Java code for extracting contents from facebook page.

GetContentsFromFbPage.java

package com.digipepper.test.html;

import java.io.BufferedReader;
import java.io.File;
import java.io.IOException;
import java.util.Iterator;
import java.util.LinkedList;
import java.util.List;

import org.apache.commons.io.FileUtils;

// http://jsoup.org/
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;


/*
 * Facebook page
 */

public class GetContentsFromFbPage {
  
	public final static void main(String[] args) throws Exception {
		
		File objRead = new File("/Users/tf0054/test/fblog.txt");
		File objWrite = new File("/Users/tf0054/test/fblog.html");
		Document doc = Jsoup.parse(objRead, "UTF-8");
		Elements timelineUnitContainers = doc.select(".timelineUnitContainer");

		int i = 0;
		Element objElement = null;
		Elements objUiLinkSubtle = null;
		Iterator<Element> objIte = timelineUnitContainers.iterator();
		while(objIte.hasNext()){
			objElement = objIte.next();
			objUiLinkSubtle = objElement.select(".uiLinkSubtle");
			if(objElement.select(".userContentWrapper").html().length() > 0){ //date(August 12, 2011)
				//System.out.print(i+","+objElement.select(".userContentWrapper").html().length());
				System.out.print(i);
				System.out.print("\t");
				System.out.print(objUiLinkSubtle.select("abbr").get(0).html().toString());
				System.out.print("\t");
				System.out.print(objElement.select(".userContent").html().replaceAll("\n", " "));
				System.out.print("\t");
				System.out.print(objElement.select(".scaledImageFitWidth").toString().replaceAll("\n", " "));
				System.out.print("\t");
				System.out.print(objElement.select(".shareUnit img").toString().replaceAll("\n", " "));
				System.out.print("\t");
				System.out.print(objElement.select(".photoUnit img").toString().replaceAll("\n", " "));
				System.out.print("\n");
				i++;
			}
		}
		FileUtils.writeStringToFile(objWrite, timelineUnitContainers.toString());
	}
	
	private static String getListFromFile(String strFilename){
		try {
			return FileUtils.readFileToString(new File(strFilename), "UTF-8");
		} catch (IOException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
			return null;
		}
	}

}