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