例子

 Directory index = SimpleFSDirectory.open(new File("d:/20101015index"));

package us.codecraft.webmagic.samples;

import java.util.List;


import us.codecraft.webmagic.Page;
import us.codecraft.webmagic.Site;
import us.codecraft.webmagic.Spider;
import us.codecraft.webmagic.processor.PageProcessor;
import us.codecraft.webmagic.selector.Html;


/**
 * @author code4crafter@gmail.com <br>
 */
public class SinaBlogProcessor implements PageProcessor {


    public static final String URL_LIST = "http://blog\\.sina\\.com\\.cn/s/articlelist_1487828712_0_\\d+\\.html";


    public static final String URL_POST = "http://blog\\.sina\\.com\\.cn/s/blog_\\w+\\.html";


    private Site site = Site
            .me()
            .setDomain("blog.sina.com.cn")
            .setSleepTime(3000)
            .setUserAgent(
                    "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_7_2) AppleWebKit/537.31 (KHTML, like Gecko) Chrome/26.0.1410.65 Safari/537.31");


    @Override
    public void process(Page page) {
   
    System.out.println(" html: "+page.getUrl());
//if (page.getUrl().toString().indexOf("mainpage.html") >= 0)
    if(false){

List<String> elems = page.getHtml().xpath("//div[@class=\"boutlist\"]/ul/li")
.all(); 

for (String elem : elems)
{
//System.out.println(elem);
}



//List<String> elems = page.getHtml().xpath("//div[@class=\"ss1\"]").all(); // links().all();
int i = 0;
for (String elem : elems) {
// System.out.println(elem);
Html pg = new Html(elem);


System.out.println("\r\n\r\n");
Html td0 = new Html(pg.xpath("//div[@class=\"tdtitle\"]").all()
.get(0));


String url = td0.xpath("//a/@href").toString();
String candy = td0.xpath("//a/@title").toString();
String title = td0.xpath("//a/text()").toString();


String ename = td0.xpath("//div[@class=\"tdtitle\"]/text()")
.toString();
ename = ename.substring(2);


// System.out.println(elem);
Html td1 = new Html(pg.xpath("//div[@class=\"tdtitle\"]").all()
.get(1));
String info = td1.xpath("//div[@class=\"tdtitle\"]/text()")
.toString();


String picurl = pg.xpath("//div[@class=\"tdimg\"]/a/img/@src")
.toString();

String dbfen = pg.xpath("//div[@class=\"ss5\"]/p/font/text()").all().get(0).toString();
String imdbfen = pg.xpath("//div[@class=\"ss5\"]/p/font/text()").all().get(1).toString();


System.out.println(i + ":" + title + " " + ename + " " + candy
+ " " + url + " " + info + " " + picurl+ " "+dbfen+" "+imdbfen);
i++;
}
}else
{
String title = page.getHtml().xpath("//div[@class=\"info\"]/h1/text()").all().get(0).toString();
String info0 = page.getHtml().xpath("//div[@class=\"info\"]/ul/li").all().get(0).toString();
String info1 = page.getHtml().xpath("//div[@class=\"info\"]/ul/li").all().get(1).toString();
String info2 = page.getHtml().xpath("//div[@class=\"info\"]/ul/li").all().get(2).toString();
String info3 = page.getHtml().xpath("//div[@class=\"info\"]/ul/li").all().get(3).toString();
String imgurl = page.getHtml().xpath("//div[@class=\"pic\"]/img/@src").all().get(0).toString();

String juqing = page.getHtml().xpath("//div[@id=\"endtext\"]/text()").all().get(0).toString();
//关于容错
List<String>  imgs = page.getHtml().xpath("//div[@id=\"endtext\"]/img/@src").all();

System.out.println(" "+title+", "+info0+", "+info1+", "+info2+", "+imgurl+" \r\n: "+juqing);

for(String img:imgs)
{
System.out.println(" "+img);
}

}
    /*
        //列表页
        if (page.getUrl().regex(URL_LIST).match()) {
            page.addTargetRequests(page.getHtml().xpath("//div[@class=\"articleList\"]").links().regex(URL_POST).all());
            page.addTargetRequests(page.getHtml().links().regex(URL_LIST).all());
            //文章页
        } else {
            page.putField("title", page.getHtml().xpath("//div[@class='articalTitle']/h2"));
            page.putField("content", page.getHtml().xpath("//div[@id='articlebody']//div[@class='articalContent']"));
            page.putField("date",
                    page.getHtml().xpath("//div[@id='articlebody']//span[@class='time SG_txtc']").regex("\\((.*)\\)"));
        }
        */
    }


    @Override
    public Site getSite() {
        return site;
    }


    public static void main(String[] args) {
        //Spider.create(new SinaBlogProcessor()).addUrl("http://127.0.0.1:8080/test/mainpage.html").run();
    Spider.create(new SinaBlogProcessor()).addUrl("http://127.0.0.1:8080/test/page.html").run();
    }

}

package jsoup;


import java.io.File;
import java.io.FileOutputStream;
import java.io.IOException;
import java.io.InputStream;
import java.net.URL;
import java.net.URLConnection;






public class test1 {


public static void main(String[] args) throws IOException {

URL   url   =   new   URL( "http://localhost:8080/test/mainpage_files/p2375744186.jpg"); 
URLConnection   uc   =   url.openConnection(); 
InputStream   is   =   uc.getInputStream(); 
File   file   =   new   File("d://1.jpg"); 
FileOutputStream   out   =   new   FileOutputStream(file); 
int   i=0; 
while   ((i=is.read())!=-1)   { 
out.write(i); 
} 
is.close();
System.out.println("ok");
}
}

<ul class="list-group">
  <li class=" list-group-item list-group-item-success"><span class="badge">14</span>


  <div class="rowmain" >
  <tbody>
  <tr>
    <td class="tdimg">
        <img class="listimg" src="../static/images/p2371145119.jpg" style="display: inline;">
        </img>
    </td>
    <td >
        <div class="p12" style="display: inline;">
        <tr>
        <a style="display: inline;">七月与安生</a>
        </tr>
        <tr>
        <p >asfafd</p></tr>
        </div>
  
    </td>
    </tr></tbody>
  <div>


  </li>
</ul>