java批量采集Android应用图标和包名 java批量采集豌豆荚网站Android应用图标和包名

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;

/**
 * 获取豌豆荚网页上安卓软件全部软件
 * 注意：执行程序前，一定要在这个工程目录下创建icon文件夹
 * 所有图标将下载到icon这个文件夹中
 * 应用名称与包名写到了icon下的packageName.txt文件里
 *
 * 这个程序用到的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/全部软件/total?page=";
  //String rootUrl = "http://www.wandoujia.com/tag/全部游戏/total?page=";
  //下载1到20页的应用图标
  for(int i = 1; i < = 20; i++) {
   System.out.println("【下载进度】：准备下载第" + i + "页");
   String currentUrl = rootUrl + i;
   HashMap<String, String> apps = new HashMap<string , String>();
   apps = getAppImageUrl(currentUrl);
   //遍历HashMap逐个下载图标
   for(Entry</string><string , String> entry : apps.entrySet()) {
    try{
     //下载图标，存储到当前工程目录下的icon目录（请事先创建icon目录)
     download(entry.getValue(), "icon/" + entry.getKey() + ".png");
    }catch(Exception e) {
     System.out.println("【下载出错】：" + entry.getKey());
     e.printStackTrace();
    }
   }
   System.out.println("【下载进度】：第" + i + "页下载完成");
  }

 }

 /**
  * 获取url网页里的所有应用的应用名及其图标网址
  * @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) {
      //获取包名
      appPackageName = ele.attr("data-pn");
      //获取图标网址
      imageEle = ele.select("img.icon");
      appImageUrl = imageEle.get(0).attr("src").replace("68_68", "256_256");
      //加入apps
      apps.put(appPackageName, appImageUrl);
      //获取app名称
      appName = nameElements.get(i).text();
      //把app名称和包名输出到文件
      write2file("【" + appName + "】" + appPackageName);
      i++;
     }
     System.out.println("【下载进度】：" + url + "下的图标网址已经获取成功");
  return apps;
 }

 /**
    * 下载文件到本地
    *
    * @param urlString
    *          被下载的文件地址
    * @param filename
    *          本地文件名
    * @throws Exception
    *           各种异常
    */
 private static void download(String urlString, String filename) throws Exception {
  System.out.println("【下载进度】：正在下载" + filename);
     // 构造URL
     URL url = new URL(urlString);
     // 打开连接
     URLConnection con = url.openConnection();
     // 输入流
     InputStream is = con.getInputStream();
     // 1K的数据缓冲
     byte[] bs = new byte[1024];
     // 读取到的数据长度
     int len;
     // 输出的文件流
     OutputStream os = new FileOutputStream(filename);
     // 开始读取
     while ((len = is.read(bs)) != -1) {
       os.write(bs, 0, len);
     }
     // 完毕，关闭所有链接
     os.close();
     is.close();
     System.out.println("【下载进度】：" + filename + "下载成功");
 }  

 /**
  * 根据URL获得所有的html信息
  * @param url
  * @return html
  */ 
 private static String getHtmlByUrl(String url){ 
     String html = null; 
     //创建httpClient对象
     HttpClient httpClient = new DefaultHttpClient(); 
     //以get方式请求该URL
     HttpGet httpget = new HttpGet(url); 
     try { 
         //得到responce对象
         HttpResponse responce = httpClient.execute(httpget);
         //返回码
         int resStatu = responce.getStatusLine().getStatusCode(); 
         //200正常  其他就不对 
         if (resStatu==HttpStatus.SC_OK) {
             //获得相应实体 
             HttpEntity entity = responce.getEntity(); 
             if (entity!=null) { 
                 //获得html源代码
                 html = EntityUtils.toString(entity); 
             } 
         } 
     } catch (Exception e) { 
         System.out.println("访问【"+url+"】出现异常!"); 
         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();
   }
   //参数true表示将输出追加到文件内容的末尾而不覆盖原来的内容
   writer = new BufferedWriter(new FileWriter(file, true));
   //输出内容
   writer.write(content);
   //换行
   writer.newLine();
  } catch(IOException e){
   System.out.println("输出出错");
   e.printStackTrace();
  } finally {
   if( writer != null) {
    try {
     writer.close();
    } catch(IOException e) {
     e.printStackTrace();
    }
   }
  }
 }

}