51 * serializers (xml, html, text ...) that write output to a stream. 52 * 53 * @xsl.usage internal 54 * @LastModified: Sept 2018 55 */ 56 abstract public class ToStream extends SerializerBase { 57 58 private static final String COMMENT_BEGIN = "<!--"; 59 private static final String COMMENT_END = "-->"; 60 61 /** Stack to keep track of disabling output escaping. */ 62 protected BoolStack m_disableOutputEscapingStates = new BoolStack(); 63 64 /** 65 * The encoding information associated with this serializer. 66 * Although initially there is no encoding, 67 * there is a dummy EncodingInfo object that will say 68 * that every character is in the encoding. This is useful 69 * for a serializer that is in temporary output state and has 70 * no associated encoding. A serializer in final output state 71 * will have an encoding, and will worry about whether 72 * single chars or surrogate pairs of high/low chars form 73 * characters in the output encoding. 74 */ 75 EncodingInfo m_encodingInfo = new EncodingInfo(null,null); 76 77 /** 78 * Method reference to the sun.io.CharToByteConverter#canConvert method 79 * for this encoding. Invalid if m_charToByteConverter is null. 80 */ 81 java.lang.reflect.Method m_canConvertMeth; 82 83 /** 84 * Boolean that tells if we already tried to get the converter. 85 */ 86 boolean m_triedToGetConverter = false; 87 88 /** 89 * Opaque reference to the sun.io.CharToByteConverter for this 90 * encoding. 91 */ 92 Object m_charToByteConverter = null; 93 94 /** 95 * Used to buffer the text nodes and the entity reference nodes if 96 * indentation is on. 97 */ 98 protected CharacterBuffer m_charactersBuffer = new CharacterBuffer(); 99 100 /** 101 * Used to decide if a text node is pretty-printed with indentation. 102 * If m_childNodeNum > 1, the text node will be indented. 103 * 104 */ 105 protected List<Integer> m_childNodeNumStack = new ArrayList<>(); 106 107 protected int m_childNodeNum = 0; 108 109 /** 110 * Used to handle xml:space attribute 111 * 112 */ 113 protected BoolStack m_preserveSpaces = new BoolStack(); 114 115 protected boolean m_ispreserveSpace = false; 116 117 118 /** 119 * State flag that tells if the previous node processed 120 * was text, so we can tell if we should preserve whitespace. 121 * 122 * Used in endDocument() and shouldIndent() but 123 * only if m_doIndent is true. 124 * If m_doIndent is false this flag has no impact. 125 */ 126 protected boolean m_isprevtext = false; 127 128 /** 129 * The maximum character size before we have to resort 130 * to escaping. 131 */ 132 protected int m_maxCharacter = Encodings.getLastPrintable(); 133 134 /** 135 * The system line separator for writing out line breaks. 136 * The default value is from the system property, 137 * but this value can be set through the xsl:output 138 * extension attribute xalan:line-separator. 139 */ 140 protected char[] m_lineSep = System.lineSeparator().toCharArray(); 141 142 /** 143 * True if the the system line separator is to be used. 144 */ 145 protected boolean m_lineSepUse = true; 146 147 /** 148 * The length of the line seperator, since the write is done 149 * one character at a time. 150 */ 151 protected int m_lineSepLen = m_lineSep.length; 152 153 /** 154 * Map that tells which characters should have special treatment, and it 155 * provides character to entity name lookup. 156 */ 157 protected CharInfo m_charInfo; 158 159 /** True if we control the buffer, and we should flush the output on endDocument. */ 160 boolean m_shouldFlush = true; 161 162 /** 163 * Add space before '/>' for XHTML. 164 */ 165 protected boolean m_spaceBeforeClose = false; 166 167 /** 168 * Flag to signal that a newline should be added. 169 * 170 * Used only in indent() which is called only if m_doIndent is true. 171 * If m_doIndent is false this flag has no impact. 172 */ 173 boolean m_startNewLine; 174 175 /** 176 * Tells if we're in an internal document type subset. 177 */ 178 protected boolean m_inDoctype = false; 179 180 /** 181 * Flag to quickly tell if the encoding is UTF8. 182 */ 183 boolean m_isUTF8 = false; 184 185 /** 186 * remembers if we are in between the startCDATA() and endCDATA() callbacks 187 */ 188 protected boolean m_cdataStartCalled = false; 189 190 /** 191 * If this flag is true DTD entity references are not left as-is, 192 * which is exiting older behavior. 193 */ 194 private boolean m_expandDTDEntities = true; 195 196 private char m_highSurrogate = 0; 197 198 /** 199 * Default constructor 200 */ 201 public ToStream() { } 202 203 /** 204 * This helper method to writes out "]]>" when closing a CDATA section. 205 * 206 * @throws org.xml.sax.SAXException 207 */ 208 protected void closeCDATA() throws org.xml.sax.SAXException { 209 try { 210 m_writer.write(CDATA_DELIMITER_CLOSE); 211 // write out a CDATA section closing "]]>" 212 m_cdataTagOpen = false; // Remember that we have done so. 213 } 214 catch (IOException e) { 215 throw new SAXException(e); 216 } 217 } 218 219 /** 220 * Serializes the DOM node. Throws an exception only if an I/O 221 * exception occured while serializing. 222 * 223 * @param node Node to serialize. 224 * @throws IOException An I/O exception occured while serializing 225 */ 226 public void serialize(Node node) throws IOException { 227 try { 228 TreeWalker walker = new TreeWalker(this); 229 walker.traverse(node); 230 } catch (org.xml.sax.SAXException se) { 231 throw new WrappedRuntimeException(se); 232 } 233 } 234 235 /** 236 * Return true if the character is the high member of a surrogate pair. 237 * 238 * NEEDSDOC @param c 239 * 240 * NEEDSDOC ($objectName$) @return 241 */ 242 static final boolean isUTF16Surrogate(char c) { 243 return (c & 0xFC00) == 0xD800; 244 } 245 246 /** 247 * Taken from XSLTC 248 */ 249 private boolean m_escaping = true; 250 251 /** 252 * Flush the formatter's result stream. 253 * 254 * @throws org.xml.sax.SAXException 255 */ 256 protected final void flushWriter() throws org.xml.sax.SAXException { 257 final Writer writer = m_writer; 258 if (null != writer) { 259 try { 260 if (writer instanceof WriterToUTF8Buffered) { 261 if (m_shouldFlush) 262 ((WriterToUTF8Buffered)writer).flush(); 263 else 264 ((WriterToUTF8Buffered)writer).flushBuffer(); 265 } 266 if (writer instanceof WriterToASCI) { 267 if (m_shouldFlush) 268 writer.flush(); 269 } else { 270 // Flush always. 271 // Not a great thing if the writer was created 272 // by this class, but don't have a choice. 273 writer.flush(); 274 } 275 } catch (IOException ioe) { 276 throw new org.xml.sax.SAXException(ioe); 277 } 278 } 279 } 280 281 OutputStream m_outputStream; 282 283 /** 284 * Get the output stream where the events will be serialized to. 285 * 286 * @return reference to the result stream, or null of only a writer was 287 * set. 288 */ 289 public OutputStream getOutputStream() { 290 return m_outputStream; 291 } 292 293 // Implement DeclHandler 294 295 /** 296 * Report an element type declaration. 297 * 298 * <p>The content model will consist of the string "EMPTY", the 299 * string "ANY", or a parenthesised group, optionally followed 300 * by an occurrence indicator. The model will be normalized so 301 * that all whitespace is removed,and will include the enclosing 302 * parentheses.</p> 303 * 304 * @param name The element type name. 305 * @param model The content model as a normalized string. 306 * @exception SAXException The application may raise an exception. 307 */ 308 public void elementDecl(String name, String model) throws SAXException 309 { 310 // Do not inline external DTD 311 if (m_inExternalDTD) 312 return; 313 try { 314 final Writer writer = m_writer; 315 DTDprolog(); 316 317 writer.write("<!ELEMENT "); 318 writer.write(name); 319 writer.write(' '); 320 writer.write(model); 321 writer.write('>'); 322 writer.write(m_lineSep, 0, m_lineSepLen); 323 } 324 catch (IOException e) 325 { 326 throw new SAXException(e); 327 } 328 329 } 330 331 /** 332 * Report an internal entity declaration. 333 * 334 * <p>Only the effective (first) declaration for each entity 335 * will be reported.</p> 336 * 337 * @param name The name of the entity. If it is a parameter 338 * entity, the name will begin with '%'. 339 * @param value The replacement text of the entity. 340 * @exception SAXException The application may raise an exception. 341 * @see #externalEntityDecl 342 * @see org.xml.sax.DTDHandler#unparsedEntityDecl 343 */ 344 public void internalEntityDecl(String name, String value) 345 throws SAXException 346 { 347 // Do not inline external DTD 348 if (m_inExternalDTD) 349 return; 350 try { 351 DTDprolog(); 352 outputEntityDecl(name, value); 353 } catch (IOException e) { 354 throw new SAXException(e); 355 } 356 357 } 358 359 /** 360 * Output the doc type declaration. 361 * 362 * @param name non-null reference to document type name. 363 * NEEDSDOC @param value 364 * 365 * @throws org.xml.sax.SAXException 366 */ 367 void outputEntityDecl(String name, String value) throws IOException 368 { 369 final Writer writer = m_writer; 370 writer.write("<!ENTITY "); 371 writer.write(name); 372 writer.write(" \""); 373 writer.write(value); 374 writer.write("\">"); 375 writer.write(m_lineSep, 0, m_lineSepLen); 376 } 377 378 /** 379 * Output a system-dependent line break. 380 * 381 * @throws org.xml.sax.SAXException 382 */ 383 protected final void outputLineSep() throws IOException { 384 m_writer.write(m_lineSep, 0, m_lineSepLen); 385 } 386 387 void setProp(String name, String val, boolean defaultVal) { 388 if (val != null) { 389 390 char first = getFirstCharLocName(name); 391 switch (first) { 392 case 'c': 393 if (OutputKeys.CDATA_SECTION_ELEMENTS.equals(name)) { 394 addCdataSectionElements(val); // val is cdataSectionNames 395 } 396 break; 397 case 'd': 398 if (OutputKeys.DOCTYPE_SYSTEM.equals(name)) { 399 this.m_doctypeSystem = val; 400 } else if (OutputKeys.DOCTYPE_PUBLIC.equals(name)) { 401 this.m_doctypePublic = val; 402 if (val.startsWith("-//W3C//DTD XHTML")) 403 m_spaceBeforeClose = true; 404 } 405 break; 406 case 'e': 407 String newEncoding = val; 408 if (OutputKeys.ENCODING.equals(name)) { 409 String possible_encoding = Encodings.getMimeEncoding(val); 410 if (possible_encoding != null) { 411 // if the encoding is being set, try to get the 412 // preferred 413 // mime-name and set it too. 414 super.setProp("mime-name", possible_encoding, 415 defaultVal); 416 } 417 final String oldExplicitEncoding = getOutputPropertyNonDefault(OutputKeys.ENCODING); 418 final String oldDefaultEncoding = getOutputPropertyDefault(OutputKeys.ENCODING); 419 if ( (defaultVal && ( oldDefaultEncoding == null || !oldDefaultEncoding.equalsIgnoreCase(newEncoding))) 420 || ( !defaultVal && (oldExplicitEncoding == null || !oldExplicitEncoding.equalsIgnoreCase(newEncoding) ))) { 421 // We are trying to change the default or the non-default setting of the encoding to a different value 422 // from what it was 423 424 EncodingInfo encodingInfo = Encodings.getEncodingInfo(newEncoding); 425 if (newEncoding != null && encodingInfo.name == null) { 426 // We tried to get an EncodingInfo for Object for the given 427 // encoding, but it came back with an internall null name 428 // so the encoding is not supported by the JDK, issue a message. 429 final String msg = Utils.messages.createMessage( 430 MsgKey.ER_ENCODING_NOT_SUPPORTED,new Object[]{ newEncoding }); 431 432 final String msg2 = 433 "Warning: encoding \"" + newEncoding + "\" not supported, using " 434 + Encodings.DEFAULT_MIME_ENCODING; 435 try { 436 // Prepare to issue the warning message 437 final Transformer tran = super.getTransformer(); 438 if (tran != null) { 439 final ErrorListener errHandler = tran 440 .getErrorListener(); 441 // Issue the warning message 442 if (null != errHandler 443 && m_sourceLocator != null) { 444 errHandler 445 .warning(new TransformerException( 446 msg, m_sourceLocator)); 447 errHandler 448 .warning(new TransformerException( 449 msg2, m_sourceLocator)); 450 } else { 451 System.out.println(msg); 452 System.out.println(msg2); 453 } 454 } else { 455 System.out.println(msg); 456 System.out.println(msg2); 457 } 458 } catch (Exception e) { 459 } 460 461 // We said we are using UTF-8, so use it 462 newEncoding = Encodings.DEFAULT_MIME_ENCODING; 463 val = Encodings.DEFAULT_MIME_ENCODING; // to store the modified value into the properties a little later 464 encodingInfo = Encodings.getEncodingInfo(newEncoding); 465 } 466 // The encoding was good, or was forced to UTF-8 above 467 468 469 // If there is already a non-default set encoding and we 470 // are trying to set the default encoding, skip the this block 471 // as the non-default value is already the one to use. 472 if (defaultVal == false || oldExplicitEncoding == null) { 473 m_encodingInfo = encodingInfo; 474 if (newEncoding != null) 475 m_isUTF8 = newEncoding.equals(Encodings.DEFAULT_MIME_ENCODING); 476 477 // if there was a previously set OutputStream 478 OutputStream os = getOutputStream(); 479 if (os != null) { 480 Writer w = getWriter(); 481 482 // If the writer was previously set, but 483 // set by the user, or if the new encoding is the same 484 // as the old encoding, skip this block 485 String oldEncoding = getOutputProperty(OutputKeys.ENCODING); 486 if ((w == null || !m_writer_set_by_user) 487 && !newEncoding.equalsIgnoreCase(oldEncoding)) { 488 // Make the change of encoding in our internal 489 // table, then call setOutputStreamInternal 490 // which will stomp on the old Writer (if any) 491 // with a new Writer with the new encoding. 492 super.setProp(name, val, defaultVal); 493 setOutputStreamInternal(os,false); 494 } 495 } 496 } 497 } 498 } 499 break; 500 case 'i': 501 if (OutputPropertiesFactory.S_KEY_INDENT_AMOUNT.equals(name)) { 502 setIndentAmount(Integer.parseInt(val)); 503 } else if (OutputKeys.INDENT.equals(name)) { 504 boolean b = val.endsWith("yes") ? true : false; 505 m_doIndent = b; 506 } 507 508 break; 509 case 'l': 510 if (OutputPropertiesFactory.S_KEY_LINE_SEPARATOR.equals(name)) { 511 m_lineSep = val.toCharArray(); 512 m_lineSepLen = m_lineSep.length; 513 } 514 515 break; 516 case 'm': 517 if (OutputKeys.MEDIA_TYPE.equals(name)) { 518 m_mediatype = val; 519 } 520 break; 521 case 'o': 522 if (OutputKeys.OMIT_XML_DECLARATION.equals(name)) { 523 boolean b = val.endsWith("yes") ? true : false; 524 this.m_shouldNotWriteXMLHeader = b; 525 } 526 break; 527 case 's': 528 // if standalone was explicitly specified 529 if (OutputKeys.STANDALONE.equals(name)) { 530 if (defaultVal) { 531 setStandaloneInternal(val); 532 } else { 533 m_standaloneWasSpecified = true; 534 setStandaloneInternal(val); 535 } 536 } 537 538 break; 539 case 'v': 540 if (OutputKeys.VERSION.equals(name)) { 541 m_version = val; 542 } 543 break; 544 default: 545 break; 546 547 } 548 super.setProp(name, val, defaultVal); 549 } 550 } 551 552 /** 553 * Specifies an output format for this serializer. It the 554 * serializer has already been associated with an output format, 555 * it will switch to the new format. This method should not be 556 * called while the serializer is in the process of serializing 557 * a document. 558 * 559 * @param format The output format to use 560 */ 561 public void setOutputFormat(Properties format) { 562 boolean shouldFlush = m_shouldFlush; 563 564 if (format != null) { 565 // Set the default values first, 566 // and the non-default values after that, 567 // just in case there is some unexpected 568 // residual values left over from over-ridden default values 569 Enumeration<?> propNames; 570 propNames = format.propertyNames(); 571 while (propNames.hasMoreElements()) { 572 String key = (String) propNames.nextElement(); 573 // Get the value, possibly a default value 574 String value = format.getProperty(key); 575 // Get the non-default value (if any). 576 String explicitValue = (String) format.get(key); 577 if (explicitValue == null && value != null) { 578 // This is a default value 579 this.setOutputPropertyDefault(key,value); 580 } 581 if (explicitValue != null) { 582 // This is an explicit non-default value 583 this.setOutputProperty(key,explicitValue); 584 } 585 } 586 } 587 588 // Access this only from the Hashtable level... we don't want to 589 // get default properties. 590 String entitiesFileName = 591 (String) format.get(OutputPropertiesFactory.S_KEY_ENTITIES); 592 593 if (null != entitiesFileName) { 594 String method = (String) format.get(OutputKeys.METHOD); 595 m_charInfo = CharInfo.getCharInfo(entitiesFileName, method); 596 } 597 598 m_shouldFlush = shouldFlush; 599 } 600 601 /** 602 * Returns the output format for this serializer. 603 * 604 * @return The output format in use 605 */ 606 public Properties getOutputFormat() { 607 Properties def = new Properties(); 608 { 609 Set<String> s = getOutputPropDefaultKeys(); 610 for (String key : s) { 611 String val = getOutputPropertyDefault(key); 612 def.put(key, val); 613 } 614 } 615 616 Properties props = new Properties(def); 617 { 618 Set<String> s = getOutputPropKeys(); 619 for (String key : s) { 620 String val = getOutputPropertyNonDefault(key); 621 if (val != null) 622 props.put(key, val); 623 } 624 } 625 return props; 626 } 627 628 /** 629 * Specifies a writer to which the document should be serialized. 630 * This method should not be called while the serializer is in 631 * the process of serializing a document. 632 * 633 * @param writer The output writer stream 634 */ 635 public void setWriter(Writer writer) { 636 setWriterInternal(writer, true); 637 } 638 639 private boolean m_writer_set_by_user; 640 private void setWriterInternal(Writer writer, boolean setByUser) { 641 m_writer_set_by_user = setByUser; 642 m_writer = writer; 643 // if we are tracing events we need to trace what 644 // characters are written to the output writer. 645 if (m_tracer != null) { 646 boolean noTracerYet = true; 647 Writer w2 = m_writer; 648 while (w2 instanceof WriterChain) { 649 if (w2 instanceof SerializerTraceWriter) { 650 noTracerYet = false; 651 break; 652 } 653 w2 = ((WriterChain)w2).getWriter(); 654 } 655 if (noTracerYet) 656 m_writer = new SerializerTraceWriter(m_writer, m_tracer); 657 } 658 } 659 660 /** 661 * Set if the operating systems end-of-line line separator should 662 * be used when serializing. If set false NL character 663 * (decimal 10) is left alone, otherwise the new-line will be replaced on 664 * output with the systems line separator. For example on UNIX this is 665 * NL, while on Windows it is two characters, CR NL, where CR is the 666 * carriage-return (decimal 13). 667 * 668 * @param use_sytem_line_break True if an input NL is replaced with the 669 * operating systems end-of-line separator. 670 * @return The previously set value of the serializer. 671 */ 672 public boolean setLineSepUse(boolean use_sytem_line_break) { 673 boolean oldValue = m_lineSepUse; 674 m_lineSepUse = use_sytem_line_break; 675 return oldValue; 676 } 677 678 /** 679 * Specifies an output stream to which the document should be 680 * serialized. This method should not be called while the 681 * serializer is in the process of serializing a document. 682 * <p> 683 * The encoding specified in the output properties is used, or 684 * if no encoding was specified, the default for the selected 685 * output method. 686 * 687 * @param output The output stream 688 */ 689 public void setOutputStream(OutputStream output) { 690 setOutputStreamInternal(output, true); 691 } 692 693 private void setOutputStreamInternal(OutputStream output, boolean setByUser) 694 { 695 m_outputStream = output; 696 String encoding = getOutputProperty(OutputKeys.ENCODING); 697 if (Encodings.DEFAULT_MIME_ENCODING.equalsIgnoreCase(encoding)) 698 { 699 // We wrap the OutputStream with a writer, but 700 // not one set by the user 701 try { 702 setWriterInternal(new WriterToUTF8Buffered(output), false); 703 } catch (UnsupportedEncodingException e) { 704 e.printStackTrace(); 705 } 706 } else if ( 707 "WINDOWS-1250".equals(encoding) 708 || "US-ASCII".equals(encoding) 709 || "ASCII".equals(encoding)) 710 { 711 setWriterInternal(new WriterToASCI(output), false); 712 } else if (encoding != null) { 713 Writer osw = null; 714 try 715 { 716 osw = Encodings.getWriter(output, encoding); 717 } 718 catch (UnsupportedEncodingException uee) 719 { 720 osw = null; 721 } 722 723 724 if (osw == null) { 725 System.out.println( 726 "Warning: encoding \"" 727 + encoding 728 + "\" not supported" 729 + ", using " 730 + Encodings.DEFAULT_MIME_ENCODING); 731 732 encoding = Encodings.DEFAULT_MIME_ENCODING; 733 setEncoding(encoding); 734 try { 735 osw = Encodings.getWriter(output, encoding); 736 } catch (UnsupportedEncodingException e) { 737 // We can't really get here, UTF-8 is always supported 738 // This try-catch exists to make the compiler happy 739 e.printStackTrace(); 740 } 741 } 742 setWriterInternal(osw,false); 743 } 744 else { 745 // don't have any encoding, but we have an OutputStream 746 Writer osw = new OutputStreamWriter(output); 747 setWriterInternal(osw,false); 748 } 749 } 750 751 752 /** 753 * @see SerializationHandler#setEscaping(boolean) 754 */ 755 public boolean setEscaping(boolean escape) 756 { 757 final boolean temp = m_escaping; 758 m_escaping = escape; 759 return temp; 760 761 } 762 763 764 /** 765 * Might print a newline character and the indentation amount 766 * of the given depth. 767 * 768 * @param depth the indentation depth (element nesting depth) 769 * 770 * @throws org.xml.sax.SAXException if an error occurs during writing. 771 */ 772 protected void indent(int depth) throws IOException 773 { 774 775 if (m_startNewLine) 776 outputLineSep(); 777 /* 778 * Default value is 4, so printSpace directly. 779 */ 780 printSpace(depth * m_indentAmount); 781 782 } 783 784 /** 785 * Indent at the current element nesting depth. 786 * @throws IOException 787 */ 788 protected void indent() throws IOException 789 { 790 indent(m_elemContext.m_currentElemDepth); 791 } 792 /** 793 * Prints <var>n</var> spaces. 794 * @param n Number of spaces to print. 795 * 796 * @throws org.xml.sax.SAXException if an error occurs when writing. 797 */ 798 private void printSpace(int n) throws IOException 799 { 800 final Writer writer = m_writer; 801 for (int i = 0; i < n; i++) 802 { 803 writer.write(' '); 804 } 805 806 } 807 808 /** 809 * Report an attribute type declaration. 810 * 811 * <p>Only the effective (first) declaration for an attribute will 812 * be reported. The type will be one of the strings "CDATA", 813 * "ID", "IDREF", "IDREFS", "NMTOKEN", "NMTOKENS", "ENTITY", 814 * "ENTITIES", or "NOTATION", or a parenthesized token group with 815 * the separator "|" and all whitespace removed.</p> 816 * 817 * @param eName The name of the associated element. 818 * @param aName The name of the attribute. 819 * @param type A string representing the attribute type. 820 * @param valueDefault A string representing the attribute default 821 * ("#IMPLIED", "#REQUIRED", or "#FIXED") or null if 822 * none of these applies. 823 * @param value A string representing the attribute's default value, 824 * or null if there is none. 825 * @exception SAXException The application may raise an exception. 826 */ 827 public void attributeDecl( 828 String eName, 829 String aName, 830 String type, 831 String valueDefault, 832 String value) 833 throws SAXException 834 { 835 // Do not inline external DTD 836 if (m_inExternalDTD) 837 return; 838 try 839 { 840 final Writer writer = m_writer; 841 DTDprolog(); 842 843 writer.write("<!ATTLIST "); 844 writer.write(eName); 845 writer.write(' '); 846 847 writer.write(aName); 848 writer.write(' '); 849 writer.write(type); 850 if (valueDefault != null) 851 { 852 writer.write(' '); 853 writer.write(valueDefault); 854 } 855 856 //writer.write(" "); 857 //writer.write(value); 858 writer.write('>'); 859 writer.write(m_lineSep, 0, m_lineSepLen); 860 } 861 catch (IOException e) 862 { 863 throw new SAXException(e); 864 } 865 } 866 867 /** 868 * Get the character stream where the events will be serialized to. 869 * 870 * @return Reference to the result Writer, or null. 871 */ 872 public Writer getWriter() 873 { 874 return m_writer; 875 } 876 877 /** 878 * Report a parsed external entity declaration. 879 * 880 * <p>Only the effective (first) declaration for each entity 881 * will be reported.</p> 882 * 883 * @param name The name of the entity. If it is a parameter 884 * entity, the name will begin with '%'. 885 * @param publicId The declared public identifier of the entity, or 886 * null if none was declared. 887 * @param systemId The declared system identifier of the entity. 888 * @exception SAXException The application may raise an exception. 889 * @see #internalEntityDecl 890 * @see org.xml.sax.DTDHandler#unparsedEntityDecl 891 */ 892 public void externalEntityDecl( 893 String name, 894 String publicId, 895 String systemId) 896 throws SAXException 897 { 898 try { 899 DTDprolog(); 900 901 m_writer.write("<!ENTITY "); 902 m_writer.write(name); 903 if (publicId != null) { 904 m_writer.write(" PUBLIC \""); 905 m_writer.write(publicId); 906 907 } 908 else { 909 m_writer.write(" SYSTEM \""); 910 m_writer.write(systemId); 911 } 912 m_writer.write("\" >"); 913 m_writer.write(m_lineSep, 0, m_lineSepLen); 914 } catch (IOException e) { 915 // TODO Auto-generated catch block 916 e.printStackTrace(); 917 } 918 919 } 920 921 /** 922 * Tell if this character can be written without escaping. 923 */ 924 protected boolean escapingNotNeeded(char ch) 925 { 926 final boolean ret; 927 if (ch < 127) 928 { 929 // This is the old/fast code here, but is this 930 // correct for all encodings? 931 if (ch >= 0x20 || (0x0A == ch || 0x0D == ch || 0x09 == ch)) 932 ret= true; 933 else 934 ret = false; 935 } 936 else { 937 ret = m_encodingInfo.isInEncoding(ch); 938 } 939 return ret; 940 } 941 942 /** 943 * Once a surrogate has been detected, write out the pair of 944 * characters if it is in the encoding, or if there is no 945 * encoding, otherwise write out an entity reference 946 * of the value of the unicode code point of the character 947 * represented by the high/low surrogate pair. 948 * <p> 949 * An exception is thrown if there is no low surrogate in the pair, 950 * because the array ends unexpectely, or if the low char is there 951 * but its value is such that it is not a low surrogate. 952 * 953 * @param c the first (high) part of the surrogate, which 954 * must be confirmed before calling this method. 955 * @param ch Character array. 956 * @param i position Where the surrogate was detected. 957 * @param end The end index of the significant characters. 958 * @return the status of writing a surrogate pair. 959 * -1 -- nothing is written 960 * 0 -- the pair is written as-is 961 * code point -- the pair is written as an entity reference 962 * 963 * @throws IOException 964 * @throws org.xml.sax.SAXException if invalid UTF-16 surrogate detected. 965 */ 966 protected int writeUTF16Surrogate(char c, char ch[], int i, int end) 967 throws IOException, SAXException 968 { 969 int status = -1; 970 if (i + 1 >= end) 971 { 972 m_highSurrogate = c; 973 return status; 974 } 975 976 char high, low; 977 if (m_highSurrogate == 0) { 978 high = c; 979 low = ch[i+1]; 980 status = 0; 981 } else { 982 high = m_highSurrogate; 983 low = c; 984 m_highSurrogate = 0; 985 } 986 987 if (!Encodings.isLowUTF16Surrogate(low)) { 988 throwIOE(high, low); 989 } 990 991 final Writer writer = m_writer; 992 993 // If we make it to here we have a valid high, low surrogate pair 994 if (m_encodingInfo.isInEncoding(high,low)) { 995 // If the character formed by the surrogate pair 996 // is in the encoding, so just write it out 997 writer.write(new char[]{high, low}, 0, 2); 998 } 999 else { 1000 // Don't know what to do with this char, it is 1001 // not in the encoding and not a high char in 1002 // a surrogate pair, so write out as an entity ref 1003 final String encoding = getEncoding(); 1004 if (encoding != null) { 1005 status = writeCharRef(writer, high, low); 1006 } else { 1007 /* The output encoding is not known, 1008 * so just write it out as-is. 1009 */ 1010 writer.write(new char[]{high, low}, 0, 2); 1011 } 1012 } 1013 // non-zero only if character reference was written out. 1014 return status; 1015 } 1016 1017 /** 1018 * Handle one of the default entities, return false if it 1019 * is not a default entity. 1020 * 1021 * @param ch character to be escaped. 1022 * @param i index into character array. 1023 * @param chars non-null reference to character array. 1024 * @param len length of chars. 1025 * @param fromTextNode true if the characters being processed 1026 * are from a text node, false if they are from an attribute value 1027 * @param escLF true if the linefeed should be escaped. 1028 * 1029 * @return i+1 if the character was written, else i. 1030 * 1031 * @throws java.io.IOException 1032 */ 1033 protected int accumDefaultEntity( 1034 Writer writer, 1035 char ch, 1036 int i, 1037 char[] chars, 1038 int len, 1039 boolean fromTextNode, 1040 boolean escLF) 1041 throws IOException 1042 { 1043 1044 if (!escLF && CharInfo.S_LINEFEED == ch) 1045 { 1046 writer.write(m_lineSep, 0, m_lineSepLen); 1047 } 1048 else 1049 { 1050 // if this is text node character and a special one of those, 1051 // or if this is a character from attribute value and a special one of those 1052 if ((fromTextNode && m_charInfo.isSpecialTextChar(ch)) || (!fromTextNode && m_charInfo.isSpecialAttrChar(ch))) 1053 { 1054 String outputStringForChar = m_charInfo.getOutputStringForChar(ch); 1055 1056 if (null != outputStringForChar) 1057 { 1058 writer.write(outputStringForChar); 1059 } 1060 else 1061 return i; 1062 } 1063 else 1064 return i; 1065 } 1066 1067 return i + 1; 1068 1069 } 1070 /** 1071 * Normalize the characters, but don't escape. 1072 * 1073 * @param ch The characters from the XML document. 1074 * @param start The start position in the array. 1075 * @param length The number of characters to read from the array. 1076 * @param isCData true if a CDATA block should be built around the characters. 1077 * @param useSystemLineSeparator true if the operating systems 1078 * end-of-line separator should be output rather than a new-line character. 1079 * 1080 * @throws IOException 1081 * @throws org.xml.sax.SAXException 1082 */ 1083 void writeNormalizedChars( 1084 char ch[], 1085 int start, 1086 int length, 1087 boolean isCData, 1088 boolean useSystemLineSeparator) 1089 throws IOException, org.xml.sax.SAXException 1090 { 1091 final Writer writer = m_writer; 1092 int end = start + length; 1093 1094 for (int i = start; i < end; i++) 1095 { 1096 char c = ch[i]; 1097 1098 if (CharInfo.S_LINEFEED == c && useSystemLineSeparator) 1099 { 1100 writer.write(m_lineSep, 0, m_lineSepLen); 1101 } 1102 else if (isCData && (!escapingNotNeeded(c))) 1103 { 1104 i = handleEscaping(writer, c, ch, i, end); 1105 } 1106 else if ( 1107 isCData 1108 && ((i < (end - 2)) 1109 && (']' == c) 1110 && (']' == ch[i + 1]) 1111 && ('>' == ch[i + 2]))) 1112 { 1113 writer.write(CDATA_CONTINUE); 1114 1115 i += 2; 1116 } 1117 else 1118 { 1119 if (escapingNotNeeded(c)) 1120 { 1121 if (isCData && !m_cdataTagOpen) 1122 { 1123 writer.write(CDATA_DELIMITER_OPEN); 1124 m_cdataTagOpen = true; 1125 } 1126 writer.write(c); 1127 } 1128 else { 1129 i = handleEscaping(writer, c, ch, i, end); 1130 } 1131 } 1132 } 1133 1134 } 1135 1136 /** 1137 * Handles escaping, writes either with a surrogate pair or a character 1138 * reference. 1139 * 1140 * @param c the current char 1141 * @param ch the character array 1142 * @param i the current position 1143 * @param end the end index of the array 1144 * @return the next index 1145 * 1146 * @throws IOException 1147 * @throws org.xml.sax.SAXException if invalid UTF-16 surrogate detected. 1148 */ 1149 private int handleEscaping(Writer writer, char c, char ch[], int i, int end) 1150 throws IOException, SAXException { 1151 if (Encodings.isHighUTF16Surrogate(c) || Encodings.isLowUTF16Surrogate(c)) 1152 { 1153 if (writeUTF16Surrogate(c, ch, i, end) >= 0) { 1154 // move the index if the low surrogate is consumed 1155 // as writeUTF16Surrogate has written the pair 1156 if (Encodings.isHighUTF16Surrogate(c)) { 1157 i++ ; 1158 } 1159 } 1160 } 1161 else 1162 { 1163 writeCharRef(writer, c); 1164 } 1165 return i; 1166 } 1167 1168 /** 1169 * Ends an un-escaping section. 1170 * 1171 * @see #startNonEscaping 1172 * 1173 * @throws org.xml.sax.SAXException 1174 */ 1175 public void endNonEscaping() throws org.xml.sax.SAXException 1176 { 1177 m_disableOutputEscapingStates.pop(); 1178 } 1179 1180 /** 1181 * Starts an un-escaping section. All characters printed within an un- 1182 * escaping section are printed as is, without escaping special characters 1183 * into entity references. Only XML and HTML serializers need to support 1184 * this method. 1185 * <p> The contents of the un-escaping section will be delivered through the 1186 * regular <tt>characters</tt> event. 1187 * 1188 * @throws org.xml.sax.SAXException 1189 */ 1190 public void startNonEscaping() throws org.xml.sax.SAXException 1191 { 1192 m_disableOutputEscapingStates.push(true); 1193 } 1194 1195 /** 1196 * Receive notification of cdata. 1197 * 1198 * <p>The Parser will call this method to report each chunk of 1199 * character data. SAX parsers may return all contiguous character 1200 * data in a single chunk, or they may split it into several 1201 * chunks; however, all of the characters in any single event 1202 * must come from the same external entity, so that the Locator 1203 * provides useful information.</p> 1204 * 1205 * <p>The application must not attempt to read from the array 1206 * outside of the specified range.</p> 1207 * 1208 * <p>Note that some parsers will report whitespace using the 1209 * ignorableWhitespace() method rather than this one (validating 1210 * parsers must do so).</p> 1211 * 1212 * @param ch The characters from the XML document. 1213 * @param start The start position in the array. 1214 * @param length The number of characters to read from the array. 1215 * @throws org.xml.sax.SAXException Any SAX exception, possibly 1216 * wrapping another exception. 1217 * @see #ignorableWhitespace 1218 * @see org.xml.sax.Locator 1219 * 1220 * @throws org.xml.sax.SAXException 1221 */ 1222 protected void cdata(char ch[], int start, final int length) 1223 throws org.xml.sax.SAXException 1224 { 1225 try 1226 { 1227 final int old_start = start; 1228 if (m_elemContext.m_startTagOpen) 1229 { 1230 closeStartTag(); 1231 m_elemContext.m_startTagOpen = false; 1232 } 1233 1234 if (!m_cdataTagOpen && shouldIndent()) 1235 indent(); 1236 1237 boolean writeCDataBrackets = 1238 (((length >= 1) && escapingNotNeeded(ch[start]))); 1239 1240 /* Write out the CDATA opening delimiter only if 1241 * we are supposed to, and if we are not already in 1242 * the middle of a CDATA section 1243 */ 1244 if (writeCDataBrackets && !m_cdataTagOpen) 1245 { 1246 m_writer.write(CDATA_DELIMITER_OPEN); 1247 m_cdataTagOpen = true; 1248 } 1249 1250 // writer.write(ch, start, length); 1251 if (isEscapingDisabled()) 1252 { 1253 charactersRaw(ch, start, length); 1254 } 1255 else 1256 writeNormalizedChars(ch, start, length, true, m_lineSepUse); 1257 1258 /* used to always write out CDATA closing delimiter here, 1259 * but now we delay, so that we can merge CDATA sections on output. 1260 * need to write closing delimiter later 1261 */ 1262 if (writeCDataBrackets) 1263 { 1264 /* if the CDATA section ends with ] don't leave it open 1265 * as there is a chance that an adjacent CDATA sections 1266 * starts with ]>. 1267 * We don't want to merge ]] with > , or ] with ]> 1268 */ 1269 if (ch[start + length - 1] == ']') 1270 closeCDATA(); 1271 } 1272 1273 // time to fire off CDATA event 1274 if (m_tracer != null) 1275 super.fireCDATAEvent(ch, old_start, length); 1276 } 1277 catch (IOException ioe) 1278 { 1279 throw new org.xml.sax.SAXException( 1280 Utils.messages.createMessage( 1281 MsgKey.ER_OIERROR, 1282 null), 1283 ioe); 1284 //"IO error", ioe); 1285 } 1286 } 1287 1288 /** 1289 * Tell if the character escaping should be disabled for the current state. 1290 * 1291 * @return true if the character escaping should be disabled. 1292 */ 1293 private boolean isEscapingDisabled() 1294 { 1295 return m_disableOutputEscapingStates.peekOrFalse(); 1296 } 1297 1298 /** 1299 * If available, when the disable-output-escaping attribute is used, 1300 * output raw text without escaping. 1301 * 1302 * @param ch The characters from the XML document. 1303 * @param start The start position in the array. 1304 * @param length The number of characters to read from the array. 1305 * 1306 * @throws org.xml.sax.SAXException 1307 */ 1308 protected void charactersRaw(char ch[], int start, int length) 1309 throws org.xml.sax.SAXException 1310 { 1311 1312 if (isInEntityRef()) 1313 return; 1314 try 1315 { 1316 if (m_elemContext.m_startTagOpen) 1317 { 1318 closeStartTag(); 1319 m_elemContext.m_startTagOpen = false; 1320 } 1321 1322 m_writer.write(ch, start, length); 1323 } 1324 catch (IOException e) 1325 { 1326 throw new SAXException(e); 1327 } 1328 1329 } 1330 1331 /** 1332 * Receive notification of character data. 1333 * 1334 * <p>The Parser will call this method to report each chunk of 1335 * character data. SAX parsers may return all contiguous character 1336 * data in a single chunk, or they may split it into several 1337 * chunks; however, all of the characters in any single event 1338 * must come from the same external entity, so that the Locator 1339 * provides useful information.</p> 1340 * 1341 * <p>The application must not attempt to read from the array 1342 * outside of the specified range.</p> 1343 * 1344 * <p>Note that some parsers will report whitespace using the 1345 * ignorableWhitespace() method rather than this one (validating 1346 * parsers must do so).</p> 1347 * 1348 * @param chars The characters from the XML document. 1349 * @param start The start position in the array. 1350 * @param length The number of characters to read from the array. 1351 * @throws org.xml.sax.SAXException Any SAX exception, possibly 1352 * wrapping another exception. 1353 * @see #ignorableWhitespace 1354 * @see org.xml.sax.Locator 1355 * 1356 * @throws org.xml.sax.SAXException 1357 */ 1358 public void characters(final char chars[], final int start, final int length) 1359 throws org.xml.sax.SAXException 1360 { 1361 // It does not make sense to continue with rest of the method if the number of 1362 // characters to read from array is 0. 1363 // Section 7.6.1 of XSLT 1.0 (http://www.w3.org/TR/xslt#value-of) suggest no text node 1364 // is created if string is empty. 1365 if (length == 0 || (isInEntityRef())) 1366 return; 1367 1368 final boolean shouldNotFormat = !shouldFormatOutput(); 1369 if (m_elemContext.m_startTagOpen) 1370 { 1371 closeStartTag(); 1372 m_elemContext.m_startTagOpen = false; 1373 } 1374 else if (m_needToCallStartDocument) 1375 { 1376 startDocumentInternal(); 1377 } 1378 1379 if (m_cdataStartCalled || m_elemContext.m_isCdataSection) 1380 { 1381 /* either due to startCDATA() being called or due to 1382 * cdata-section-elements atribute, we need this as cdata 1383 */ 1384 cdata(chars, start, length); 1385 1386 return; 1387 } 1388 1389 if (m_cdataTagOpen) 1390 closeCDATA(); 1391 // the check with _escaping is a bit of a hack for XLSTC 1392 1393 if (m_disableOutputEscapingStates.peekOrFalse() || (!m_escaping)) 1394 { 1395 if (shouldNotFormat) { 1396 charactersRaw(chars, start, length); 1397 m_isprevtext = true; 1398 } else { 1399 m_charactersBuffer.addRawText(chars, start, length); 1400 } 1401 // time to fire off characters generation event 1402 if (m_tracer != null) 1403 super.fireCharEvent(chars, start, length); 1404 1405 return; 1406 } 1407 1408 if (m_elemContext.m_startTagOpen) 1409 { 1410 closeStartTag(); 1411 m_elemContext.m_startTagOpen = false; 1412 } 1413 1414 if (shouldNotFormat) { 1415 outputCharacters(chars, start, length); 1416 } else { 1417 m_charactersBuffer.addText(chars, start, length); 1418 } 1419 1420 // time to fire off characters generation event 1421 if (m_tracer != null) 1422 super.fireCharEvent(chars, start, length); 1423 } 1424 1425 1426 /** 1427 * This method checks if the content in current element should be formatted. 1428 * 1429 * @return True if the content should be formatted. 1430 */ 1431 protected boolean shouldFormatOutput() { 1432 return m_doIndent && !m_ispreserveSpace; 1433 } 1434 1435 /** 1436 * @return True if the content in current element should be formatted. 1437 */ 1438 public boolean getIndent() { 1439 return shouldFormatOutput(); 1440 } 1441 1442 /** 1443 * Write out the characters. 1444 * 1445 * @param chars The characters of the text. 1446 * @param start The start position in the char array. 1447 * @param length The number of characters from the char array. 1448 */ 1449 private void outputCharacters(final char chars[], final int start, final int length) throws SAXException { 1450 try 1451 { 1452 int i; 1453 char ch1; 1454 int startClean; 1455 1456 // skip any leading whitspace 1457 // don't go off the end and use a hand inlined version 1458 // of isWhitespace(ch) 1459 final int end = start + length; 1460 int lastDirty = start - 1; // last character that needed processing 1461 for (i = start; 1462 ((i < end) 1463 && ((ch1 = chars[i]) == 0x20 1464 || (ch1 == 0xA && m_lineSepUse) 1465 || ch1 == 0xD 1466 || ch1 == 0x09)); 1467 i++) 1468 { 1469 /* 1470 * We are processing leading whitespace, but are doing the same 1471 * processing for dirty characters here as for non-whitespace. 1472 * 1473 */ 1474 if (!m_charInfo.isTextASCIIClean(ch1)) 1475 { 1476 lastDirty = processDirty(chars,end, i,ch1, lastDirty, true); 1477 i = lastDirty; 1478 } 1479 } 1480 1481 // int lengthClean; // number of clean characters in a row 1482 // final boolean[] isAsciiClean = m_charInfo.getASCIIClean(); 1483 1484 final boolean isXML10 = XMLVERSION10.equals(getVersion()); 1485 // we've skipped the leading whitespace, now deal with the rest 1486 for (; i < end; i++) 1487 { 1488 { 1489 // A tight loop to skip over common clean chars 1490 // This tight loop makes it easier for the JIT 1491 // to optimize. 1492 char ch2; 1493 while (i<end 1494 && ((ch2 = chars[i])<127) 1495 && m_charInfo.isTextASCIIClean(ch2)) 1496 i++; 1497 if (i == end) 1498 break; 1499 } 1500 1501 final char ch = chars[i]; 1502 /* The check for isCharacterInC0orC1Ranger and 1503 * isNELorLSEPCharacter has been added 1504 * to support Control Characters in XML 1.1 1505 */ 1506 if (!isCharacterInC0orC1Range(ch) && 1507 (isXML10 || !isNELorLSEPCharacter(ch)) && 1508 (escapingNotNeeded(ch) && (!m_charInfo.isSpecialTextChar(ch))) 1509 || ('"' == ch)) 1510 { 1511 ; // a character needing no special processing 1512 } 1513 else 1514 { 1515 lastDirty = processDirty(chars,end, i, ch, lastDirty, true); 1516 i = lastDirty; 1517 } 1518 } 1519 1520 // we've reached the end. Any clean characters at the 1521 // end of the array than need to be written out? 1522 startClean = lastDirty + 1; 1523 if (i > startClean) 1524 { 1525 int lengthClean = i - startClean; 1526 m_writer.write(chars, startClean, lengthClean); 1527 } 1528 1529 // For indentation purposes, mark that we've just writen text out 1530 m_isprevtext = true; 1531 } 1532 catch (IOException e) 1533 { 1534 throw new SAXException(e); 1535 } 1536 } 1537 1538 /** 1539 * Used to flush the buffered characters when indentation is on, this method 1540 * will be called when the next node is traversed. 1541 * 1542 */ 1543 final protected void flushCharactersBuffer() throws SAXException { 1544 try { 1545 if (shouldFormatOutput() && m_charactersBuffer.isAnyCharactersBuffered()) { 1546 if (m_elemContext.m_isCdataSection) { 1547 /* 1548 * due to cdata-section-elements atribute, we need this as 1549 * cdata 1550 */ 1551 char[] chars = m_charactersBuffer.toChars(); 1552 cdata(chars, 0, chars.length); 1553 return; 1554 } 1555 1556 m_childNodeNum++; 1557 boolean skipBeginningNewlines = false; 1558 if (shouldIndentForText()) { 1559 indent(); 1560 m_startNewLine = true; 1561 // newline has always been added here because if this is the 1562 // text before the first element, shouldIndent() won't 1563 // return true. 1564 skipBeginningNewlines = true; 1565 } 1566 m_charactersBuffer.flush(skipBeginningNewlines); 1567 } 1568 } catch (IOException e) { 1569 throw new SAXException(e); 1570 } finally { 1571 m_charactersBuffer.clear(); 1572 } 1573 } 1574 1575 /** 1576 * True if should indent in flushCharactersBuffer method. 1577 * This method may be overridden in sub-class. 1578 * 1579 */ 1580 protected boolean shouldIndentForText() { 1581 return (shouldIndent() && m_childNodeNum > 1); 1582 } 1583 1584 /** 1585 * This method checks if a given character is between C0 or C1 range 1586 * of Control characters. 1587 * This method is added to support Control Characters for XML 1.1 1588 * If a given character is TAB (0x09), LF (0x0A) or CR (0x0D), this method 1589 * return false. Since they are whitespace characters, no special processing is needed. 1590 * 1591 * @param ch 1592 * @return boolean 1593 */ 1594 private static boolean isCharacterInC0orC1Range(char ch) 1595 { 1596 if(ch == 0x09 || ch == 0x0A || ch == 0x0D) 1597 return false; 1598 else 1599 return (ch >= 0x7F && ch <= 0x9F)|| (ch >= 0x01 && ch <= 0x1F); 1600 } 1601 /** 1602 * This method checks if a given character either NEL (0x85) or LSEP (0x2028) 1603 * These are new end of line charcters added in XML 1.1. These characters must be 1604 * written as Numeric Character References (NCR) in XML 1.1 output document. 1605 * 1606 * @param ch 1607 * @return boolean 1608 */ 1609 private static boolean isNELorLSEPCharacter(char ch) 1610 { 1611 return (ch == 0x85 || ch == 0x2028); 1612 } 1613 /** 1614 * Process a dirty character and any preeceding clean characters 1615 * that were not yet processed. 1616 * @param chars array of characters being processed 1617 * @param end one (1) beyond the last character 1618 * in chars to be processed 1619 * @param i the index of the dirty character 1620 * @param ch the character in chars[i] 1621 * @param lastDirty the last dirty character previous to i 1622 * @param fromTextNode true if the characters being processed are 1623 * from a text node, false if they are from an attribute value. 1624 * @return the index of the last character processed 1625 */ 1626 private int processDirty( 1627 char[] chars, 1628 int end, 1629 int i, 1630 char ch, 1631 int lastDirty, 1632 boolean fromTextNode) throws IOException, SAXException 1633 { 1634 int startClean = lastDirty + 1; 1635 // if we have some clean characters accumulated 1636 // process them before the dirty one. 1637 if (i > startClean) 1638 { 1639 int lengthClean = i - startClean; 1640 m_writer.write(chars, startClean, lengthClean); 1641 } 1642 1643 // process the "dirty" character 1644 if (CharInfo.S_LINEFEED == ch && fromTextNode) 1645 { 1646 m_writer.write(m_lineSep, 0, m_lineSepLen); 1647 } 1648 else 1649 { 1650 startClean = 1651 accumDefaultEscape( 1652 m_writer, 1653 ch, 1654 i, 1655 chars, 1656 end, 1657 fromTextNode, 1658 false); 1659 i = startClean - 1; 1660 } 1661 // Return the index of the last character that we just processed 1662 // which is a dirty character. 1663 return i; 1664 } 1665 1666 /** 1667 * Receive notification of character data. 1668 * 1669 * @param s The string of characters to process. 1670 * 1671 * @throws org.xml.sax.SAXException 1672 */ 1673 public void characters(String s) throws org.xml.sax.SAXException 1674 { 1675 if (isInEntityRef()) 1676 return; 1677 final int length = s.length(); 1678 if (length > m_charsBuff.length) 1679 { 1680 m_charsBuff = new char[length * 2 + 1]; 1681 } 1682 s.getChars(0, length, m_charsBuff, 0); 1683 characters(m_charsBuff, 0, length); 1684 } 1685 1686 /** 1687 * Escape and writer.write a character. 1688 * 1689 * @param ch character to be escaped. 1690 * @param i index into character array. 1691 * @param chars non-null reference to character array. 1692 * @param len length of chars. 1693 * @param fromTextNode true if the characters being processed are 1694 * from a text node, false if the characters being processed are from 1695 * an attribute value. 1696 * @param escLF true if the linefeed should be escaped. 1697 * 1698 * @return i+1 if a character was written, i+2 if two characters 1699 * were written out, else return i. 1700 * 1701 * @throws org.xml.sax.SAXException 1702 */ 1703 protected int accumDefaultEscape( 1704 Writer writer, 1705 char ch, 1706 int i, 1707 char[] chars, 1708 int len, 1709 boolean fromTextNode, 1710 boolean escLF) 1711 throws IOException, SAXException 1712 { 1713 1714 int pos = accumDefaultEntity(writer, ch, i, chars, len, fromTextNode, escLF); 1715 1716 if (i == pos) 1717 { 1718 if (m_highSurrogate != 0) { 1719 if (!(Encodings.isLowUTF16Surrogate(ch))) { 1720 throwIOE(m_highSurrogate, ch); 1721 } 1722 writeCharRef(writer, m_highSurrogate, ch); 1723 m_highSurrogate = 0; 1724 return ++pos; 1725 } 1726 1727 if (Encodings.isHighUTF16Surrogate(ch)) 1728 { 1729 if (i + 1 >= len) 1730 { 1731 // save for the next read 1732 m_highSurrogate = ch; 1733 pos++; 1734 } 1735 else 1736 { 1737 // the next should be the UTF-16 low surrogate of the hig/low pair. 1738 char next = chars[++i]; 1739 if (!(Encodings.isLowUTF16Surrogate(next))) 1740 throwIOE(ch, next); 1741 1742 writeCharRef(writer, ch, next); 1743 pos += 2; // count the two characters that went into writing out this entity 1744 } 1745 } 1746 else 1747 { 1748 /* This if check is added to support control characters in XML 1.1. 1749 * If a character is a Control Character within C0 and C1 range, it is desirable 1750 * to write it out as Numeric Character Reference(NCR) regardless of XML Version 1751 * being used for output document. 1752 */ 1753 if (isCharacterInC0orC1Range(ch) || 1754 (XMLVERSION11.equals(getVersion()) && isNELorLSEPCharacter(ch))) 1755 { 1756 writeCharRef(writer, ch); 1757 } 1758 else if ((!escapingNotNeeded(ch) || 1759 ( (fromTextNode && m_charInfo.isSpecialTextChar(ch)) 1760 || (!fromTextNode && m_charInfo.isSpecialAttrChar(ch)))) 1761 && m_elemContext.m_currentElemDepth > 0) 1762 { 1763 writeCharRef(writer, ch); 1764 } 1765 else 1766 { 1767 writer.write(ch); 1768 } 1769 pos++; // count the single character that was processed 1770 } 1771 1772 } 1773 return pos; 1774 } 1775 1776 /** 1777 * Writes out a character reference. 1778 * @param writer the writer 1779 * @param c the character 1780 * @throws IOException 1781 */ 1782 private void writeCharRef(Writer writer, char c) throws IOException, SAXException { 1783 if (m_cdataTagOpen) 1784 closeCDATA(); 1785 writer.write("&#"); 1786 writer.write(Integer.toString(c)); 1787 writer.write(';'); 1788 } 1789 1790 /** 1791 * Writes out a pair of surrogates as a character reference 1792 * @param writer the writer 1793 * @param high the high surrogate 1794 * @param low the low surrogate 1795 * @throws IOException 1796 */ 1797 private int writeCharRef(Writer writer, char high, char low) throws IOException, SAXException { 1798 if (m_cdataTagOpen) 1799 closeCDATA(); 1800 // Unicode code point formed from the high/low pair. 1801 int codePoint = Encodings.toCodePoint(high, low); 1802 writer.write("&#"); 1803 writer.write(Integer.toString(codePoint)); 1804 writer.write(';'); 1805 return codePoint; 1806 } 1807 1808 private void throwIOE(char ch, char next) throws IOException { 1809 throw new IOException(Utils.messages.createMessage( 1810 MsgKey.ER_INVALID_UTF16_SURROGATE, 1811 new Object[] {Integer.toHexString(ch) + " " 1812 + Integer.toHexString(next)})); 1813 } 1814 1815 /** 1816 * Receive notification of the beginning of an element, although this is a 1817 * SAX method additional namespace or attribute information can occur before 1818 * or after this call, that is associated with this element. 1819 * 1820 * 1821 * @param namespaceURI The Namespace URI, or the empty string if the 1822 * element has no Namespace URI or if Namespace 1823 * processing is not being performed. 1824 * @param localName The local name (without prefix), or the 1825 * empty string if Namespace processing is not being 1826 * performed. 1827 * @param name The element type name. 1828 * @param atts The attributes attached to the element, if any. 1829 * @throws org.xml.sax.SAXException Any SAX exception, possibly 1830 * wrapping another exception. 1831 * @see org.xml.sax.ContentHandler#startElement 1832 * @see org.xml.sax.ContentHandler#endElement 1833 * @see org.xml.sax.AttributeList 1834 * 1835 * @throws org.xml.sax.SAXException 1836 */ 1837 public void startElement( 1838 String namespaceURI, 1839 String localName, 1840 String name, 1841 Attributes atts) 1842 throws org.xml.sax.SAXException 1843 { 1844 if (isInEntityRef()) 1845 return; 1846 1847 if (m_doIndent) { 1848 m_childNodeNum++; 1849 flushCharactersBuffer(); 1850 } 1851 1852 if (m_needToCallStartDocument) 1853 { 1854 startDocumentInternal(); 1855 m_needToCallStartDocument = false; 1856 } 1857 else if (m_cdataTagOpen) 1858 closeCDATA(); 1859 try 1860 { 1861 if ((true == m_needToOutputDocTypeDecl) 1862 && (null != getDoctypeSystem())) 1863 { 1864 outputDocTypeDecl(name, true); 1865 } 1866 1867 m_needToOutputDocTypeDecl = false; 1868 1869 /* before we over-write the current elementLocalName etc. 1870 * lets close out the old one (if we still need to) 1871 */ 1872 if (m_elemContext.m_startTagOpen) 1873 { 1874 closeStartTag(); 1875 m_elemContext.m_startTagOpen = false; 1876 } 1877 1878 if (namespaceURI != null) 1879 ensurePrefixIsDeclared(namespaceURI, name); 1880 1881 if (shouldIndent() && m_startNewLine) 1882 { 1883 indent(); 1884 } 1885 1886 m_startNewLine = true; 1887 1888 final Writer writer = m_writer; 1889 writer.write('<'); 1890 writer.write(name); 1891 } 1892 catch (IOException e) 1893 { 1894 throw new SAXException(e); 1895 } 1896 1897 // process the attributes now, because after this SAX call they might be gone 1898 if (atts != null) 1899 addAttributes(atts); 1900 1901 if (m_doIndent) { 1902 m_ispreserveSpace = m_preserveSpaces.peekOrFalse(); 1903 m_preserveSpaces.push(m_ispreserveSpace); 1904 1905 m_childNodeNumStack.add(m_childNodeNum); 1906 m_childNodeNum = 0; 1907 } 1908 1909 m_elemContext = m_elemContext.push(namespaceURI,localName,name); 1910 m_isprevtext = false; 1911 1912 if (m_tracer != null){ 1913 firePseudoAttributes(); 1914 } 1915 1916 } 1917 1918 /** 1919 * Receive notification of the beginning of an element, additional 1920 * namespace or attribute information can occur before or after this call, 1921 * that is associated with this element. 1922 * 1923 * 1924 * @param elementNamespaceURI The Namespace URI, or the empty string if the 1925 * element has no Namespace URI or if Namespace 1926 * processing is not being performed. 1927 * @param elementLocalName The local name (without prefix), or the 1928 * empty string if Namespace processing is not being 1929 * performed. 1930 * @param elementName The element type name. 1931 * @throws org.xml.sax.SAXException Any SAX exception, possibly 1932 * wrapping another exception. 1933 * @see org.xml.sax.ContentHandler#startElement 1934 * @see org.xml.sax.ContentHandler#endElement 1935 * @see org.xml.sax.AttributeList 1936 * 1937 * @throws org.xml.sax.SAXException 1938 */ 1939 public void startElement( 1940 String elementNamespaceURI, 1941 String elementLocalName, 1942 String elementName) 1943 throws SAXException 1944 { 1945 startElement(elementNamespaceURI, elementLocalName, elementName, null); 1946 } 1947 1948 public void startElement(String elementName) throws SAXException 1949 { 1950 startElement(null, null, elementName, null); 1951 } 1952 1953 /** 1954 * Output the doc type declaration. 1955 * 1956 * @param name non-null reference to document type name. 1957 * NEEDSDOC @param closeDecl 1958 * 1959 * @throws java.io.IOException 1960 */ 1961 void outputDocTypeDecl(String name, boolean closeDecl) throws SAXException 1962 { 1963 if (m_cdataTagOpen) 1964 closeCDATA(); 1965 try 1966 { 1967 final Writer writer = m_writer; 1968 writer.write("<!DOCTYPE "); 1969 writer.write(name); 1970 1971 String doctypePublic = getDoctypePublic(); 1972 if (null != doctypePublic) 1973 { 1974 writer.write(" PUBLIC \""); 1975 writer.write(doctypePublic); 1976 writer.write('\"'); 1977 } 1978 1979 String doctypeSystem = getDoctypeSystem(); 1980 if (null != doctypeSystem) 1981 { 1982 if (null == doctypePublic) 1983 writer.write(" SYSTEM \""); 1984 else 1985 writer.write(" \""); 1986 1987 writer.write(doctypeSystem); 1988 1989 if (closeDecl) 1990 { 1991 writer.write("\">"); 1992 writer.write(m_lineSep, 0, m_lineSepLen); 1993 closeDecl = false; // done closing 1994 } 1995 else 1996 writer.write('\"'); 1997 } 1998 boolean dothis = false; 1999 if (dothis) 2000 { 2001 // at one point this code seemed right, 2002 // but not anymore - Brian M. 2003 if (closeDecl) 2004 { 2005 writer.write('>'); 2006 writer.write(m_lineSep, 0, m_lineSepLen); 2007 } 2008 } 2009 } 2010 catch (IOException e) 2011 { 2012 throw new SAXException(e); 2013 } 2014 } 2015 2016 /** 2017 * Process the attributes, which means to write out the currently 2018 * collected attributes to the writer. The attributes are not 2019 * cleared by this method 2020 * 2021 * @param writer the writer to write processed attributes to. 2022 * @param nAttrs the number of attributes in m_attributes 2023 * to be processed 2024 * 2025 * @throws java.io.IOException 2026 * @throws org.xml.sax.SAXException 2027 */ 2028 public void processAttributes(Writer writer, int nAttrs) throws IOException, SAXException 2029 { 2030 /* real SAX attributes are not passed in, so process the 2031 * attributes that were collected after the startElement call. 2032 * _attribVector is a "cheap" list for Stream serializer output 2033 * accumulated over a series of calls to attribute(name,value) 2034 */ 2035 String encoding = getEncoding(); 2036 for (int i = 0; i < nAttrs; i++) 2037 { 2038 // elementAt is JDK 1.1.8 2039 final String name = m_attributes.getQName(i); 2040 final String value = m_attributes.getValue(i); 2041 writer.write(' '); 2042 writer.write(name); 2043 writer.write("=\""); 2044 writeAttrString(writer, value, encoding); 2045 writer.write('\"'); 2046 } 2047 } 2048 2049 /** 2050 * Returns the specified <var>string</var> after substituting <VAR>specials</VAR>, 2051 * and UTF-16 surrogates for chracter references <CODE>&#xnn</CODE>. 2052 * 2053 * @param string String to convert to XML format. 2054 * @param encoding CURRENTLY NOT IMPLEMENTED. 2055 * 2056 * @throws java.io.IOException 2057 */ 2058 public void writeAttrString( 2059 Writer writer, 2060 String string, 2061 String encoding) 2062 throws IOException, SAXException 2063 { 2064 final int len = string.length(); 2065 if (len > m_attrBuff.length) 2066 { 2067 m_attrBuff = new char[len*2 + 1]; 2068 } 2069 string.getChars(0,len, m_attrBuff, 0); 2070 final char[] stringChars = m_attrBuff; 2071 2072 for (int i = 0; i < len; ) 2073 { 2074 char ch = stringChars[i]; 2075 if (escapingNotNeeded(ch) && (!m_charInfo.isSpecialAttrChar(ch))) 2076 { 2077 writer.write(ch); 2078 i++; 2079 } 2080 else 2081 { // I guess the parser doesn't normalize cr/lf in attributes. -sb 2082 // if ((CharInfo.S_CARRIAGERETURN == ch) 2083 // && ((i + 1) < len) 2084 // && (CharInfo.S_LINEFEED == stringChars[i + 1])) 2085 // { 2086 // i++; 2087 // ch = CharInfo.S_LINEFEED; 2088 // } 2089 2090 i = accumDefaultEscape(writer, ch, i, stringChars, len, false, true); 2091 } 2092 } 2093 2094 } 2095 2096 /** 2097 * Receive notification of the end of an element. 2098 * 2099 * 2100 * @param namespaceURI The Namespace URI, or the empty string if the 2101 * element has no Namespace URI or if Namespace 2102 * processing is not being performed. 2103 * @param localName The local name (without prefix), or the 2104 * empty string if Namespace processing is not being 2105 * performed. 2106 * @param name The element type name 2107 * @throws org.xml.sax.SAXException Any SAX exception, possibly 2108 * wrapping another exception. 2109 * 2110 * @throws org.xml.sax.SAXException 2111 */ 2112 public void endElement(String namespaceURI, String localName, String name) 2113 throws org.xml.sax.SAXException 2114 { 2115 2116 if (isInEntityRef()) 2117 return; 2118 2119 if (m_doIndent) { 2120 flushCharactersBuffer(); 2121 } 2122 // namespaces declared at the current depth are no longer valid 2123 // so get rid of them 2124 m_prefixMap.popNamespaces(m_elemContext.m_currentElemDepth, null); 2125 2126 try 2127 { 2128 final Writer writer = m_writer; 2129 if (m_elemContext.m_startTagOpen) 2130 { 2131 if (m_tracer != null) 2132 super.fireStartElem(m_elemContext.m_elementName); 2133 int nAttrs = m_attributes.getLength(); 2134 if (nAttrs > 0) 2135 { 2136 processAttributes(m_writer, nAttrs); 2137 // clear attributes object for re-use with next element 2138 m_attributes.clear(); 2139 } 2140 if (m_spaceBeforeClose) 2141 writer.write(" />"); 2142 else 2143 writer.write("/>"); 2144 /* don't need to pop cdataSectionState because 2145 * this element ended so quickly that we didn't get 2146 * to push the state. 2147 */ 2148 2149 } 2150 else 2151 { 2152 if (m_cdataTagOpen) 2153 closeCDATA(); 2154 2155 if (shouldIndent() && (m_childNodeNum > 1 || !m_isprevtext)) 2156 indent(m_elemContext.m_currentElemDepth - 1); 2157 writer.write('<'); 2158 writer.write('/'); 2159 writer.write(name); 2160 writer.write('>'); 2161 } 2162 } 2163 catch (IOException e) 2164 { 2165 throw new SAXException(e); 2166 } 2167 2168 if (m_doIndent) { 2169 m_ispreserveSpace = m_preserveSpaces.popAndTop(); 2170 m_childNodeNum = m_childNodeNumStack.remove(m_childNodeNumStack.size() - 1); 2171 2172 m_isprevtext = false; 2173 } 2174 2175 // fire off the end element event 2176 if (m_tracer != null) 2177 super.fireEndElem(name); 2178 m_elemContext = m_elemContext.m_prev; 2179 } 2180 2181 /** 2182 * Receive notification of the end of an element. 2183 * @param name The element type name 2184 * @throws org.xml.sax.SAXException Any SAX exception, possibly 2185 * wrapping another exception. 2186 */ 2187 public void endElement(String name) throws org.xml.sax.SAXException 2188 { 2189 endElement(null, null, name); 2190 } 2191 2192 /** 2193 * Begin the scope of a prefix-URI Namespace mapping 2194 * just before another element is about to start. 2195 * This call will close any open tags so that the prefix mapping 2196 * will not apply to the current element, but the up comming child. 2197 * 2198 * @see org.xml.sax.ContentHandler#startPrefixMapping 2199 * 2200 * @param prefix The Namespace prefix being declared. 2201 * @param uri The Namespace URI the prefix is mapped to. 2202 * 2203 * @throws org.xml.sax.SAXException The client may throw 2204 * an exception during processing. 2205 * 2206 */ 2207 public void startPrefixMapping(String prefix, String uri) 2208 throws org.xml.sax.SAXException 2209 { 2210 // the "true" causes the flush of any open tags 2211 startPrefixMapping(prefix, uri, true); 2212 } 2213 2214 /** 2215 * Handle a prefix/uri mapping, which is associated with a startElement() 2216 * that is soon to follow. Need to close any open start tag to make 2217 * sure than any name space attributes due to this event are associated wih 2218 * the up comming element, not the current one. 2219 * @see ExtendedContentHandler#startPrefixMapping 2220 * 2221 * @param prefix The Namespace prefix being declared. 2222 * @param uri The Namespace URI the prefix is mapped to. 2223 * @param shouldFlush true if any open tags need to be closed first, this 2224 * will impact which element the mapping applies to (open parent, or its up 2225 * comming child) 2226 * @return returns true if the call made a change to the current 2227 * namespace information, false if it did not change anything, e.g. if the 2228 * prefix/namespace mapping was already in scope from before. 2229 * 2230 * @throws org.xml.sax.SAXException The client may throw 2231 * an exception during processing. 2232 * 2233 * 2234 */ 2235 public boolean startPrefixMapping( 2236 String prefix, 2237 String uri, 2238 boolean shouldFlush) 2239 throws org.xml.sax.SAXException 2240 { 2241 2242 /* Remember the mapping, and at what depth it was declared 2243 * This is one greater than the current depth because these 2244 * mappings will apply to the next depth. This is in 2245 * consideration that startElement() will soon be called 2246 */ 2247 2248 boolean pushed; 2249 int pushDepth; 2250 if (shouldFlush) 2251 { 2252 flushPending(); 2253 // the prefix mapping applies to the child element (one deeper) 2254 pushDepth = m_elemContext.m_currentElemDepth + 1; 2255 } 2256 else 2257 { 2258 // the prefix mapping applies to the current element 2259 pushDepth = m_elemContext.m_currentElemDepth; 2260 } 2261 pushed = m_prefixMap.pushNamespace(prefix, uri, pushDepth); 2262 2263 if (pushed) 2264 { 2265 /* Brian M.: don't know if we really needto do this. The 2266 * callers of this object should have injected both 2267 * startPrefixMapping and the attributes. We are 2268 * just covering our butt here. 2269 */ 2270 String name; 2271 if (EMPTYSTRING.equals(prefix)) 2272 { 2273 name = "xmlns"; 2274 addAttributeAlways(XMLNS_URI, name, name, "CDATA", uri, false); 2275 } 2276 else 2277 { 2278 if (!EMPTYSTRING.equals(uri)) 2279 // hack for XSLTC attribset16 test 2280 { // that maps ns1 prefix to "" URI 2281 name = "xmlns:" + prefix; 2282 2283 /* for something like xmlns:abc="w3.pretend.org" 2284 * the uri is the value, that is why we pass it in the 2285 * value, or 5th slot of addAttributeAlways() 2286 */ 2287 addAttributeAlways(XMLNS_URI, prefix, name, "CDATA", uri, false); 2288 } 2289 } 2290 } 2291 return pushed; 2292 } 2293 2294 /** 2295 * Receive notification of an XML comment anywhere in the document. This 2296 * callback will be used for comments inside or outside the document 2297 * element, including comments in the external DTD subset (if read). 2298 * @param ch An array holding the characters in the comment. 2299 * @param start The starting position in the array. 2300 * @param length The number of characters to use from the array. 2301 * @throws org.xml.sax.SAXException The application may raise an exception. 2302 */ 2303 public void comment(char ch[], int start, int length) 2304 throws org.xml.sax.SAXException 2305 { 2306 2307 int start_old = start; 2308 if (isInEntityRef()) 2309 return; 2310 if (m_doIndent) { 2311 m_childNodeNum++; 2312 flushCharactersBuffer(); 2313 } 2314 if (m_elemContext.m_startTagOpen) 2315 { 2316 closeStartTag(); 2317 m_elemContext.m_startTagOpen = false; 2318 } 2319 else if (m_needToCallStartDocument) 2320 { 2321 startDocumentInternal(); 2322 m_needToCallStartDocument = false; 2323 } 2324 2325 try 2326 { 2327 if (shouldIndent() && m_isStandalone) 2328 indent(); 2329 2330 final int limit = start + length; 2331 boolean wasDash = false; 2332 if (m_cdataTagOpen) 2333 closeCDATA(); 2334 2335 if (shouldIndent() && !m_isStandalone) 2336 indent(); 2337 2338 final Writer writer = m_writer; 2339 writer.write(COMMENT_BEGIN); 2340 // Detect occurrences of two consecutive dashes, handle as necessary. 2341 for (int i = start; i < limit; i++) 2342 { 2343 if (wasDash && ch[i] == '-') 2344 { 2345 writer.write(ch, start, i - start); 2346 writer.write(" -"); 2347 start = i + 1; 2348 } 2349 wasDash = (ch[i] == '-'); 2350 } 2351 2352 // if we have some chars in the comment 2353 if (length > 0) 2354 { 2355 // Output the remaining characters (if any) 2356 final int remainingChars = (limit - start); 2357 if (remainingChars > 0) 2358 writer.write(ch, start, remainingChars); 2359 // Protect comment end from a single trailing dash 2360 if (ch[limit - 1] == '-') 2361 writer.write(' '); 2362 } 2363 writer.write(COMMENT_END); 2364 } 2365 catch (IOException e) 2366 { 2367 throw new SAXException(e); 2368 } 2369 2370 /* 2371 * Don't write out any indentation whitespace now, 2372 * because there may be non-whitespace text after this. 2373 * 2374 * Simply mark that at this point if we do decide 2375 * to indent that we should 2376 * add a newline on the end of the current line before 2377 * the indentation at the start of the next line. 2378 */ 2379 m_startNewLine = true; 2380 // time to generate comment event 2381 if (m_tracer != null) 2382 super.fireCommentEvent(ch, start_old,length); 2383 } 2384 2385 /** 2386 * Report the end of a CDATA section. 2387 * @throws org.xml.sax.SAXException The application may raise an exception. 2388 * 2389 * @see #startCDATA 2390 */ 2391 public void endCDATA() throws org.xml.sax.SAXException 2392 { 2393 if (m_cdataTagOpen) 2394 closeCDATA(); 2395 m_cdataStartCalled = false; 2396 } 2397 2398 /** 2399 * Report the end of DTD declarations. 2400 * @throws org.xml.sax.SAXException The application may raise an exception. 2401 * @see #startDTD 2402 */ 2403 public void endDTD() throws org.xml.sax.SAXException 2404 { 2405 try 2406 { 2407 // Don't output doctype declaration until startDocumentInternal 2408 // has been called. Otherwise, it can appear before XML decl. 2409 if (m_needToCallStartDocument) { 2410 return; 2411 } 2412 2413 if (m_needToOutputDocTypeDecl) 2414 { 2415 outputDocTypeDecl(m_elemContext.m_elementName, false); 2416 m_needToOutputDocTypeDecl = false; 2417 } 2418 final Writer writer = m_writer; 2419 if (!m_inDoctype) 2420 writer.write("]>"); 2421 else 2422 { 2423 writer.write('>'); 2424 } 2425 2426 writer.write(m_lineSep, 0, m_lineSepLen); 2427 } 2428 catch (IOException e) 2429 { 2430 throw new SAXException(e); 2431 } 2432 2433 } 2434 2435 /** 2436 * End the scope of a prefix-URI Namespace mapping. 2437 * @see org.xml.sax.ContentHandler#endPrefixMapping 2438 * 2439 * @param prefix The prefix that was being mapping. 2440 * @throws org.xml.sax.SAXException The client may throw 2441 * an exception during processing. 2442 */ 2443 public void endPrefixMapping(String prefix) throws org.xml.sax.SAXException 2444 { // do nothing 2445 } 2446 2447 /** 2448 * Receive notification of ignorable whitespace in element content. 2449 * 2450 * Not sure how to get this invoked quite yet. 2451 * 2452 * @param ch The characters from the XML document. 2453 * @param start The start position in the array. 2454 * @param length The number of characters to read from the array. 2455 * @throws org.xml.sax.SAXException Any SAX exception, possibly 2456 * wrapping another exception. 2457 * @see #characters 2458 * 2459 * @throws org.xml.sax.SAXException 2460 */ 2461 public void ignorableWhitespace(char ch[], int start, int length) 2462 throws org.xml.sax.SAXException 2463 { 2464 2465 if (0 == length) 2466 return; 2467 characters(ch, start, length); 2468 } 2469 2470 /** 2471 * Receive notification of a skipped entity. 2472 * @see org.xml.sax.ContentHandler#skippedEntity 2473 * 2474 * @param name The name of the skipped entity. If it is a 2475 * parameter entity, the name will begin with '%', 2476 * and if it is the external DTD subset, it will be the string 2477 * "[dtd]". 2478 * @throws org.xml.sax.SAXException Any SAX exception, possibly wrapping 2479 * another exception. 2480 */ 2481 public void skippedEntity(String name) throws org.xml.sax.SAXException 2482 { // TODO: Should handle 2483 } 2484 2485 /** 2486 * Report the start of a CDATA section. 2487 * 2488 * @throws org.xml.sax.SAXException The application may raise an exception. 2489 * @see #endCDATA 2490 */ 2491 public void startCDATA() throws org.xml.sax.SAXException 2492 { 2493 if (m_doIndent) { 2494 m_childNodeNum++; 2495 flushCharactersBuffer(); 2496 } 2497 2498 m_cdataStartCalled = true; 2499 } 2500 2501 /** 2502 * Report the beginning of an entity. 2503 * 2504 * The start and end of the document entity are not reported. 2505 * The start and end of the external DTD subset are reported 2506 * using the pseudo-name "[dtd]". All other events must be 2507 * properly nested within start/end entity events. 2508 * 2509 * @param name The name of the entity. If it is a parameter 2510 * entity, the name will begin with '%'. 2511 * @throws org.xml.sax.SAXException The application may raise an exception. 2512 * @see #endEntity 2513 * @see org.xml.sax.ext.DeclHandler#internalEntityDecl 2514 * @see org.xml.sax.ext.DeclHandler#externalEntityDecl 2515 */ 2516 public void startEntity(String name) throws org.xml.sax.SAXException 2517 { 2518 if (name.equals("[dtd]")) 2519 m_inExternalDTD = true; 2520 2521 // if this is not the magic [dtd] name 2522 if (!m_inExternalDTD) { 2523 // if it's not in nested entity reference 2524 if (!isInEntityRef()) { 2525 if (shouldFormatOutput()) { 2526 m_charactersBuffer.addEntityReference(name); 2527 } else { 2528 outputEntityReference(name); 2529 } 2530 } 2531 m_inEntityRef++; 2532 } 2533 } 2534 2535 /** 2536 * Write out the entity reference with the form as "&entityName;". 2537 * 2538 * @param name The name of the entity. 2539 */ 2540 private void outputEntityReference(String name) throws SAXException { 2541 startNonEscaping(); 2542 characters("&" + name + ';'); 2543 endNonEscaping(); 2544 m_isprevtext = true; 2545 } 2546 2547 /** 2548 * For the enclosing elements starting tag write out 2549 * out any attributes followed by ">" 2550 * 2551 * @throws org.xml.sax.SAXException 2552 */ 2553 protected void closeStartTag() throws SAXException 2554 { 2555 if (m_elemContext.m_startTagOpen) 2556 { 2557 2558 try 2559 { 2560 if (m_tracer != null) 2561 super.fireStartElem(m_elemContext.m_elementName); 2562 int nAttrs = m_attributes.getLength(); 2563 if (nAttrs > 0) 2564 { 2565 processAttributes(m_writer, nAttrs); 2566 // clear attributes object for re-use with next element 2567 m_attributes.clear(); 2568 } 2569 m_writer.write('>'); 2570 } 2571 catch (IOException e) 2572 { 2573 throw new SAXException(e); 2574 } 2575 2576 /* whether Xalan or XSLTC, we have the prefix mappings now, so 2577 * lets determine if the current element is specified in the cdata- 2578 * section-elements list. 2579 */ 2580 if (m_StringOfCDATASections != null) 2581 m_elemContext.m_isCdataSection = isCdataSection(); 2582 } 2583 2584 } 2585 2586 /** 2587 * Report the start of DTD declarations, if any. 2588 * 2589 * Any declarations are assumed to be in the internal subset unless 2590 * otherwise indicated. 2591 * 2592 * @param name The document type name. 2593 * @param publicId The declared public identifier for the 2594 * external DTD subset, or null if none was declared. 2595 * @param systemId The declared system identifier for the 2596 * external DTD subset, or null if none was declared. 2597 * @throws org.xml.sax.SAXException The application may raise an 2598 * exception. 2599 * @see #endDTD 2600 * @see #startEntity 2601 */ 2602 public void startDTD(String name, String publicId, String systemId) 2603 throws org.xml.sax.SAXException 2604 { 2605 setDoctypeSystem(systemId); 2606 setDoctypePublic(publicId); 2607 2608 m_elemContext.m_elementName = name; 2609 m_inDoctype = true; 2610 } 2611 2612 /** 2613 * Returns the m_indentAmount. 2614 * @return int 2615 */ 2616 public int getIndentAmount() 2617 { 2618 return m_indentAmount; 2619 } 2620 2621 /** 2622 * Sets the m_indentAmount. 2623 * 2624 * @param m_indentAmount The m_indentAmount to set 2625 */ 2626 public void setIndentAmount(int m_indentAmount) 2627 { 2628 this.m_indentAmount = m_indentAmount; 2629 } 2630 2631 /** 2632 * Tell if, based on space preservation constraints and the doIndent property, 2633 * if an indent should occur. 2634 * 2635 * @return True if an indent should occur. 2636 */ 2637 protected boolean shouldIndent() 2638 { 2639 return shouldFormatOutput() && (m_elemContext.m_currentElemDepth > 0 || m_isStandalone); 2640 } 2641 2642 /** 2643 * Searches for the list of qname properties with the specified key in the 2644 * property list. If the key is not found in this property list, the default 2645 * property list, and its defaults, recursively, are then checked. The 2646 * method returns <code>null</code> if the property is not found. 2647 * 2648 * @param key the property key. 2649 * @param props the list of properties to search in. 2650 * 2651 * Sets the ArrayList of local-name/URI pairs of the cdata section elements 2652 * specified in the cdata-section-elements property. 2653 * 2654 * This method is essentially a copy of getQNameProperties() from 2655 * OutputProperties. Eventually this method should go away and a call 2656 * to setCdataSectionElements(List<String> v) should be made directly. 2657 */ 2658 private void setCdataSectionElements(String key, Properties props) { 2659 String s = props.getProperty(key); 2660 2661 if (null != s) { 2662 // List<String> of URI/LocalName pairs 2663 List<String> al = new ArrayList<>(); 2664 int l = s.length(); 2665 boolean inCurly = false; 2666 StringBuilder buf = new StringBuilder(); 2667 2668 // parse through string, breaking on whitespaces. I do this instead 2669 // of a tokenizer so I can track whitespace inside of curly brackets, 2670 // which theoretically shouldn't happen if they contain legal URLs. 2671 for (int i = 0; i < l; i++) 2672 { 2673 char c = s.charAt(i); 2674 2675 if (Character.isWhitespace(c)) 2676 { 2677 if (!inCurly) 2678 { 2679 if (buf.length() > 0) 2680 { 2681 addCdataSectionElement(buf.toString(), al); 2682 buf.setLength(0); 2683 } 2684 continue; 2685 } 2686 } 2687 else if ('{' == c) 2688 inCurly = true; 2689 else if ('}' == c) 2690 inCurly = false; 2691 2692 buf.append(c); 2693 } 2694 2695 if (buf.length() > 0) 2696 { 2697 addCdataSectionElement(buf.toString(), al); 2698 buf.setLength(0); 2699 } 2700 // call the official, public method to set the collected names 2701 setCdataSectionElements(al); 2702 } 2703 2704 } 2705 2706 /** 2707 * Adds a URI/LocalName pair of strings to the list. 2708 * 2709 * @param URI_and_localName String of the form "{uri}local" or "local" 2710 * 2711 * @return a QName object 2712 */ 2713 private void addCdataSectionElement(String URI_and_localName, List<String> al) { 2714 StringTokenizer tokenizer = new StringTokenizer(URI_and_localName, "{}", false); 2715 String s1 = tokenizer.nextToken(); 2716 String s2 = tokenizer.hasMoreTokens() ? tokenizer.nextToken() : null; 2717 2718 if (null == s2) { 2719 // add null URI and the local name 2720 al.add(null); 2721 al.add(s1); 2722 } else { 2723 // add URI, then local name 2724 al.add(s1); 2725 al.add(s2); 2726 } 2727 } 2728 2729 /** 2730 * Remembers the cdata sections specified in the cdata-section-elements. 2731 * The "official way to set URI and localName pairs. 2732 * This method should be used by both Xalan and XSLTC. 2733 * 2734 * @param URI_and_localNames an ArrayList of pairs of Strings (URI/local) 2735 */ 2736 public void setCdataSectionElements(List<String> URI_and_localNames) { 2737 // convert to the new way. 2738 if (URI_and_localNames != null) { 2739 final int len = URI_and_localNames.size() - 1; 2740 if (len > 0) { 2741 final StringBuilder sb = new StringBuilder(); 2742 for (int i = 0; i < len; i += 2) { 2743 // whitspace separated "{uri1}local1 {uri2}local2 ..." 2744 if (i != 0) 2745 sb.append(' '); 2746 final String uri = URI_and_localNames.get(i); 2747 final String localName = URI_and_localNames.get(i + 1); 2748 if (uri != null) { 2749 // If there is no URI don't put this in, just the localName then. 2750 sb.append('{'); 2751 sb.append(uri); 2752 sb.append('}'); 2753 } 2754 sb.append(localName); 2755 } 2756 m_StringOfCDATASections = sb.toString(); 2757 } 2758 } 2759 initCdataElems(m_StringOfCDATASections); 2760 } 2761 2762 /** 2763 * Makes sure that the namespace URI for the given qualified attribute name 2764 * is declared. 2765 * @param ns the namespace URI 2766 * @param rawName the qualified name 2767 * @return returns null if no action is taken, otherwise it returns the 2768 * prefix used in declaring the namespace. 2769 * @throws SAXException 2770 */ 2771 protected String ensureAttributesNamespaceIsDeclared( 2772 String ns, 2773 String localName, 2774 String rawName) 2775 throws org.xml.sax.SAXException 2776 { 2777 2778 if (ns != null && ns.length() > 0) 2779 { 2780 2781 // extract the prefix in front of the raw name 2782 int index = 0; 2783 String prefixFromRawName = 2784 (index = rawName.indexOf(":")) < 0 2785 ? "" 2786 : rawName.substring(0, index); 2787 2788 if (index > 0) 2789 { 2790 // we have a prefix, lets see if it maps to a namespace 2791 String uri = m_prefixMap.lookupNamespace(prefixFromRawName); 2792 if (uri != null && uri.equals(ns)) 2793 { 2794 // the prefix in the raw name is already maps to the given namespace uri 2795 // so we don't need to do anything 2796 return null; 2797 } 2798 else 2799 { 2800 // The uri does not map to the prefix in the raw name, 2801 // so lets make the mapping. 2802 this.startPrefixMapping(prefixFromRawName, ns, false); 2803 this.addAttribute( 2804 "http://www.w3.org/2000/xmlns/", 2805 prefixFromRawName, 2806 "xmlns:" + prefixFromRawName, 2807 "CDATA", 2808 ns, false); 2809 return prefixFromRawName; 2810 } 2811 } 2812 else 2813 { 2814 // we don't have a prefix in the raw name. 2815 // Does the URI map to a prefix already? 2816 String prefix = m_prefixMap.lookupPrefix(ns); 2817 if (prefix == null) 2818 { 2819 // uri is not associated with a prefix, 2820 // so lets generate a new prefix to use 2821 prefix = m_prefixMap.generateNextPrefix(); 2822 this.startPrefixMapping(prefix, ns, false); 2823 this.addAttribute( 2824 "http://www.w3.org/2000/xmlns/", 2825 prefix, 2826 "xmlns:" + prefix, 2827 "CDATA", 2828 ns, false); 2829 } 2830 2831 return prefix; 2832 2833 } 2834 } 2835 return null; 2836 } 2837 2838 void ensurePrefixIsDeclared(String ns, String rawName) 2839 throws org.xml.sax.SAXException 2840 { 2841 2842 if (ns != null && ns.length() > 0) 2843 { 2844 int index; 2845 final boolean no_prefix = ((index = rawName.indexOf(":")) < 0); 2846 String prefix = (no_prefix) ? "" : rawName.substring(0, index); 2847 2848 if (null != prefix) 2849 { 2850 String foundURI = m_prefixMap.lookupNamespace(prefix); 2851 2852 if ((null == foundURI) || !foundURI.equals(ns)) 2853 { 2854 this.startPrefixMapping(prefix, ns); 2855 2856 // Bugzilla1133: Generate attribute as well as namespace event. 2857 // SAX does expect both. 2858 2859 this.addAttributeAlways( 2860 "http://www.w3.org/2000/xmlns/", 2861 no_prefix ? "xmlns" : prefix, // local name 2862 no_prefix ? "xmlns" : ("xmlns:"+ prefix), // qname 2863 "CDATA", 2864 ns, 2865 false); 2866 } 2867 2868 } 2869 } 2870 } 2871 2872 /** 2873 * This method flushes any pending events, which can be startDocument() 2874 * closing the opening tag of an element, or closing an open CDATA section. 2875 */ 2876 public void flushPending() throws SAXException 2877 { 2878 if (m_needToCallStartDocument) 2879 { 2880 startDocumentInternal(); 2881 m_needToCallStartDocument = false; 2882 } 2883 if (m_elemContext.m_startTagOpen) 2884 { 2885 closeStartTag(); 2886 m_elemContext.m_startTagOpen = false; 2887 } 2888 2889 if (m_cdataTagOpen) 2890 { 2891 closeCDATA(); 2892 m_cdataTagOpen = false; 2893 } 2894 } 2895 2896 public void setContentHandler(ContentHandler ch) 2897 { 2898 // this method is really only useful in the ToSAXHandler classes but it is 2899 // in the interface. If the method defined here is ever called 2900 // we are probably in trouble. 2901 } 2902 2903 /** 2904 * Adds the given attribute to the set of attributes, even if there is 2905 * no currently open element. This is useful if a SAX startPrefixMapping() 2906 * should need to add an attribute before the element name is seen. 2907 * 2908 * This method is a copy of its super classes method, except that some 2909 * tracing of events is done. This is so the tracing is only done for 2910 * stream serializers, not for SAX ones. 2911 * 2912 * @param uri the URI of the attribute 2913 * @param localName the local name of the attribute 2914 * @param rawName the qualified name of the attribute 2915 * @param type the type of the attribute (probably CDATA) 2916 * @param value the value of the attribute 2917 * @param xslAttribute true if this attribute is coming from an xsl:attribute element. 2918 * @return true if the attribute value was added, 2919 * false if the attribute already existed and the value was 2920 * replaced with the new value. 2921 */ 2922 public boolean addAttributeAlways( 2923 String uri, 2924 String localName, 2925 String rawName, 2926 String type, 2927 String value, 2928 boolean xslAttribute) 2929 { 2930 if (!m_charactersBuffer.isAnyCharactersBuffered()) { 2931 return doAddAttributeAlways(uri, localName, rawName, type, value, xslAttribute); 2932 } else { 2933 /* 2934 * If stylesheet includes xsl:copy-of an attribute node, XSLTC will 2935 * fire an addAttribute event. When a text node is handling in 2936 * ToStream, addAttribute has no effect. But closeStartTag call is 2937 * delayed to flushCharactersBuffer() method if the text node is 2938 * buffered, so here we ignore the attribute to avoid corrupting the 2939 * start tag content. 2940 * 2941 */ 2942 return m_attributes.getIndex(rawName) < 0; 2943 } 2944 } 2945 2946 /** 2947 * Does really add the attribute to the set of attributes. 2948 */ 2949 private boolean doAddAttributeAlways( 2950 String uri, 2951 String localName, 2952 String rawName, 2953 String type, 2954 String value, 2955 boolean xslAttribute) 2956 { 2957 boolean was_added; 2958 int index; 2959 //if (uri == null || localName == null || uri.length() == 0) 2960 index = m_attributes.getIndex(rawName); 2961 // Don't use 'localName' as it gives incorrect value, rely only on 'rawName' 2962 /*else { 2963 index = m_attributes.getIndex(uri, localName); 2964 }*/ 2965 if (index >= 0) 2966 { 2967 String old_value = null; 2968 if (m_tracer != null) 2969 { 2970 old_value = m_attributes.getValue(index); 2971 if (value.equals(old_value)) 2972 old_value = null; 2973 } 2974 2975 /* We've seen the attribute before. 2976 * We may have a null uri or localName, but all we really 2977 * want to re-set is the value anyway. 2978 */ 2979 m_attributes.setValue(index, value); 2980 was_added = false; 2981 if (old_value != null){ 2982 firePseudoAttributes(); 2983 } 2984 2985 } 2986 else 2987 { 2988 // the attribute doesn't exist yet, create it 2989 if (xslAttribute) 2990 { 2991 /* 2992 * This attribute is from an xsl:attribute element so we take some care in 2993 * adding it, e.g. 2994 * <elem1 foo:attr1="1" xmlns:foo="uri1"> 2995 * <xsl:attribute name="foo:attr2">2</xsl:attribute> 2996 * </elem1> 2997 * 2998 * We are adding attr1 and attr2 both as attributes of elem1, 2999 * and this code is adding attr2 (the xsl:attribute ). 3000 * We could have a collision with the prefix like in the example above. 3001 */ 3002 3003 // In the example above, is there a prefix like foo ? 3004 final int colonIndex = rawName.indexOf(':'); 3005 if (colonIndex > 0) 3006 { 3007 String prefix = rawName.substring(0,colonIndex); 3008 NamespaceMappings.MappingRecord existing_mapping = m_prefixMap.getMappingFromPrefix(prefix); 3009 3010 /* Before adding this attribute (foo:attr2), 3011 * is the prefix for it (foo) already mapped at the current depth? 3012 */ 3013 if (existing_mapping != null 3014 && existing_mapping.m_declarationDepth == m_elemContext.m_currentElemDepth 3015 && !existing_mapping.m_uri.equals(uri)) 3016 { 3017 /* 3018 * There is an existing mapping of this prefix, 3019 * it differs from the one we need, 3020 * and unfortunately it is at the current depth so we 3021 * can not over-ride it. 3022 */ 3023 3024 /* 3025 * Are we lucky enough that an existing other prefix maps to this URI ? 3026 */ 3027 prefix = m_prefixMap.lookupPrefix(uri); 3028 if (prefix == null) 3029 { 3030 /* Unfortunately there is no existing prefix that happens to map to ours, 3031 * so to avoid a prefix collision we must generated a new prefix to use. 3032 * This is OK because the prefix URI mapping 3033 * defined in the xsl:attribute is short in scope, 3034 * just the xsl:attribute element itself, 3035 * and at this point in serialization the body of the 3036 * xsl:attribute, if any, is just a String. Right? 3037 * . . . I sure hope so - Brian M. 3038 */ 3039 prefix = m_prefixMap.generateNextPrefix(); 3040 } 3041 3042 rawName = prefix + ':' + localName; 3043 } 3044 } 3045 3046 try 3047 { 3048 /* This is our last chance to make sure the namespace for this 3049 * attribute is declared, especially if we just generated an alternate 3050 * prefix to avoid a collision (the new prefix/rawName will go out of scope 3051 * soon and be lost ... last chance here. 3052 */ 3053 String prefixUsed = 3054 ensureAttributesNamespaceIsDeclared( 3055 uri, 3056 localName, 3057 rawName); 3058 } 3059 catch (SAXException e) 3060 { 3061 // TODO Auto-generated catch block 3062 e.printStackTrace(); 3063 } 3064 } 3065 3066 m_attributes.addAttribute(uri, localName, rawName, type, value); 3067 was_added = true; 3068 if (m_tracer != null){ 3069 firePseudoAttributes(); 3070 } 3071 } 3072 3073 if (m_doIndent && rawName.equals("xml:space")) { 3074 if (value.equals("preserve")) { 3075 m_ispreserveSpace = true; 3076 if (m_preserveSpaces.size() > 0) 3077 m_preserveSpaces.setTop(m_ispreserveSpace); 3078 } else if (value.equals("default")) { 3079 m_ispreserveSpace = false; 3080 if (m_preserveSpaces.size() > 0) 3081 m_preserveSpaces.setTop(m_ispreserveSpace); 3082 } 3083 } 3084 3085 return was_added; 3086 } 3087 3088 /** 3089 * To fire off the pseudo characters of attributes, as they currently 3090 * exist. This method should be called everytime an attribute is added, 3091 * or when an attribute value is changed, or an element is created. 3092 */ 3093 protected void firePseudoAttributes() { 3094 if (m_tracer != null) { 3095 try { 3096 // flush out the "<elemName" if not already flushed 3097 m_writer.flush(); 3098 3099 // make a StringBuffer to write the name="value" pairs to. 3100 StringBuffer sb = new StringBuffer(); 3101 int nAttrs = m_attributes.getLength(); 3102 if (nAttrs > 0) { 3103 // make a writer that internally appends to the same 3104 // StringBuffer 3105 Writer writer = new ToStream.WritertoStringBuffer(sb); 3106 3107 processAttributes(writer, nAttrs); 3108 // Don't clear the attributes! 3109 // We only want to see what would be written out 3110 // at this point, we don't want to loose them. 3111 } 3112 sb.append('>'); // the potential > after the attributes. 3113 // convert the StringBuffer to a char array and 3114 // emit the trace event that these characters "might" 3115 // be written 3116 char ch[] = sb.toString().toCharArray(); 3117 m_tracer.fireGenerateEvent( 3118 SerializerTrace.EVENTTYPE_OUTPUT_PSEUDO_CHARACTERS, 3119 ch, 3120 0, 3121 ch.length); 3122 } catch (IOException ioe) { 3123 // ignore ? 3124 } catch (SAXException se) { 3125 // ignore ? 3126 } 3127 } 3128 } 3129 3130 /** 3131 * This inner class is used only to collect attribute values 3132 * written by the method writeAttrString() into a string buffer. 3133 * In this manner trace events, and the real writing of attributes will use 3134 * the same code. 3135 */ 3136 private class WritertoStringBuffer extends Writer { 3137 final private StringBuffer m_stringbuf; 3138 3139 /** 3140 * @see java.io.Writer#write(char[], int, int) 3141 */ 3142 WritertoStringBuffer(StringBuffer sb) { 3143 m_stringbuf = sb; 3144 } 3145 3146 public void write(char[] arg0, int arg1, int arg2) throws IOException { 3147 m_stringbuf.append(arg0, arg1, arg2); 3148 } 3149 3150 /** 3151 * @see java.io.Writer#flush() 3152 */ 3153 public void flush() throws IOException {} 3154 3155 /** 3156 * @see java.io.Writer#close() 3157 */ 3158 public void close() throws IOException {} 3159 3160 public void write(int i) { 3161 m_stringbuf.append((char) i); 3162 } 3163 3164 public void write(String s) { 3165 m_stringbuf.append(s); 3166 } 3167 } 3168 3169 /** 3170 * @see SerializationHandler#setTransformer(Transformer) 3171 */ 3172 public void setTransformer(Transformer transformer) { 3173 super.setTransformer(transformer); 3174 if (m_tracer != null && !(m_writer instanceof SerializerTraceWriter)) { 3175 m_writer = new SerializerTraceWriter(m_writer, m_tracer); 3176 } 3177 } 3178 3179 /** 3180 * Try's to reset the super class and reset this class for 3181 * re-use, so that you don't need to create a new serializer 3182 * (mostly for performance reasons). 3183 * 3184 * @return true if the class was successfuly reset. 3185 */ 3186 public boolean reset() { 3187 boolean wasReset = false; 3188 if (super.reset()) { 3189 resetToStream(); 3190 wasReset = true; 3191 } 3192 return wasReset; 3193 } 3194 3195 /** 3196 * Reset all of the fields owned by ToStream class 3197 * 3198 */ 3199 private void resetToStream() { 3200 this.m_cdataStartCalled = false; 3201 /* The stream is being reset. It is one of 3202 * ToXMLStream, ToHTMLStream ... and this type can't be changed 3203 * so neither should m_charInfo which is associated with the 3204 * type of Stream. Just leave m_charInfo as-is for the next re-use. 3205 */ 3206 // this.m_charInfo = null; // don't set to null 3207 3208 this.m_disableOutputEscapingStates.clear(); 3209 3210 this.m_escaping = true; 3211 // Leave m_format alone for now - Brian M. 3212 // this.m_format = null; 3213 this.m_inDoctype = false; 3214 this.m_ispreserveSpace = false; 3215 this.m_preserveSpaces.clear(); 3216 this.m_childNodeNum = 0; 3217 this.m_childNodeNumStack.clear(); 3218 this.m_charactersBuffer.clear(); 3219 this.m_isprevtext = false; 3220 this.m_isUTF8 = false; // ?? used anywhere ?? 3221 this.m_shouldFlush = true; 3222 this.m_spaceBeforeClose = false; 3223 this.m_startNewLine = false; 3224 this.m_lineSepUse = true; 3225 // DON'T SET THE WRITER TO NULL, IT MAY BE REUSED !! 3226 // this.m_writer = null; 3227 this.m_expandDTDEntities = true; 3228 3229 } 3230 3231 /** 3232 * Sets the character encoding coming from the xsl:output encoding stylesheet attribute. 3233 * @param encoding the character encoding 3234 */ 3235 public void setEncoding(String encoding) 3236 { 3237 setOutputProperty(OutputKeys.ENCODING,encoding); 3238 } 3239 3240 /** 3241 * Simple stack for boolean values. 3242 * 3243 * This class is a copy of the one in com.sun.org.apache.xml.internal.utils. 3244 * It exists to cut the serializers dependancy on that package. 3245 * A minor changes from that package are: 3246 * doesn't implement Clonable 3247 * 3248 * @xsl.usage internal 3249 */ 3250 static final class BoolStack { 3251 /** Array of boolean values */ 3252 private boolean m_values[]; 3253 3254 /** Array size allocated */ 3255 private int m_allocatedSize; 3256 3257 /** Index into the array of booleans */ 3258 private int m_index; 3259 3260 /** 3261 * Default constructor. Note that the default 3262 * block size is very small, for small lists. 3263 */ 3264 public BoolStack() { 3265 this(32); 3266 } 3267 3268 /** 3269 * Construct a IntVector, using the given block size. 3270 * 3271 * @param size array size to allocate 3272 */ 3273 public BoolStack(int size) { 3274 m_allocatedSize = size; 3275 m_values = new boolean[size]; 3276 m_index = -1; 3277 } 3278 3279 /** 3280 * Get the length of the list. 3281 * 3282 * @return Current length of the list 3283 */ 3284 public final int size() { 3285 return m_index + 1; 3286 } 3287 3288 /** 3289 * Clears the stack. 3290 * 3291 */ 3292 public final void clear() { 3293 m_index = -1; 3294 } 3295 3296 /** 3297 * Pushes an item onto the top of this stack. 3298 * 3299 * 3300 * @param val the boolean to be pushed onto this stack. 3301 * @return the <code>item</code> argument. 3302 */ 3303 public final boolean push(boolean val) { 3304 if (m_index == m_allocatedSize - 1) 3305 grow(); 3306 3307 return (m_values[++m_index] = val); 3308 } 3309 3310 /** 3311 * Removes the object at the top of this stack and returns that 3312 * object as the value of this function. 3313 * 3314 * @return The object at the top of this stack. 3315 * @throws EmptyStackException if this stack is empty. 3316 */ 3317 public final boolean pop() { 3318 return m_values[m_index--]; 3319 } 3320 3321 /** 3322 * Removes the object at the top of this stack and returns the 3323 * next object at the top as the value of this function. 3324 * 3325 * 3326 * @return Next object to the top or false if none there 3327 */ 3328 public final boolean popAndTop() { 3329 m_index--; 3330 return (m_index >= 0) ? m_values[m_index] : false; 3331 } 3332 3333 /** 3334 * Set the item at the top of this stack 3335 * 3336 * 3337 * @param b Object to set at the top of this stack 3338 */ 3339 public final void setTop(boolean b) { 3340 m_values[m_index] = b; 3341 } 3342 3343 /** 3344 * Looks at the object at the top of this stack without removing it 3345 * from the stack. 3346 * 3347 * @return the object at the top of this stack. 3348 * @throws EmptyStackException if this stack is empty. 3349 */ 3350 public final boolean peek() { 3351 return m_values[m_index]; 3352 } 3353 3354 /** 3355 * Looks at the object at the top of this stack without removing it 3356 * from the stack. If the stack is empty, it returns false. 3357 * 3358 * @return the object at the top of this stack. 3359 */ 3360 public final boolean peekOrFalse() { 3361 return (m_index > -1) ? m_values[m_index] : false; 3362 } 3363 3364 /** 3365 * Looks at the object at the top of this stack without removing it 3366 * from the stack. If the stack is empty, it returns true. 3367 * 3368 * @return the object at the top of this stack. 3369 */ 3370 public final boolean peekOrTrue() { 3371 return (m_index > -1) ? m_values[m_index] : true; 3372 } 3373 3374 /** 3375 * Tests if this stack is empty. 3376 * 3377 * @return <code>true</code> if this stack is empty; 3378 * <code>false</code> otherwise. 3379 */ 3380 public boolean isEmpty() { 3381 return (m_index == -1); 3382 } 3383 3384 /** 3385 * Grows the size of the stack 3386 * 3387 */ 3388 private void grow() { 3389 m_allocatedSize *= 2; 3390 boolean newVector[] = new boolean[m_allocatedSize]; 3391 System.arraycopy(m_values, 0, newVector, 0, m_index + 1); 3392 m_values = newVector; 3393 } 3394 } 3395 3396 3397 /** 3398 * This inner class is used to buffer the text nodes and the entity 3399 * reference nodes if indentation is on. There is only one CharacterBuffer 3400 * instance in ToStream, it contains a queue of GenericCharacters, 3401 * GenericCharacters can be a text node or an entity reference node. The 3402 * text nodes and entity reference nodes are joined together and then are 3403 * flushed. 3404 */ 3405 private class CharacterBuffer { 3406 /** 3407 * GenericCharacters is immutable. 3408 */ 3409 private abstract class GenericCharacters { 3410 /** 3411 * @return True if all characters in this Text are newlines. 3412 */ 3413 abstract boolean flush(boolean skipBeginningNewlines) throws SAXException; 3414 3415 /** 3416 * Converts this GenericCharacters to a new character array. This 3417 * method is used to handle cdata-section-elements attribute in 3418 * xsl:output. Therefore it doesn't need to consider 3419 * skipBeginningNewlines because the text will be involved with CDATA 3420 * tag. 3421 */ 3422 abstract char[] toChars(); 3423 } 3424 3425 private List<GenericCharacters> bufferedCharacters = new ArrayList<>(); 3426 3427 /** 3428 * Append a text node to the buffer. 3429 */ 3430 public void addText(final char chars[], final int start, final int length) { 3431 bufferedCharacters.add(new GenericCharacters() { 3432 char[] text; 3433 3434 { 3435 text = Arrays.copyOfRange(chars, start, start + length); 3436 } 3437 3438 boolean flush(boolean skipBeginningNewlines) throws SAXException { 3439 int start = 0; 3440 while (skipBeginningNewlines && text[start] == '\n') { 3441 start++; 3442 if (start == text.length) { 3443 return true; 3444 } 3445 } 3446 outputCharacters(text, start, text.length - start); 3447 return false; 3448 } 3449 3450 char[] toChars() { 3451 return text; 3452 } 3453 }); 3454 } 3455 3456 /** 3457 * Append an entity reference to the buffer. 3458 */ 3459 public void addEntityReference(String entityName) { 3460 bufferedCharacters.add(new GenericCharacters() { 3461 boolean flush(boolean skipBeginningNewlines) throws SAXException { 3462 if (m_elemContext.m_startTagOpen) 3463 { 3464 closeStartTag(); 3465 m_elemContext.m_startTagOpen = false; 3466 } 3467 if (m_cdataTagOpen) 3468 closeCDATA(); 3469 char[] cs = toChars(); 3470 try { 3471 m_writer.write(cs, 0, cs.length); 3472 m_isprevtext = true; 3473 } catch (IOException e) { 3474 throw new SAXException(e); 3475 } 3476 return false; 3477 } 3478 3479 char[] toChars() { 3480 return ("&" + entityName + ";").toCharArray(); 3481 } 3482 }); 3483 } 3484 3485 /** 3486 * Append a raw text to the buffer. Used to handle raw characters event. 3487 */ 3488 public void addRawText(final char chars[], final int start, final int length) { 3489 bufferedCharacters.add(new GenericCharacters() { 3490 char[] text; 3491 3492 { 3493 text = Arrays.copyOfRange(chars, start, start + length); 3494 } 3495 3496 boolean flush(boolean skipBeginningNewlines) throws SAXException { 3497 try { 3498 int start = 0; 3499 while (skipBeginningNewlines && text[start] == '\n') { 3500 start++; 3501 if (start == text.length) { 3502 return true; 3503 } 3504 } 3505 m_writer.write(text, start, text.length - start); 3506 m_isprevtext = true; 3507 } catch (IOException e) { 3508 throw new SAXException(e); 3509 } 3510 return false; 3511 } 3512 3513 char[] toChars() { 3514 return text; 3515 } 3516 }); 3517 } 3518 3519 /** 3520 * @return True if any GenericCharacters are buffered. 3521 */ 3522 public boolean isAnyCharactersBuffered() { 3523 return bufferedCharacters.size() > 0; 3524 } 3525 3526 /** 3527 * Flush all buffered GenericCharacters. 3528 */ 3529 public void flush(boolean skipBeginningNewlines) throws SAXException { 3530 Iterator<GenericCharacters> itr = bufferedCharacters.iterator(); 3531 3532 boolean continueSkipBeginningNewlines = skipBeginningNewlines; 3533 while (itr.hasNext()) { 3534 GenericCharacters element = itr.next(); 3535 continueSkipBeginningNewlines = element.flush(continueSkipBeginningNewlines); 3536 itr.remove(); 3537 } 3538 } 3539 3540 /** 3541 * Converts all buffered GenericCharacters to a new character array. 3542 */ 3543 public char[] toChars() { 3544 StringBuilder sb = new StringBuilder(); 3545 for (GenericCharacters element : bufferedCharacters) { 3546 sb.append(element.toChars()); 3547 } 3548 return sb.toString().toCharArray(); 3549 } 3550 3551 /** 3552 * Clear the buffer. 3553 */ 3554 public void clear() { 3555 bufferedCharacters.clear(); 3556 } 3557 } 3558 3559 3560 // Implement DTDHandler 3561 /** 3562 * If this method is called, the serializer is used as a 3563 * DTDHandler, which changes behavior how the serializer 3564 * handles document entities. 3565 * @see org.xml.sax.DTDHandler#notationDecl(java.lang.String, java.lang.String, java.lang.String) 3566 */ 3567 public void notationDecl(String name, String pubID, String sysID) throws SAXException { 3568 // TODO Auto-generated method stub 3569 try { 3570 DTDprolog(); 3571 3572 m_writer.write("<!NOTATION "); 3573 m_writer.write(name); 3574 if (pubID != null) { 3575 m_writer.write(" PUBLIC \""); 3576 m_writer.write(pubID); 3577 3578 } 3579 else { 3580 m_writer.write(" SYSTEM \""); 3581 m_writer.write(sysID); 3582 } 3583 m_writer.write("\" >"); 3584 m_writer.write(m_lineSep, 0, m_lineSepLen); 3585 } catch (IOException e) { 3586 // TODO Auto-generated catch block 3587 e.printStackTrace(); 3588 } 3589 } 3590 3591 /** 3592 * If this method is called, the serializer is used as a 3593 * DTDHandler, which changes behavior how the serializer 3594 * handles document entities. 3595 * @see org.xml.sax.DTDHandler#unparsedEntityDecl(java.lang.String, java.lang.String, java.lang.String, java.lang.String) 3596 */ 3597 public void unparsedEntityDecl(String name, String pubID, String sysID, String notationName) throws SAXException { 3598 // TODO Auto-generated method stub 3599 try { 3600 DTDprolog(); 3601 3602 m_writer.write("<!ENTITY "); 3603 m_writer.write(name); 3604 if (pubID != null) { 3605 m_writer.write(" PUBLIC \""); 3606 m_writer.write(pubID); 3607 3608 } 3609 else { 3610 m_writer.write(" SYSTEM \""); 3611 m_writer.write(sysID); 3612 } 3613 m_writer.write("\" NDATA "); 3614 m_writer.write(notationName); 3615 m_writer.write(" >"); 3616 m_writer.write(m_lineSep, 0, m_lineSepLen); 3617 } catch (IOException e) { 3618 // TODO Auto-generated catch block 3619 e.printStackTrace(); 3620 } 3621 } 3622 3623 /** 3624 * A private helper method to output the 3625 * @throws SAXException 3626 * @throws IOException 3627 */ 3628 private void DTDprolog() throws SAXException, IOException { 3629 final Writer writer = m_writer; 3630 if (m_needToOutputDocTypeDecl) { 3631 outputDocTypeDecl(m_elemContext.m_elementName, false); 3632 m_needToOutputDocTypeDecl = false; 3633 } 3634 if (m_inDoctype) { 3635 writer.write(" ["); 3636 writer.write(m_lineSep, 0, m_lineSepLen); 3637 m_inDoctype = false; 3638 } 3639 } 3640 3641 /** 3642 * If set to false the serializer does not expand DTD entities, 3643 * but leaves them as is, the default value is true; 3644 */ 3645 public void setDTDEntityExpansion(boolean expand) { 3646 m_expandDTDEntities = expand; 3647 } 3648 3649 /** 3650 * Remembers the cdata sections specified in the cdata-section-elements by appending the given 3651 * cdata section elements to the list. This method can be called multiple times, but once an 3652 * element is put in the list of cdata section elements it can not be removed. 3653 * This method should be used by both Xalan and XSLTC. 3654 * 3655 * @param URI_and_localNames a whitespace separated list of element names, each element 3656 * is a URI in curly braces (optional) and a local name. An example of such a parameter is: 3657 * "{http://company.com}price {myURI2}book chapter" 3658 */ 3659 public void addCdataSectionElements(String URI_and_localNames) 3660 { 3661 if (URI_and_localNames != null) 3662 initCdataElems(URI_and_localNames); 3663 if (m_StringOfCDATASections == null) 3664 m_StringOfCDATASections = URI_and_localNames; 3665 else 3666 m_StringOfCDATASections += (" " + URI_and_localNames); 3667 } 3668 }