/*
    * Copyright 2001-2005 (C) MetaStuff, Ltd. All Rights Reserved.
    *
    * This software is open source.
    * See the bottom of this file for the licence.
    */
   
   package org.dom4j.io;
   
  import java.io.IOException;
  import java.io.OutputStream;
  import java.io.StringWriter;
  import java.io.UnsupportedEncodingException;
  import java.io.Writer;
  import java.util.HashSet;
  import java.util.Iterator;
  import java.util.Set;
  import java.util.Stack;
  
  import org.dom4j.Document;
  import org.dom4j.DocumentHelper;
  import org.dom4j.Element;
  import org.dom4j.Entity;
  import org.dom4j.Node;
  
  import org.xml.sax.SAXException;
  
  /***
   * <p>
   * <code>HTMLWriter</code> takes a DOM4J tree and formats it to a stream as
   * HTML. This formatter is similar to XMLWriter but it outputs the text of CDATA
   * and Entity sections rather than the serialised format as in XML, it has an
   * XHTML mode, it retains whitespace in certain elements such as &lt;PRE&gt;,
   * and it supports certain elements which have no corresponding close tag such
   * as for &lt;BR&gt; and &lt;P&gt;.
   * </p>
   * 
   * <p>
   * The OutputFormat passed in to the constructor is checked for isXHTML() and
   * isExpandEmptyElements(). See {@link OutputFormat OutputFormat}for details.
   * Here are the rules for <b>this class </b> based on an OutputFormat, "format",
   * passed in to the constructor: <br/><br/>
   * 
   * <ul>
   * <li>If an element is in {@link #getOmitElementCloseSet()
   * getOmitElementCloseSet}, then it is treated specially:
   * 
   * <ul>
   * <li>It never expands, since some browsers treat this as two separate
   * Horizontal Rules: &lt;HR&gt;&lt;/HR&gt;</li>
   * <li>If {@link org.dom4j.io.OutputFormat#isXHTML() format.isXHTML()}, then
   * it has a space before the closing single-tag slash, since Netscape 4.x-
   * treats this: &lt;HR /&gt; as an element named "HR" with an attribute named
   * "/", but that's better than when it refuses to recognize this: &lt;hr/&gt;
   * which it thinks is an element named "HR/".</li>
   * </ul>
   * 
   * </li>
   * <li>If {@link org.dom4j.io.OutputFormat#isXHTML() format.isXHTML()}, all
   * elements must have either a close element, or be a closed single tag.</li>
   * <li>If {@link org.dom4j.io.OutputFormat#isExpandEmptyElements()
   * format.isExpandEmptyElements()}() is true, all elements are expanded except
   * as above.</li>
   * </ul>
   * 
   * <b>Examples </b>
   * </p>
   * 
   * <p>
   * </p>
   * 
   * <p>
   * If isXHTML == true, CDATA sections look like this:
   * 
   * <PRE>
   * 
   * <b>&lt;myelement&gt;&lt;![CDATA[My data]]&gt;&lt;/myelement&gt; </b>
   * 
   * </PRE>
   * 
   * Otherwise, they look like this:
   * 
   * <PRE>
   * 
   * <b>&lt;myelement&gt;My data&lt;/myelement&gt; </b>
   * 
   * </PRE>
   * 
   * </p>
   * 
   * <p>
   * Basically, {@link OutputFormat.isXHTML() OutputFormat.isXHTML()} ==
   * <code>true</code> will produce valid XML, while {@link
   * org.dom4j.io.OutputFormat#isExpandEmptyElements()
   * format.isExpandEmptyElements()} determines whether empty elements are
   * expanded if isXHTML is true, excepting the special HTML single tags.
   * </p>
   * 
   * <p>
  * Also, HTMLWriter handles tags whose contents should be preformatted, that is,
  * whitespace-preserved. By default, this set includes the tags &lt;PRE&gt;,
  * &lt;SCRIPT&gt;, &lt;STYLE&gt;, and &lt;TEXTAREA&gt;, case insensitively. It
  * does not include &lt;IFRAME&gt;. Other tags, such as &lt;CODE&gt;,
  * &lt;KBD&gt;, &lt;TT&gt;, &lt;VAR&gt;, are usually rendered in a different
  * font in most browsers, but don't preserve whitespace, so they also don't
  * appear in the default list. HTML Comments are always whitespace-preserved.
  * However, the parser you use may store comments with linefeed-only text nodes
  * (\n) even if your platform uses another line.separator character, and
  * HTMLWriter outputs Comment nodes exactly as the DOM is set up by the parser.
  * See examples and discussion here: {@link#setPreformattedTags(java.util.Set)
  * setPreformattedTags}
  * </p>
  * 
  * <p>
  * <b>Examples </b>
  * </p>
  * <blockquote>
  * <p>
  * <b>Pretty Printing </b>
  * </p>
  * 
  * <p>
  * This example shows how to pretty print a string containing a valid HTML
  * document to a string. You can also just call the static methods of this
  * class: <br>
  * {@link #prettyPrintHTML(String) prettyPrintHTML(String)}or <br>
  * {@link #prettyPrintHTML(String,boolean,boolean,boolean,boolean)
  * prettyPrintHTML(String,boolean,boolean,boolean,boolean)} or, <br>
  * {@link #prettyPrintXHTML(String) prettyPrintXHTML(String)}for XHTML (note
  * the X)
  * </p>
  * 
  * <pre>
  * String testPrettyPrint(String html) {
  *     StringWriter sw = new StringWriter();
  *     OutputFormat format = OutputFormat.createPrettyPrint();
  *     // These are the default values for createPrettyPrint,
  *     // so you needn't set them:
  *     // format.setNewlines(true);
  *     // format.setTrimText(true);&lt;/font&gt;
  *     format.setXHTML(true);
  *     HTMLWriter writer = new HTMLWriter(sw, format);
  *     Document document = DocumentHelper.parseText(html);
  *     writer.write(document);
  *     writer.flush();
  *     return sw.toString();
  * }
  * </pre>
  * 
  * <p>
  * This example shows how to create a "squeezed" document, but one that will
  * work in browsers even if the browser line length is limited. No newlines are
  * included, no extra whitespace at all, except where it it required by
  * {@link #setPreformattedTags(java.util.Set) setPreformattedTags}.
  * </p>
  * 
  * <pre>
  * String testCrunch(String html) {
  *     StringWriter sw = new StringWriter();
  *     OutputFormat format = OutputFormat.createPrettyPrint();
  *     format.setNewlines(false);
  *     format.setTrimText(true);
  *     format.setIndent(&quot;&quot;);
  *     format.setXHTML(true);
  *     format.setExpandEmptyElements(false);
  *     format.setNewLineAfterNTags(20);
  *     org.dom4j.io.HTMLWriter writer = new HTMLWriter(sw, format);
  *     org.dom4j.Document document = DocumentHelper.parseText(html);
  *     writer.write(document);
  *     writer.flush();
  *     return sw.toString();
  * }
  * </pre>
  * 
  * </blockquote>
  * 
  * @author <a href="mailto:james.strachan@metastuff.com">James Strachan </a>
  * @author Laramie Crocker
  * @version $Revision: 1.21 $
  */
 public class HTMLWriter extends XMLWriter {
     private static String lineSeparator = System.getProperty("line.separator");
 
     protected static final HashSet DEFAULT_PREFORMATTED_TAGS;
 
     static {
         // If you change this list, update the javadoc examples, above in the
         // class javadoc, in writeElement, and in setPreformattedTags().
         DEFAULT_PREFORMATTED_TAGS = new HashSet();
         DEFAULT_PREFORMATTED_TAGS.add("PRE");
         DEFAULT_PREFORMATTED_TAGS.add("SCRIPT");
         DEFAULT_PREFORMATTED_TAGS.add("STYLE");
         DEFAULT_PREFORMATTED_TAGS.add("TEXTAREA");
     }
 
     protected static final OutputFormat DEFAULT_HTML_FORMAT;
 
     static {
         DEFAULT_HTML_FORMAT = new OutputFormat("  ", true);
         DEFAULT_HTML_FORMAT.setTrimText(true);
         DEFAULT_HTML_FORMAT.setSuppressDeclaration(true);
     }
 
     private Stack formatStack = new Stack();
 
     private String lastText = "";
 
     private int tagsOuput = 0;
 
     // legal values are 0+, but -1 signifies lazy initialization.
     private int newLineAfterNTags = -1;
 
     private HashSet preformattedTags = DEFAULT_PREFORMATTED_TAGS;
 
     /***
      * Used to store the qualified element names which should have no close
      * element tag
      */
     private HashSet omitElementCloseSet;
 
     public HTMLWriter(Writer writer) {
         super(writer, DEFAULT_HTML_FORMAT);
     }
 
     public HTMLWriter(Writer writer, OutputFormat format) {
         super(writer, format);
     }
 
     public HTMLWriter() throws UnsupportedEncodingException {
         super(DEFAULT_HTML_FORMAT);
     }
 
     public HTMLWriter(OutputFormat format) throws UnsupportedEncodingException {
         super(format);
     }
 
     public HTMLWriter(OutputStream out) throws UnsupportedEncodingException {
         super(out, DEFAULT_HTML_FORMAT);
     }
 
     public HTMLWriter(OutputStream out, OutputFormat format)
             throws UnsupportedEncodingException {
         super(out, format);
     }
 
     public void startCDATA() throws SAXException {
     }
 
     public void endCDATA() throws SAXException {
     }
 
     // Overloaded methods
     // added isXHTML() stuff so you get the CDATA brackets if you desire.
     protected void writeCDATA(String text) throws IOException {
         // XXX: Should we escape entities?
         // writer.write( escapeElementEntities( text ) );
         if (getOutputFormat().isXHTML()) {
             super.writeCDATA(text);
         } else {
             writer.write(text);
         }
 
         lastOutputNodeType = Node.CDATA_SECTION_NODE;
     }
 
     protected void writeEntity(Entity entity) throws IOException {
         writer.write(entity.getText());
         lastOutputNodeType = Node.ENTITY_REFERENCE_NODE;
     }
 
     protected void writeDeclaration() throws IOException {
     }
 
     protected void writeString(String text) throws IOException {
         /*
          * DOM stores \n at the end of text nodes that are newlines. This is
          * significant if we are in a PRE section. However, we only want to
          * output the system line.separator, not \n. This is a little brittle,
          * but this function appears to be called with these lineseparators as a
          * separate TEXT_NODE. If we are in a preformatted section, output the
          * right line.separator, otherwise ditch. If the single \n character is
          * not the text, then do the super thing to output the text.
          * 
          * Also, we store the last text that was not a \n since it may be used
          * by writeElement in this class to line up preformatted tags.
          */
         if (text.equals("\n")) {
             if (!formatStack.empty()) {
                 super.writeString(lineSeparator);
             }
 
             return;
         }
 
         lastText = text;
 
         if (formatStack.empty()) {
             super.writeString(text.trim());
         } else {
             super.writeString(text);
         }
     }
 
     /***
      * Overriden method to not close certain element names to avoid wierd
      * behaviour from browsers for versions up to 5.x
      * 
      * @param qualifiedName
      *            DOCUMENT ME!
      * 
      * @throws IOException
      *             DOCUMENT ME!
      */
     protected void writeClose(String qualifiedName) throws IOException {
         if (!omitElementClose(qualifiedName)) {
             super.writeClose(qualifiedName);
         }
     }
 
     protected void writeEmptyElementClose(String qualifiedName)
             throws IOException {
         if (getOutputFormat().isXHTML()) {
             // xhtml, always check with format object whether to expand or not.
             if (omitElementClose(qualifiedName)) {
                 // it was a special omit tag, do it the XHTML way: "<br/>",
                 // ignoring the expansion option, since <br></br> is OK XML,
                 // but produces twice the linefeeds desired in the browser.
                 // for netscape 4.7, though all are fine with it, write a space
                 // before the close slash.
                 writer.write(" />");
             } else {
                 super.writeEmptyElementClose(qualifiedName);
             }
         } else {
             // html, not xhtml
             if (omitElementClose(qualifiedName)) {
                 // it was a special omit tag, do it the old html way: "<br>".
                 writer.write(">");
             } else {
                 // it was NOT a special omit tag, check with format object
                 // whether to expand or not.
                 super.writeEmptyElementClose(qualifiedName);
             }
         }
     }
 
     protected boolean omitElementClose(String qualifiedName) {
         return internalGetOmitElementCloseSet().contains(
                 qualifiedName.toUpperCase());
     }
 
     private HashSet internalGetOmitElementCloseSet() {
         if (omitElementCloseSet == null) {
             omitElementCloseSet = new HashSet();
             loadOmitElementCloseSet(omitElementCloseSet);
         }
 
         return omitElementCloseSet;
     }
 
     // If you change this, change the javadoc for getOmitElementCloseSet.
     protected void loadOmitElementCloseSet(Set set) {
         set.add("AREA");
         set.add("BASE");
         set.add("BR");
         set.add("COL");
         set.add("HR");
         set.add("IMG");
         set.add("INPUT");
         set.add("LINK");
         set.add("META");
         set.add("P");
         set.add("PARAM");
     }
 
     // let the people see the set, but not modify it.
 
     /***
      * A clone of the Set of elements that can have their close-tags omitted. By
      * default it should be "AREA", "BASE", "BR", "COL", "HR", "IMG", "INPUT",
      * "LINK", "META", "P", "PARAM"
      * 
      * @return A clone of the Set.
      */
     public Set getOmitElementCloseSet() {
         return (Set) (internalGetOmitElementCloseSet().clone());
     }
 
     /***
      * To use the empty set, pass an empty Set, or null:
      * 
      * <pre>
      * 
      * 
      *       setOmitElementCloseSet(new HashSet());
      *     or
      *       setOmitElementCloseSet(null);
      * 
      *  
      * </pre>
      * 
      * @param newSet
      *            DOCUMENT ME!
      */
     public void setOmitElementCloseSet(Set newSet) {
         // resets, and safely empties it out if newSet is null.
         omitElementCloseSet = new HashSet();
 
         if (newSet != null) {
             omitElementCloseSet = new HashSet();
 
             Object aTag;
             Iterator iter = newSet.iterator();
 
             while (iter.hasNext()) {
                 aTag = iter.next();
 
                 if (aTag != null) {
                     omitElementCloseSet.add(aTag.toString().toUpperCase());
                 }
             }
         }
     }
 
     /***
      * @see #setPreformattedTags(java.util.Set) setPreformattedTags
      */
     public Set getPreformattedTags() {
         return (Set) (preformattedTags.clone());
     }
 
     /***
      * <p>
      * Override the default set, which includes PRE, SCRIPT, STYLE, and
      * TEXTAREA, case insensitively.
      * </p>
      * 
      * <p>
      * <b>Setting Preformatted Tags </b>
      * </p>
      * 
      * <p>
      * Pass in a Set of Strings, one for each tag name that should be treated
      * like a PRE tag. You may pass in null or an empty Set to assign the empty
      * set, in which case no tags will be treated as preformatted, except that
      * HTML Comments will continue to be preformatted. If a tag is included in
      * the set of preformatted tags, all whitespace within the tag will be
      * preserved, including whitespace on the same line preceding the close tag.
      * This will generally make the close tag not line up with the start tag,
      * but it preserves the intention of the whitespace within the tag.
      * </p>
      * 
      * <p>
      * The browser considers leading whitespace before the close tag to be
      * significant, but leading whitespace before the open tag to be
      * insignificant. For example, if the HTML author doesn't put the close
      * TEXTAREA tag flush to the left margin, then the TEXTAREA control in the
      * browser will have spaces on the last line inside the control. This may be
      * the HTML author's intent. Similarly, in a PRE, the browser treats a
      * flushed left close PRE tag as different from a close tag with leading
      * whitespace. Again, this must be left up to the HTML author.
      * </p>
      * 
      * <p>
      * <b>Examples </b>
      * </p>
      * <blockquote>
      * <p>
      * Here is an example of how you can set the PreformattedTags list using
      * setPreformattedTags to include IFRAME, as well as the default set, if you
      * have an instance of this class named myHTMLWriter:
      * 
      * <pre>
      * Set current = myHTMLWriter.getPreformattedTags();
      * current.add(&quot;IFRAME&quot;);
      * myHTMLWriter.setPreformattedTags(current);
      * 
      * //The set is now &lt;b&gt;PRE, SCRIPT, STYLE, TEXTAREA, IFRAME&lt;/b&gt;
      * 
      * 
      * </pre>
      * 
      * Similarly, you can simply replace it with your own:
      * 
      * <pre>
      * 
      * 
      *       HashSet newset = new HashSet();
      *       newset.add(&quot;PRE&quot;);
      *       newset.add(&quot;TEXTAREA&quot;);
      *       myHTMLWriter.setPreformattedTags(newset);
      * 
      *       //The set is now &lt;b&gt;{PRE, TEXTAREA}&lt;/b&gt;
      * 
      *  
      * </pre>
      * 
      * You can remove all tags from the preformatted tags list, with an empty
      * set, like this:
      * 
      * <pre>
      * 
      * 
      *       myHTMLWriter.setPreformattedTags(new HashSet());
      * 
      *       //The set is now &lt;b&gt;{}&lt;/b&gt;
      * 
      *  
      * </pre>
      * 
      * or with null, like this:
      * 
      * <pre>
      * 
      * 
      *       myHTMLWriter.setPreformattedTags(null);
      * 
      *       //The set is now &lt;b&gt;{}&lt;/b&gt;
      * 
      *  
      * </pre>
      * 
      * </p>
      * </blockquote>
      * 
      * @param newSet
      *            DOCUMENT ME!
      */
     public void setPreformattedTags(Set newSet) {
         // no fancy merging, just set it, assuming they did a
         // getExcludeTrimTags() first if they wanted to preserve the default
         // set.
         // resets, and safely empties it out if newSet is null.
         preformattedTags = new HashSet();
 
         if (newSet != null) {
             Object aTag;
             Iterator iter = newSet.iterator();
 
             while (iter.hasNext()) {
                 aTag = iter.next();
 
                 if (aTag != null) {
                     preformattedTags.add(aTag.toString().toUpperCase());
                 }
             }
         }
     }
 
     /***
      * DOCUMENT ME!
      * 
      * @param qualifiedName
      *            DOCUMENT ME!
      * 
      * @return true if the qualifiedName passed in matched (case-insensitively)
      *         a tag in the preformattedTags set, or false if not found or if
      *         the set is empty or null.
      * 
      * @see #setPreformattedTags(java.util.Set) setPreformattedTags
      */
     public boolean isPreformattedTag(String qualifiedName) {
         // A null set implies that the user called setPreformattedTags(null),
         // which means they want no tags to be preformatted.
         return (preformattedTags != null)
                 && (preformattedTags.contains(qualifiedName.toUpperCase()));
     }
 
     /***
      * This override handles any elements that should not remove whitespace,
      * such as &lt;PRE&gt;, &lt;SCRIPT&gt;, &lt;STYLE&gt;, and &lt;TEXTAREA&gt;.
      * Note: the close tags won't line up with the open tag, but we can't alter
      * that. See javadoc note at setPreformattedTags.
      * 
      * @param element
      *            DOCUMENT ME!
      * 
      * @throws IOException
      *             When the stream could not be written to.
      * 
      * @see #setPreformattedTags(java.util.Set) setPreformattedTags
      */
     protected void writeElement(Element element) throws IOException {
         if (newLineAfterNTags == -1) { // lazy initialization check
             lazyInitNewLinesAfterNTags();
         }
 
         if (newLineAfterNTags > 0) {
             if ((tagsOuput > 0) && ((tagsOuput % newLineAfterNTags) == 0)) {
                 super.writer.write(lineSeparator);
             }
         }
 
         tagsOuput++;
 
         String qualifiedName = element.getQualifiedName();
         String saveLastText = lastText;
         int size = element.nodeCount();
 
         if (isPreformattedTag(qualifiedName)) {
             OutputFormat currentFormat = getOutputFormat();
             boolean saveNewlines = currentFormat.isNewlines();
             boolean saveTrimText = currentFormat.isTrimText();
             String currentIndent = currentFormat.getIndent();
 
             // You could have nested PREs, or SCRIPTS within PRE... etc.,
             // therefore use push and pop.
             formatStack.push(new FormatState(saveNewlines, saveTrimText,
                     currentIndent));
 
             try {
                 // do this manually, since it won't be done while outputting
                 // the tag.
                 super.writePrintln();
 
                 if ((saveLastText.trim().length() == 0)
                         && (currentIndent != null)
                         && (currentIndent.length() > 0)) {
                     // We are indenting, but we want to line up with the close
                     // tag. lastText was the indent (whitespace, no \n) before
                     // the preformatted start tag. So write it out instead of
                     // the current indent level. This makes it line up with its
                     // close tag.
                     super.writer.write(justSpaces(saveLastText));
                 }
 
                 // actually, newlines are handled in this class by writeString,
                 // depending on if the stack is empty.
                 currentFormat.setNewlines(false);
                 currentFormat.setTrimText(false);
                 currentFormat.setIndent("");
 
                 // This line is the recursive one:
                 super.writeElement(element);
             } finally {
                 FormatState state = (FormatState) formatStack.pop();
                 currentFormat.setNewlines(state.isNewlines());
                 currentFormat.setTrimText(state.isTrimText());
                 currentFormat.setIndent(state.getIndent());
             }
         } else {
             super.writeElement(element);
         }
     }
 
     private String justSpaces(String text) {
         int size = text.length();
         StringBuffer res = new StringBuffer(size);
         char c;
 
         for (int i = 0; i < size; i++) {
             c = text.charAt(i);
 
             switch (c) {
                 case '\r':
                 case '\n':
 
                     continue;
 
                 default:
                     res.append(c);
             }
         }
 
         return res.toString();
     }
 
     private void lazyInitNewLinesAfterNTags() {
         if (getOutputFormat().isNewlines()) {
             // don't bother, newlines are going to happen anyway.
             newLineAfterNTags = 0;
         } else {
             newLineAfterNTags = getOutputFormat().getNewLineAfterNTags();
         }
     }
 
     // Convenience methods, static, with bunch-o-defaults
 
     /***
      * Convenience method to just get a String result.
      * 
      * @param html
      *            DOCUMENT ME!
      * 
      * @return a pretty printed String from the source string, preserving
      *         whitespace in the defaultPreformattedTags set, and leaving the
      *         close tags off of the default omitElementCloseSet set. Use one of
      *         the write methods if you want stream output.
      * 
      * @throws java.io.IOException
      * @throws java.io.UnsupportedEncodingException
      * @throws org.dom4j.DocumentException
      */
     public static String prettyPrintHTML(String html)
             throws java.io.IOException, java.io.UnsupportedEncodingException,
             org.dom4j.DocumentException {
         return prettyPrintHTML(html, true, true, false, true);
     }
 
     /***
      * Convenience method to just get a String result, but <b>As XHTML </b>.
      * 
      * @param html
      *            DOCUMENT ME!
      * 
      * @return a pretty printed String from the source string, preserving
      *         whitespace in the defaultPreformattedTags set, but conforming to
      *         XHTML: no close tags are omitted (though if empty, they will be
      *         converted to XHTML empty tags: &lt;HR/&gt; Use one of the write
      *         methods if you want stream output.
      * 
      * @throws java.io.IOException
      * @throws java.io.UnsupportedEncodingException
      * @throws org.dom4j.DocumentException
      */
     public static String prettyPrintXHTML(String html)
             throws java.io.IOException, java.io.UnsupportedEncodingException,
             org.dom4j.DocumentException {
         return prettyPrintHTML(html, true, true, true, false);
     }
 
     /***
      * DOCUMENT ME!
      * 
      * @param html
      *            DOCUMENT ME!
      * @param newlines
      *            DOCUMENT ME!
      * @param trim
      *            DOCUMENT ME!
      * @param isXHTML
      *            DOCUMENT ME!
      * @param expandEmpty
      *            DOCUMENT ME!
      * 
      * @return a pretty printed String from the source string, preserving
      *         whitespace in the defaultPreformattedTags set, and leaving the
      *         close tags off of the default omitElementCloseSet set. This
      *         override allows you to specify various formatter options. Use one
      *         of the write methods if you want stream output.
      * 
      * @throws java.io.IOException
      * @throws java.io.UnsupportedEncodingException
      * @throws org.dom4j.DocumentException
      */
     public static String prettyPrintHTML(String html, boolean newlines,
             boolean trim, boolean isXHTML, boolean expandEmpty)
             throws java.io.IOException, java.io.UnsupportedEncodingException,
             org.dom4j.DocumentException {
         StringWriter sw = new StringWriter();
         OutputFormat format = OutputFormat.createPrettyPrint();
         format.setNewlines(newlines);
         format.setTrimText(trim);
         format.setXHTML(isXHTML);
         format.setExpandEmptyElements(expandEmpty);
 
         HTMLWriter writer = new HTMLWriter(sw, format);
         Document document = DocumentHelper.parseText(html);
         writer.write(document);
         writer.flush();
 
         return sw.toString();
     }
 
     // Allows us to the current state of the format in this struct on the
     // formatStack.
     private class FormatState {
         private boolean newlines = false;
 
         private boolean trimText = false;
 
         private String indent = "";
 
         public FormatState(boolean newLines, boolean trimText, String indent) {
             this.newlines = newLines;
             this.trimText = trimText;
             this.indent = indent;
         }
 
         public boolean isNewlines() {
             return newlines;
         }
 
         public boolean isTrimText() {
             return trimText;
         }
 
         public String getIndent() {
             return indent;
         }
     }
 }
 
 /*
  * <html> <head> <title>My Title </title> <style> .foo { text-align: Right; }
  * </style> <script> function mojo(){ return "bar"; } </script> <script
  * language="JavaScript"> <!-- //this is the canonical javascript hiding.
  * function foo(){ return "foo"; } //--> </script> </head> <!-- this is a
  * comment --> <body bgcolor="#A4BFDD" mojo="&amp;"> entities: &#160; &amp;
  * &quot; &lt; &gt; %23 <p></p> <mojo> </mojo> <foo /> <table border="1"> <tr>
  * <td><pre> line0 <hr /> line1 <b>line2, should line up, indent-wise </b> line
  * 3 line 4 </pre></td><td></td></tr> </table> <myCDATAElement> <![CDATA[My
  * data]]> </myCDATAElement> </body> </html>
  */
 
 /*
  * Redistribution and use of this software and associated documentation
  * ("Software"), with or without modification, are permitted provided that the
  * following conditions are met:
  * 
  * 1. Redistributions of source code must retain copyright statements and
  * notices. Redistributions must also contain a copy of this document.
  * 
  * 2. Redistributions in binary form must reproduce the above copyright notice,
  * this list of conditions and the following disclaimer in the documentation
  * and/or other materials provided with the distribution.
  * 
  * 3. The name "DOM4J" must not be used to endorse or promote products derived
  * from this Software without prior written permission of MetaStuff, Ltd. For
  * written permission, please contact dom4j-info@metastuff.com.
  * 
  * 4. Products derived from this Software may not be called "DOM4J" nor may
  * "DOM4J" appear in their names without prior written permission of MetaStuff,
  * Ltd. DOM4J is a registered trademark of MetaStuff, Ltd.
  * 
  * 5. Due credit should be given to the DOM4J Project - http://www.dom4j.org
  * 
  * THIS SOFTWARE IS PROVIDED BY METASTUFF, LTD. AND CONTRIBUTORS ``AS IS'' AND
  * ANY EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  * ARE DISCLAIMED. IN NO EVENT SHALL METASTUFF, LTD. OR ITS CONTRIBUTORS BE
  * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
  * POSSIBILITY OF SUCH DAMAGE.
  * 
  * Copyright 2001-2005 (C) MetaStuff, Ltd. All Rights Reserved.
  */