package us.codecraft.webmagic.model.samples;
import us.codecraft.webmagic.Site;
import us.codecraft.webmagic.model.ConsolePageModelPipeline;
import us.codecraft.webmagic.model.OOSpider;
import us.codecraft.webmagic.model.annotation.ConfigInfo;
import us.codecraft.webmagic.model.annotation.ExtractBy;
import us.codecraft.webmagic.model.annotation.ExtractByUrl;
import us.codecraft.webmagic.model.annotation.HelpUrl;
import us.codecraft.webmagic.model.annotation.TargetUrl;
/**
* @author code4crafter@gmail.com <br>
* Date: 13-8-11 <br>
* Time: 下午9:29 <br>
*/
@TargetUrl("http://www.36kr.com/p/\\d+.html")
@HelpUrl("http://www.36kr.com/#/page/\\d+")
public class Kr36NewsModel {
@ExtractBy("//h1[@class='entry-title sep10']")
private String title;
@ExtractBy(value="//div[@class='mainContent sep-10']",configure=@ConfigInfo(isRemoveTag=true))
private String content;
@ExtractByUrl
private String url;
public static void main(String[] args) {
OOSpider.create(Site.me().addStartUrl("http://www.36kr.com/"), new ConsolePageModelPipeline(),
Kr36NewsModel.class).run();
}
}