import java.io.BufferedWriter;
import java.io.File;
import java.io.FileOutputStream;
import java.io.FileWriter;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
import java.net.URL;
import java.net.URLConnection;
import java.util.HashMap;
import java.util.Map.Entry;

import org.apache.http.HttpEntity;
import org.apache.http.HttpResponse;
import org.apache.http.HttpStatus;
import org.apache.http.client.HttpClient;
import org.apache.http.client.methods.HttpGet;
import org.apache.http.impl.client.DefaultHttpClient;
import org.apache.http.util.EntityUtils;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;

/**
 * 獲取豌豆莢網(wǎng)頁上安卓軟件全部軟件
 * 注意：執(zhí)行程序前，一定要在這個工程目錄下創(chuàng)建icon文件夾
 * 所有圖標將下載到icon這個文件夾中
 * 應(yīng)用名稱與包名寫到了icon下的packageName.txt文件里
 *
 * 這個程序用到的jar包：
 * commons-logging-1.1.3.jar
 * httpclient-4.1.2.jar
 * httpcore-4.3.jar
 * jsoup-1.6.1.jar
 *
 *
 */
public class AppIconDownloader {

 /**
  * @param args
  */
 public static void main(String[] args) {

  String rootUrl = "http://www.wandoujia.com/tag/全部軟件/total?page=";
  //String rootUrl = "http://www.wandoujia.com/tag/全部游戲/total?page=";
  //下載1到20頁的應(yīng)用圖標
  for(int i = 1; i < = 20; i++) {
   System.out.println("【下載進度】：準備下載第" + i + "頁");
   String currentUrl = rootUrl + i;
   HashMap<String, String> apps = new HashMap<string , String>();
   apps = getAppImageUrl(currentUrl);
   //遍歷HashMap逐個下載圖標
   for(Entry</string><string , String> entry : apps.entrySet()) {
    try{
     //下載圖標，存儲到當前工程目錄下的icon目錄（請事先創(chuàng)建icon目錄)
     download(entry.getValue(), "icon/" + entry.getKey() + ".png");
    }catch(Exception e) {
     System.out.println("【下載出錯】：" + entry.getKey());
     e.printStackTrace();
    }
   }
   System.out.println("【下載進度】：第" + i + "頁下載完成");
  }

 }

 /**
  * 獲取url網(wǎng)頁里的所有應(yīng)用的應(yīng)用名及其圖標網(wǎng)址
  * @param appPackageName
  * @return
  */
 private static HashMap</string><string , String> getAppImageUrl(String url) {

  HashMap</string><string , String> apps = new HashMap</string><string , String>();
  String appPackageName = "";
  String appImageUrl = "";
  String appName = "";

  String html = getHtmlByUrl(url);
  Document doc = Jsoup.parse(html);
     Elements elements = doc.select("div.container.clearfix>section.main-col>div.app-blocks>div.app-block>ul.app-list.clearfix>li.app>a.icon-area"); 
     Elements nameElements = doc.select("div.container.clearfix>section.main-col>div.app-blocks>div.app-block>ul.app-list.clearfix>li.app>div.operate>a.name>span.txt"); 
     Elements imageEle;

     int i = 0;
     for(Element ele : elements) {
      //獲取包名
      appPackageName = ele.attr("data-pn");
      //獲取圖標網(wǎng)址
      imageEle = ele.select("img.icon");
      appImageUrl = imageEle.get(0).attr("src").replace("68_68", "256_256");
      //加入apps
      apps.put(appPackageName, appImageUrl);
      //獲取app名稱
      appName = nameElements.get(i).text();
      //把app名稱和包名輸出到文件
      write2file("【" + appName + "】" + appPackageName);
      i++;
     }
     System.out.println("【下載進度】：" + url + "下的圖標網(wǎng)址已經(jīng)獲取成功");
  return apps;
 }

 /**
    * 下載文件到本地
    *
    * @param urlString
    *          被下載的文件地址
    * @param filename
    *          本地文件名
    * @throws Exception
    *           各種異常
    */
 private static void download(String urlString, String filename) throws Exception {
  System.out.println("【下載進度】：正在下載" + filename);
     // 構(gòu)造URL
     URL url = new URL(urlString);
     // 打開連接
     URLConnection con = url.openConnection();
     // 輸入流
     InputStream is = con.getInputStream();
     // 1K的數(shù)據(jù)緩沖
     byte[] bs = new byte[1024];
     // 讀取到的數(shù)據(jù)長度
     int len;
     // 輸出的文件流
     OutputStream os = new FileOutputStream(filename);
     // 開始讀取
     while ((len = is.read(bs)) != -1) {
       os.write(bs, 0, len);
     }
     // 完畢，關(guān)閉所有鏈接
     os.close();
     is.close();
     System.out.println("【下載進度】：" + filename + "下載成功");
 }  

 /**
  * 根據(jù)URL獲得所有的html信息
  * @param url
  * @return html
  */ 
 private static String getHtmlByUrl(String url){ 
     String html = null; 
     //創(chuàng)建httpClient對象
     HttpClient httpClient = new DefaultHttpClient(); 
     //以get方式請求該URL
     HttpGet httpget = new HttpGet(url); 
     try { 
         //得到responce對象
         HttpResponse responce = httpClient.execute(httpget);
         //返回碼
         int resStatu = responce.getStatusLine().getStatusCode(); 
         //200正常  其他就不對 
         if (resStatu==HttpStatus.SC_OK) {
             //獲得相應(yīng)實體 
             HttpEntity entity = responce.getEntity(); 
             if (entity!=null) { 
                 //獲得html源代碼
                 html = EntityUtils.toString(entity); 
             } 
         } 
     } catch (Exception e) { 
         System.out.println("訪問【"+url+"】出現(xiàn)異常!"); 
         e.printStackTrace(); 
     } finally { 
         httpClient.getConnectionManager().shutdown(); 
     } 
     return html; 
 }

 private static void write2file(String content) {

  File file = new File("icon/packageName.txt");
  BufferedWriter writer = null;
  try{
   if(!file.exists()) {
    file.createNewFile();
   }
   //參數(shù)true表示將輸出追加到文件內(nèi)容的末尾而不覆蓋原來的內(nèi)容
   writer = new BufferedWriter(new FileWriter(file, true));
   //輸出內(nèi)容
   writer.write(content);
   //換行
   writer.newLine();
  } catch(IOException e){
   System.out.println("輸出出錯");
   e.printStackTrace();
  } finally {
   if( writer != null) {
    try {
     writer.close();
    } catch(IOException e) {
     e.printStackTrace();
    }
   }
  }
 }

}