/* this belongs to the WebRaveler package */
package webraveler.html; 
import java.io.*;
import au.id.jericho.lib.html.*;
import java.util.*;

import webraveler.http.HTTPRequest;
import webraveler.http.HTTPResponse;



/**
 * An HTMLPage2 is essentially just a String.  The interesting
 * stuff all happens in the methods, but there's no extra 
 * overhead of carrying around vectors or hashes or whatnot.  
 * Just a String that we modify with the handy Jericho HTML
 * Parser.
 * 
 * @author The Web Raveler Development Squad
 * @version 1.0
 */
public class HTMLPage2 implements Serializable {
	private String html;
	
	/**
	 * Make a new HTMLPage2 from the raw HTML string.
	 * 
	 * @param html a String with html source to store
	 */
	public HTMLPage2(String html) {
		this.html = html;
	}//HTMLPage2(String)
	
	
	/**
	 * Set the page's title to the given String.
	 * 
	 * @param html
	 */
	public void setTitle(String title) {
		//make a new OutputDocument
		OutputDocument outputDocument=new OutputDocument(html);
		
		//make a Segment which spans the entire document
		Source source=new Source(html);
		
		//get a list of the BODY tags (though there should only be one)
		List linkStartTags=source.findAllStartTags(Tag.TITLE);
		
		//iterate over the TITLE tags
		for (Iterator i=linkStartTags.iterator(); i.hasNext();) {
			StartTag startTag=(StartTag)i.next();
			
			//get the text following the TITLE tag
			Segment textSegment = startTag.getFollowingTextSegment();
			
			//replace the above segment with the given String
			outputDocument.add(new StringOutputSegment(textSegment,title));
		}//for
		
		//dump the resulting document to our html string
		html = outputDocument.toString();
	}//void setTitle(String)
	
	
	/**
	 * Set the page's HTML to the specified string.
	 * 
	 * @param html a String containing the new HTML
	 */
	public void setText(String html) {
		this.html = html;
	}//void setText(String)
	
	
	/**
	 * Replace text matching the given pattern (a regular expression) with
	 * the replacement String.
	 * 
	 * @param pattern a String regex whose matches we will replace
	 * @param replacement the String containing the replacement text
	 */
	public void replaceText(String pattern, String replacement) {
	        String oldText; //storage for original text
		String newText; //storage for replacement text
	
		//make a new OutputDocument
		OutputDocument outputDocument=new OutputDocument(html);
		
		//make a Segment which spans the entire document
		Source source=new Source(html);
		
		//get a list of the tags
		List startTags=source.findAllStartTags();
		
		//iterate over the tags
		for (Iterator i=startTags.iterator(); i.hasNext();) {
			StartTag startTag=(StartTag)i.next(); //the next tag
			
			//check for an "alt" attribute, replacing text as necessary
			Attributes attribs = startTag.getAttributes();
			if ( attribs != null ) {
				Attribute attr = attribs.get("alt");
				if ( attr != null ) {
					Segment valueSegment = attr.getValueSegment();
					oldText = valueSegment.getSourceText();
					newText = oldText.replaceAll(pattern, replacement);
					outputDocument.add(new StringOutputSegment(valueSegment,newText));
				}//if
			}//if
			
			//get the text following the tag
			Segment textSegment = startTag.getFollowingTextSegment();
			
			//replace the pattern in the above segment
			oldText = textSegment.getSourceText();
			newText = oldText.replaceAll(pattern,replacement);
			outputDocument.add(new StringOutputSegment(textSegment,newText));
		}//for
		
		//dump the resulting document to our html string
		html = outputDocument.toString();
	}//void replaceText(String,String)
	 
	 
	
	/**
	 * Insert some text immediately after the HEAD tag (???).
	 * 
	 * @param textToInsert
	 */
	public void insertAtTop(String textToInsert) {
		//make a new OutputDocument
		OutputDocument outputDocument=new OutputDocument(html);
		
		//make a Segment which spans the entire document
		Source source=new Source(html);
		
		//get a list of the BODY tags (though there should only be one)
		List linkStartTags=source.findAllStartTags(Tag.BODY);
		
		//iterate over the BODY tags
		for (Iterator i=linkStartTags.iterator(); i.hasNext();) {
			StartTag startTag=(StartTag)i.next();
			
			//get the text following the BODY tag
			Segment textSegment = startTag.getFollowingTextSegment();
			
			//add the new text to the start of the above text segment
			outputDocument.add(new StringOutputSegment(textSegment,textToInsert+textSegment.getSourceText()));
		}//for
		
		//dump the resulting document to our html string
		html = outputDocument.toString();
	}//void insertAtTop(String)
	
	
	/**
	 * Insert some text immediately before the end tag for the 
	 * BODY element.
	 * 
	 * @param textToInsert
	 */
	public void insertAtBottom(String textToInsert) {
		//make a new OutputDocument
		OutputDocument outputDocument=new OutputDocument(html);
		
		//make a Segment which spans the entire document
		Source source=new Source(html);
		
		//get a list of the BODY tags (though there should only be one)
		List linkStartTags=source.findAllStartTags(Tag.BODY);
		
		//iterate over the BODY tags
		for (Iterator i=linkStartTags.iterator(); i.hasNext();) {
			StartTag startTag=(StartTag)i.next();
			
			//get the matching end tag
			EndTag endTag = startTag.getElement().getEndTag();
			if ( endTag == null ) return; //if there's no closing tag, quit
			
			//get the index in the source where the end tag starts
			int insertionPoint = endTag.getBegin();
			
			//add the new text to the start of the above text segment
			outputDocument.add(new StringOutputSegment(insertionPoint,insertionPoint,textToInsert));
		}//for
		
		//dump the resulting document to our html string
		html = outputDocument.toString();
	}//void insertAtBottom(String)
	
	
	/**
	 * Insert some text immediately after the HEAD tag.
	 * 
	 * @param textToInsert
	 */
	public void insertInHead(String textToInsert) {
		//make a new OutputDocument
		OutputDocument outputDocument=new OutputDocument(html);
		
		//make a Segment which spans the entire document
		Source source=new Source(html);
		
		//get a list of the HEAD tags (though there should only be one)
		List linkStartTags=source.findAllStartTags(Tag.HEAD);
		
		//iterate over the HEAD tags
		for (Iterator i=linkStartTags.iterator(); i.hasNext();) {
			StartTag startTag=(StartTag)i.next();
			
			//get the text following the HEAD tag
			Segment textSegment = startTag.getFollowingTextSegment();
			
			//add the new text to the start of the above text segment
			outputDocument.add(new StringOutputSegment(textSegment,textToInsert+textSegment.getSourceText()));
		}//for
		
		//dump the resulting document to our html string
		html = outputDocument.toString();
	}//void insertAtBottom(String)
	
	/**
	 * Simply return the html source
	 * 
	 * @return a String representation of this page
	 */
	public String toString() {
		return html;
	}//String toString()
	
	
	/**
	 * Return an Enumeration of Strings containing all the
	 * text between HTML tags or comments.
	 * 
	 * @return an Enumeration of Strings of untagged text in the document
	 */
	public Enumeration getTextBlocks() {
		// we'll put the strings in a vector, then use Vector.elements()
		// to get the enumeration
		Vector vector = new Vector();

		//make a new OutputDocument
		OutputDocument outputDocument=new OutputDocument(html);
		
		//make a Segment which spans the entire document
		Source source=new Source(html);
		
		//get a list of the tags
		List startTags=source.findAllStartTags();
		
		//iterate over the tags
		for (Iterator i=startTags.iterator(); i.hasNext();) {
			StartTag startTag=(StartTag)i.next(); //the next tag
			
			//get the text following the tag, and put it in the vector
			Segment textSegment = startTag.getFollowingTextSegment();
			String text = textSegment.getSourceText();
			vector.add(text);
		}//for

		return vector.elements();
	}// Enumeration getTextBlocks()
	
	
	public void makeURLsAbsolute(HTTPRequest request, HTTPResponse response) {
		//System.out.println("HTMLPage2: Making urls absolute");
		String serverRoot;
		String uriString = request.uriString();
		String docRoot;

		if (response.getHtmlPage() == null) return;
		
		if (uriString.equalsIgnoreCase("http://webraveler2.banner.bannerhtml")) return;
		
		/* figure out the document root and server root */		       
		int thirdSlash = request.uriString().indexOf("/",7);
		
		if ( thirdSlash == -1 ) {
			uriString += "/";  
			thirdSlash = uriString.length()-1;
		}//if
		
		/* the prefix for URLs that start with "/" */
		serverRoot = uriString.substring(0,thirdSlash);
		
		/* Fix for uris with /~name */
		if (uriString.length() > thirdSlash+1 && uriString.charAt(thirdSlash+1) == '~') {
			int fourthSlash = uriString.indexOf('/',thirdSlash+1);
			if ( fourthSlash == -1 ) {
				uriString += "/";  
				fourthSlash = uriString.length()-1;
			} //if
			serverRoot = uriString.substring(0,fourthSlash);
		}//if
		
		/* it's nice to know where the last slash in the URL is, so that 
		 * we know the base for URLs that don't start with "/" */
		int lastSlash = uriString.lastIndexOf("/");
		if ( lastSlash == 6 ) {
			uriString += "/";
			lastSlash = uriString.length()-1;
		}//if
		
		/* this string is the prefix for URLs that don't start with "/" or "http://" */
		docRoot = uriString.substring(0,lastSlash+1);
		
		/* make URLs absolute for the following attribute names */		       
		makeAttributeAbsolute(new String[]{"href","src","action"},docRoot,serverRoot);
	}
	
	
	
	private void makeAttributeAbsolute(String[] attributeNames,String docRoot,String serverRoot) {
		//Vector srcVector = (Vector) attributes.get(attributeName);
		
		//iterate over all StartTags in the document, 
		//iterating over all matching Attribute elements.
		
		//make a new OutputDocument
		OutputDocument outputDocument=new OutputDocument(html);
		
		//make a Segment which spans the entire document
		Source source=new Source(html);
		
		//iterate over all tags in the document
		for ( Iterator i = source.getNextTagIterator(0) ; i.hasNext() ; ) {
			try {
				StartTag tag = (StartTag) i.next();
				Attributes attribs = tag.getAttributes();
				if ( attribs == null ) continue;
				for ( int k = 0 ; k < attributeNames.length ; k++ ) {
					Attribute attr = attribs.get(attributeNames[k]);
					if ( attr == null ) continue;
					Segment valueSegment = attr.getValueSegment();
					StringBuffer sourceText = new StringBuffer(valueSegment.getSourceText());
					
					//don't change urls that start with '#'
					if ( sourceText.charAt(0) == '#' ) continue;
					
					//a path from the server root
					if ( sourceText.charAt(0) == '/' ) {
						sourceText.insert(0, serverRoot);
					} else if (sourceText.toString().startsWith("http://") ) {
						// add trailing "/" to urls with no explicit paths from server root 
						if ( sourceText.indexOf("/",7) == -1 ) { 
							sourceText.insert(sourceText.length(),"/");
						}
					} else {
						// get the document root, allowing for ".." in the path
						// ( e.g., "../../foo.html" )
						String tempDocRoot = docRoot;
						while ( sourceText.toString().startsWith("../") ) {
							tempDocRoot = tempDocRoot.substring(0,1 + tempDocRoot.lastIndexOf('/',tempDocRoot.length()-2));
							sourceText = sourceText.delete(0,3);
						}//while
						
						//for, e.g., "./foo.html"
						if ( sourceText.toString().startsWith("./") ) {
							sourceText = sourceText.delete(0,2);
						}//if
						
						sourceText.insert(0, tempDocRoot); //just add the docroot
					}//else
					
					// if it's an href, don't add the vurl
					if ( !attributeNames[k].equals("href") || !tag.getName().equalsIgnoreCase("a") ) { 
						sourceText.insert(0,"http://temp.webraveler2.getunframeddocument/getunframeddocument!");
					} else {
						//for hrefs in "a" tags, make the "target" attribute equal '_top'
						//in the enclosing tag
						Attribute targetAttr = attribs.get("target");
						if ( targetAttr == null ) {
							//add a new "target" attribute to the tag
							outputDocument.add(new StringOutputSegment(tag.getEnd()-1,tag.getEnd()-1," target='_top'"));
						} else {
							//modify the existing "target" attribute
							Segment value = targetAttr.getValueSegment();
							outputDocument.add(new StringOutputSegment(value,"_top"));
						}//else
					}//else
					
					//add the changes to our OutputDocument
					outputDocument.add(new StringOutputSegment(valueSegment,sourceText.toString()));
				}//for
			} catch( Exception e ) { }
		}//for
		html = outputDocument.toString();
	}//makeAttributeAbsolute
	
	
	
	/**
	 * Do a little self-testing...
	 * 
	 * @param args
	 */
	public static void main( String[] args ) {
		String text = "<html>\n<head>\n<title>FOO</title>\n</head>\n";
		text += "<body>\n<a href='http://www.google.com'>Hello, world!</a>\n<img src=\"foo.jpg\" ALT=\"Google\">\n<br>\nGoogle was here!\n<img src='foo.jpg' alt=\"FOO\"></body>\n</html>\n";
		System.out.println("text:======>\n"+text+"\n");
		HTMLPage2 foo = new HTMLPage2(text);
		foo.insertAtTop("INSERTED_AT_THE_TOP");
		System.out.println("HTMLPage2 output:======>\n"+foo);
		foo.insertAtBottom("INSERTED_AT_THE_BOTTOM");
		System.out.println("HTMLPage2 output 2:======>\n"+foo);
		foo.setTitle("HAHAHAHAHA! I CHANGED THE TITLE!");
		System.out.println("HTMLPage2 output 3:======>\n"+foo);
		foo.replaceText("[Gg]oogle","REPLACED SOME TEXT");
		System.out.println("HTMLPage2 output 4:======>\n"+foo);
		foo.replaceText("LACE","----");
		System.out.println("HTMLPage2 output 5:======>\n"+foo);

		System.out.println("\nGetting text blocks:");
		//test the getTextBlocks() method
		int i=1;
		for ( Enumeration e = foo.getTextBlocks() ; e.hasMoreElements() ; i++ ) {
		    String str = (String) e.nextElement();
		    System.out.println(i+" - '"+str+"'");
		}//for
	}//main
} // class HTMLPage