用java抓取cnproxy代理服务器地址

package org.mingyuan.fetcher;
import java.io.BufferedOutputStream;
import java.io.BufferedReader;
import java.io.File;
import java.io.FileOutputStream;
import java.io.IOException;
import java.io.InputStreamReader;
import java.net.URL;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.Set;
public class StringTest {
/**
* 通过url，得到url源文件
* @param url
* @return
* @throws IOException
*/
private static String getURLContent(String url) throws IOException {
URL ipListUrl = new URL(url);
BufferedReader in = new BufferedReader(new InputStreamReader(ipListUrl
.openStream()));
String str = null;
String html = "";
while ((str = in.readLine()) != null) {
html += str;
}
in.close();
return html;
}
/**
*
* @param htmlContent
* @param fileName
* @return ip地址和ip地理信息
* @throws IOException
*/
private static HashMap<String,String> getIpList(String htmlContent,String fileName) throws IOException {
String str1[] = htmlContent
.split("<tr><td width=\"140\">IP:Port</td><td width=\"40\">Type</td><td width=\"90\">Speed</td><td width=\"160\"> Country/Area</td></tr>");
String str2 = str1[1];// ip及余下部分
String str3[] = str2
.split("</table>");
String str4 = str3[0];// <tr><td>24.25.26.128<SCRIPT type=text/javascript>document.write(":"+q+d)</SCRIPT></td><td>HTTP</td><td>296,984,984</td><td>美国 维吉尼亚州</td></tr>

String str5[]=str4.split("<tr><td>");
HashMap<String,String> map=