// Chamando Robot public static void robot() throws Exception { Database db = new Database(); db.connect(); ResultSet rs = Page.findAll(db); Page p = null; while ((p = Page.next(rs)) != null) { String body = Robot.get(p.getUrl()); // procurar por urls dentro do body // buscar por essas paginas // String expr = "href=\"([^\"]*)"; String ereg = "href=\"https{0,1}:\\/\\/([^\"]*)\""; Pattern pt = Pattern.compile(ereg); Matcher m = pt.matcher(body); while (m.find()) { System.out.println(m.group()); String[] _url = m.group().split("\""); Page.newUrl(_url[1]); } p.setBody(body); p.update(); } db.close(); }
public static void newUrl(String url) throws Exception { String body = Robot.get(url); Page p = new Page(); p.setTmsp("2011-03-11 11:00:00"); p.setUrl(url); p.setBody(body); p.insert(); }