VisClient/org/hfbk/vis/source/SourceYoutube.java

Go to the documentation of this file.

package org.hfbk.vis.source;

import java.io.IOException;
import java.io.UnsupportedEncodingException;
import java.net.URL;
import java.util.HashSet;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

import org.dronus.graph.Node;
import org.hfbk.util.HTTPUtils;
import org.hfbk.util.Sleeper;

public class SourceYoutube extends SourceRegExp {

Pattern endOfInterestingPart=Pattern.compile(matcher());
        HashSet<String> parsed=new HashSet<String>();
        
        int subfetches=0;
        
        
        public SourceYoutube(URL url) {
                super(url);
                root=add(keyword, "imagefield", null); 
                String start=getParam(url, "start");
                if (start==null) start="0";
                add(""+Integer.parseInt(start)+1, "next", root);
        }

        
        String matcher() {
                return "(?i)href=\"(/watch\\?v=[^\"]*)\".*?img.*?src=\"([^\"]*)\".*?title=\"[^\"]*\"";
        }

        String url() throws UnsupportedEncodingException {
                String start=getParam(url, "start");
                if (start==null) start="";
                return "http://youtube.com/results?search_query="+HTTPUtils.encode(keyword)
                        +"&page="+HTTPUtils.encode(start);

        }

        void parse(final String[] matches) throws IOException {
                if (parsed.contains(matches[1]) || parsed.size()>=8) return;
                parsed.add(matches[1]);
                
                // now launch a subfetch to parse the movie page for this search result
                Thread subfetcher=new Thread(){
                        public void run(){
                                String playerURL="http://youtube.com"+matches[1];
                                
                                String playerpage=HTTPUtils.fetchUntil(playerURL, endOfInterestingPart, silent);
                
                                Pattern p=Pattern.compile("(?i)\"video_id\": \"(.*?)\".*?\"t\": \"(.*?)\"");
                                Matcher m=p.matcher(playerpage);
                                m.find();
                                String fetchURL="http://www.youtube.com/get_video?video_id="+m.group(1)+"&t="+m.group(2);
                
                                synchronized(graph){
                                        Node video=add(fetchURL, "video", root);
                                        add(matches[2], "thumbnail", video);
                                        add(playerURL, "URL", video);
                                }
                                subfetches--;
                        }
                };
                subfetcher.start();
                subfetches++;
        }
        
        void buildGraph() throws IOException {
                super.buildGraph();
                while(subfetches>0)
                        Sleeper.sleep(10);              
        }
}

---