原码
/**
* GET页面的内容
*
* @param cityMsg
* @return 返回HTML源代码
*/
private String getHtmlCode(String url,String code) {
String html=null;
try{
//为了避免并发量采集,睡眠半秒钟。
//Thread.sleep(500);
String sCurrentLine;
StringBuffer sTotalString = new StringBuffer("");
sCurrentLine="";
java.io.InputStream l_urlStream;
//cityMsg=java.net.URLEncoder.encode(cityMsg, "UTF-8");
java.net.URL l_url = new java.net.URL(url);
java.net.HttpURLConnection l_connection = (java.net.HttpURLConnection) l_url.openConnection();
//l_connection.setRequestProperty("User-Agent","Mozilla/6.0 (compatible; MSIE 5.0; Windows NT; DigExt)");
int state=l_connection.getResponseCode();
if(state==200){
l_connection.connect();
l_urlStream = l_connection.getInputStream();
java.io.BufferedReader l_reader = new java.io.BufferedReader(new java.io.InputStreamReader(l_urlStream,code));
//得到返回的信息 while ((sCurrentLine = l_reader.readLine()) != null) { sTotalString.append(sCurrentLine); sTotalString.append("\n"); } l_reader.close(); html = sTotalString.toString(); }else{ html=null; } l_connection.disconnect(); }catch(Exception ex){ log.error("读取网页异常:"+url, ex); } return html; }