Followings is my codes.
Any suggestion is welcome.
thank you
May goodness be with you all
--------------------------------------------------
code MyVisitor.java
--------------------------------------------------
import org.htmlparser.Parser;
import org.htmlparser.tags.LinkTag;
import org.htmlparser.util.ParserException;
import org.htmlparser.visitors.NodeVisitor;
import org.htmlparser.*;
import org.htmlparser.tags.ScriptTag;
import org.htmlparser.tags.ImageTag;
import java.util.Vector;
import java.util.Enumeration;
import org.htmlparser.tags.*;
import org.htmlparser.tags.StyleTag;
public class MyVisitor extends NodeVisitor
{
public MyVisitor()
{
}
public void visitTag(Tag tag)
{
Vector attributes;
Attribute attribute;
attributes = tag.getAttributesEx();
attribute = (Attribute)attributes.elementAt(0);
attribute.setName(attribute.getName().toLowerCase());
// System.out.print(tag.toHtml());
char MyQuote='"';
Vector AttributeVector;
AttributeVector=tag.getAttributesEx();
for (Enumeration e = AttributeVector.elements() ;
e.hasMoreElements()
{
((Attribute)e).setQuote(MyQuote);
}
if (tag instanceof LinkTag){
System.out.print(tag.getAttribute("href"));
System.out.print("****");
tag.setAttribute("href","
http://www.yzu.edu.tw/"+tag.getAttribute("href"
);
System.out.print(tag.getAttribute("href"));
System.out.print("////");
Vector AttributeLinkTagVector;
String JavaScriptbuffer;
AttributeLinkTagVector=tag.getAttributesEx();
for (Enumeration e2th=AttributeVector.elements() ;
e2th.hasMoreElements()
{
JavaScriptbuffer=((Attribute)e2th).getRawValue();
if(JavaScriptbuffer.startsWith("JavaScript"))
{
tag.removeAttribute("href");
tag.removeAttribute("onMouseOver");
tag.removeAttribute("onMouseOut");
}
}
}
if (tag instanceof DoctypeTag){
//System.out.print(tag.toHtml());
}
if (tag instanceof ImageTag){
tag.setAttribute("alt","*");
//System.out.print(tag.toHtml());
System.out.print(tag.getAttribute("src"));
System.out.print("****");
tag.setAttribute("src","
http://www.yzu.edu.tw"+tag.getAttribute("src"
);
System.out.print(tag.getAttribute("src"));
System.out.print("////");
}
if (tag instanceof Html){
/* String DOCTYPEHEADER1="<!DOCTYPE html PUBLIC \"-//W3C//DTD XHTML
1.0 Transitional//EN\"";
String DOCTYPEHEADER2="\"
http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd\">";;
String DOCTYPEHEADER=DOCTYPEHEADER1+DOCTYPEHEADER2;
*/
tag.setAttribute("xmlns","
http://www.w3.org/1999/xhtml"
;
tag.setAttribute("xml:lang","en-US");
tag.setAttribute("lang","en-US");
//<html xmlns="
http://www.w3.org/1999/xhtml"; xml:lang="en-US"
lang="en-US">
// Set attribute with given key, value pair.
System.out.print(tag.toHtml());
}
if(tag instanceof StyleTag){
tag.setAttribute("type","text/css",'"');
System.out.print(tag.toHtml());
}
if(tag instanceof StyleTag){
tag.setAttribute("type","text/css",'"');
System.out.print(tag.toHtml());
}
}
}
--------------------------------------------------
code ToHtmlDemoTest2.java
--------------------------------------------------
import org.htmlparser.Parser;
import org.htmlparser.util.NodeIterator;
import org.htmlparser.util.*;
import org.htmlparser.util.ParserException;
import org.htmlparser.visitors.HtmlPage;
import org.htmlparser.tags.*;
import org.htmlparser.visitors.NodeVisitor;
import org.htmlparser.*;
import org.htmlparser.filters.*;
import org.htmlparser.filters.*;
import java.io.*;
public class ToHtmlDemoTest2
{
public static void main (String[] args) throws ParserException
{
Parser parser = new Parser("
http://www.yzu.edu.tw"
;
MyVisitor visitor = new MyVisitor();
NodeList list = new NodeList();
NodeFilter filter=new NotFilter(new TagNameFilter("Script"));
NodeIterator iterator = parser.elements();
while(iterator.hasMoreNodes()){
list.add(iterator.nextNode());
}
list.keepAllNodesThatMatch(filter,true);
parser.visitAllNodesWith(visitor);
}
}
