source: trunk/greenstone3-extensions/gs3build/src/org/greenstone/gsdl3/gs3build/util/HTMLTidy.java@ 12188

package org.greenstone.gsdl3.gs3build.util;

import java.io.File;
import java.io.FileReader;

import java.net.URL;

import org.w3c.dom.*;

import org.xml.sax.XMLReader;
import org.xml.sax.InputSource;
import org.xml.sax.SAXException;
import org.xml.sax.Attributes;
import org.xml.sax.helpers.XMLReaderFactory;
import org.xml.sax.helpers.DefaultHandler;
import org.xml.sax.ext.LexicalHandler;

import org.greenstone.gsdl3.gs3build.doctypes.DocumentID;
import org.greenstone.gsdl3.gs3build.doctypes.DocumentInterface;
import org.greenstone.gsdl3.gs3build.doctypes.DocumentList;
import org.greenstone.gsdl3.gs3build.doctypes.GMLDocument;

public class HTMLTidy extends DefaultHandler
{
    String label;
    File   file;
    StringBuffer value;
    boolean inElement;
    boolean accumulate;
    Document document;
    Element  currentElement;
    Element root;

    class HTMLCommentHandler implements LexicalHandler
    {
    HTMLTidy parent;

    HTMLCommentHandler(HTMLTidy tidier)
    { 
        this.parent = tidier;
    }

    public void comment(char ch[], int start, int length)
    { 
        String comment = new String(ch, start, length);
        parent.comment(comment);
    }
    
    public void endCDATA()
    {
    }

    public void endDTD()
    {
    }

    public void endEntity(String name)
    {
    }

    public void startCDATA()
    {
    }

    public void startDTD(String name, String publicId, String systemId)
    {
    }

    public void startEntity(String name)
    {
    }
    }

    public HTMLTidy(URL url)
    { 
    // TODO: complete
    }

    public HTMLTidy(File file)
    { 
    this.file = null;
    this.label = null;
    this.value = null;
    
    this.document = new org.apache.html.dom.HTMLDocumentImpl();
    
    this.currentElement = this.document.getDocumentElement();
    this.root = this.currentElement;
    
    try {
        XMLReader handler  = new org.ccil.cowan.tagsoup.Parser();
        HTMLCommentHandler commentHandler = new HTMLCommentHandler(this);
        handler.setContentHandler(this);
        handler.setProperty("http://xml.org/sax/properties/lexical-handler", commentHandler);
        handler.parse(file.toURI().toString());
    }
    catch (java.io.IOException io)
        {
        System.out.println(io);
        }
    catch (org.xml.sax.SAXException saxEx)
        {
        System.out.println(saxEx);
        }
    }
    
    public void startElement(String URI, String localName, String qName, Attributes attributes)
    { 
    //System.out.println("<"+localName+">");
    
    if (localName.equals("html") == false) {
        Element childElement = this.document.createElement(localName);
        
        this.currentElement.appendChild(childElement);
        this.currentElement = childElement;
        
        for (int a = 0; a < attributes.getLength(); a ++){
        String name = attributes.getLocalName(a);
        String value = attributes.getValue(a);
        
        childElement.setAttribute(name, value);
        }
    }
    }
    
    public void comment(String text)
    {
    Comment comment = document.createComment(text);
    this.currentElement.appendChild(comment);
    //System.out.println("<!-- "+text+" -->");
    }
    
    public void endElement(String URI, String localName, String qName)
    { 
    // System.out.println("Ending " + localName);
    // System.out.println("</"+localName+">");
    
    if (localName.equals("html")) {
        return;
    }
    
    // TODO: check is the same name as current node
    if (!this.currentElement.getTagName().equals(localName.toUpperCase())) {
        // TODO: error handling
        System.out.println("Bad end at " + localName + " " + this.currentElement.getTagName());
        return;
    }
    
    // Optimisation: is this an empty tag?
    boolean isEmpty = false;
    
    if (this.currentElement.getChildNodes().getLength() == 1){
        Node child = this.currentElement.getFirstChild();
        
        if (child.getNodeType() == Node.TEXT_NODE) {
        String text = child.getNodeValue();
        
        int i = 0;
        while (i < text.length()){
            if (text.charAt(i) > 32 && text.charAt(i) != 160) {
            break;
            }
            i++;
        }
        if (text.length() == i) {
            isEmpty = true;
        }
        }
    }
    else if (this.currentElement.getChildNodes().getLength() == 0){
        isEmpty = true;
    }
    
    boolean isSingleton = false;
    String tagName = this.currentElement.getTagName().toLowerCase();
    
    if (tagName.equals("img") ||
        tagName.equals("br") ||
        tagName.equals("meta") ||
        tagName.equals("input") ||
        tagName.equals("area") ||
        tagName.equals("link") ||
        tagName.equals("base") ||
        //tagName.equals("img") ||
        tagName.equals("hr")) {
        isSingleton = true;
    }
    
    Element parent =  (Element) this.currentElement.getParentNode();

    if (isEmpty == true && !isSingleton) {
        // trim the current child
        parent.removeChild(this.currentElement);
        //      System.out.println("***Removing child***");
    }
    
    this.currentElement = parent;
    }
    
    public void characters(char c[], int start, int length)
    { 
    String string = new String(c, start, length);
    Node text_node = this.document.createTextNode(string);
    this.currentElement.appendChild(text_node);
    //    System.out.println(string);
    }
    
    public void recursePrint(Element element, int indent)
    {
    for (int i = 0; i < indent; i++) {
        System.out.print(" ");
    }
    System.out.println(element.toString());
    for (int c = 0; c < element.getChildNodes().getLength(); c ++){
        Node n = element.getChildNodes().item(c);
        
        if (n.getNodeType() == org.w3c.dom.Node.ELEMENT_NODE) {
        Element e = (Element) n;
        
        this.recursePrint(e, indent+2);
        }
    }
    }
    
    public Document getDocument()
    { 
    //    this.recursePrint(this.root, 0);
    //    System.out.println(this.document.toString());
    //    System.out.println(this.root);
    return this.document;
    }
}