WebMagic轻量级爬虫框架实战-根据关键词爬取某网站数据

java 复制代码

package com.tjp.www.xuexiangbanweb.webmagic;

import cn.hutool.core.io.FileUtil;
import com.alibaba.fastjson.JSONObject;
import com.tjp.www.xuexiangbanweb.entity.FHNewxVo;
import com.tjp.www.xuexiangbanweb.entity.FHResp;
import com.tjp.www.xuexiangbanweb.entity.NewsCollect;
import com.tjp.www.xuexiangbanweb.service.NewsCollectService;
import com.tjp.www.xuexiangbanweb.util.DownloadFile;
import com.tjp.www.xuexiangbanweb.util.HttpUtils;
import org.apache.commons.lang3.StringUtils;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;
import us.codecraft.webmagic.Request;
import us.codecraft.webmagic.Spider;

import java.util.ArrayList;
import java.util.List;
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.CopyOnWriteArrayList;


@Component
public class WebMagicStarterFenghuang {
    @Autowired
    NewsCollectService newsCollectService;
    @Autowired
    CustomerPipelinefh customerPipelinefh;
    @Autowired
    NewsPageProcessor2 newsPageProcessor2;

    public void doStartReptile() {
        String[] keywords=new String[]{"北海舰队","海军航空兵","南沙群岛","辽宁舰","海军人员","南海岛礁","中国海军","南海局势","中国领海","山东舰","军事强国","北海某海域","中国某海域",
                "黄岩岛","黄海某海域","南海某海域","战区海军","南海舰队","海军航空兵","执法船","中国海洋资源","中国轮船","海事部门","海洋强国","渔业捕捞","海洋战略","海上应急",
                "中国岛礁","渔政","海洋战略","水产品","船舶管理","海洋资源","中国轮船","海上执法","海洋部门","东海某海域","海上演练","海航","中国渔船","渔业政策","黄岩岛",
        "驱逐舰","舰艇","海上演习","钢铁海洋","渔政部门","渔民","战斗舰","海洋鱼类","海洋生态","海上打捞","禁渔期","海上船只","海洋政策","海上巡航","建设海洋生态","发展海洋","海洋经济",
        "船事故","海警部门","中国水产","台湾海峡","海军部队","沿海港航","沿海航行"};
        List<Request> list = new ArrayList<>();
        for (String keywordss : keywords) {
          //爬取目标网站url,
            String urls = "https://www.xxxxxx.com/xxxx/xxxxx/xxx/"+keywordss+"/1/xxxx?callback=getSoFengDataCallback&_=16502467818311";
            String s = HttpUtils.get(urls);
            String ss = s.replace("getSoFengDataCallback", "");
            String substring = ss.substring(1, ss.length() - 1);
            FHResp resp = JSONObject.parseObject(substring, FHResp.class);
            int totalPage = resp.getData().getTotalPage();
            for (int i = 1; i <=totalPage; i++) {
                String url = "https://shankapi.ifeng.com/season/getSoFengData/all/"+keywordss+"/"+i+"/getSoFengDataCallback?callback=getSoFengDataCallback&_=16502467818311";
                String sss = HttpUtils.get(url);
                String spjson = sss.replace("getSoFengDataCallback", "");
                String substrings = spjson.substring(1, spjson.length() - 1);

                FHResp fhResp = JSONObject.parseObject(substrings, FHResp.class);
                if(fhResp.getData().getItems().size()>0){
                    List<FHNewxVo> items = fhResp.getData().getItems();
                    for (FHNewxVo item : items) {
                        if(StringUtils.isNotBlank(item.getUrl())){
                            Map<String,Object> map=new ConcurrentHashMap<>();
                            map.put("keyword",keywordss);
                            String deatilUrl="https:"+item.getUrl();
                            Request request = new Request(deatilUrl);
                            request.setExtras(map);
                            list.add(request);

                        }
                    }
                }
            }
        }
        System.out.println("数量--》"+list.size());

//获取所有详情页URL,进行请求
        Request[] requests = list.toArray(new Request[list.size()]);
        Spider.create(newsPageProcessor2)
                //.addUrl(url)
                .addRequest(requests)
                //开启5个线程抓取
                .addPipeline(customerPipelinefh)
                .thread(10)
                //启动爬虫
                .run();
        System.out.println("读到--》"+CustomerPipelinefh.map.size());
        //System.out.println("读到--》"+CustomerPipelinefh.map.toString());
        /* 图片存放分文件夹 */
        String dowFilePath = "E:/images";
        if (!FileUtil.exist(dowFilePath)) {
            FileUtil.mkdir(dowFilePath);
        }
        //取数据 存数据库
        List<NewsCollect> lists=new CopyOnWriteArrayList<>();
        CustomerPipelinefh.map.forEach((k, v) -> {
            lists.add(v);
        });
        System.out.println("读取到的数据，去重前---》"+lists.size());
        int bs=0;
        int as=0;
        for (NewsCollect newsCollect : lists) {
            String s = newsCollect.getPublishtime().split("/")[0];
            if (StringUtils.isNotBlank(newsCollect.getAttachments())) {
                String jpg = newsCollect.getAttachments().substring(newsCollect.getAttachments().lastIndexOf("."));
                String timePre = newsCollect.getPublishtime();
                if (StringUtils.isNotBlank(newsCollect.getPublishtime())) {
                    String[] timeSpilt = newsCollect.getPublishtime().substring(0, 10).split("-");
                    System.out.println("时间---》"+timeSpilt.toString());
                    timePre = timeSpilt[0] + timeSpilt[1] + timeSpilt[2];
                }
                String jpgName = timePre + "_" +
                        newsCollect.getPublishpersoninfo() + "_" +
                        newsCollect.getTitle().replaceAll("\\s*", "") + "_" + newsCollect.getNewssource() + jpg;
                boolean b = DownloadFile.downloadA(newsCollect.getAttachments(), "E:\\images\\" + jpgName);
                if (b) {//成功了在设置
                    newsCollect.setAttachmentsurl("images/" + jpgName);
                    System.out.println("成功下载了---》" + (++bs));
                }
                System.out.println("下载了---》" + (++as));
            }

        }

        System.out.println("数据量--》"+lists.size());
      //  System.out.println("数据量内容--》"+lists.toString());
        //存数据库都是有效的
       newsCollectService.saveBatch(lists);
        CustomerPipelinefh.map.clear();
        System.out.println("清除map-->"+CustomerPipelinefh.map.size());
    }

}

Piepeline组件逻辑

java 复制代码

package com.tjp.www.xuexiangbanweb.webmagic;

import cn.hutool.core.util.ObjectUtil;
import com.tjp.www.xuexiangbanweb.entity.NewsCollect;
import com.tjp.www.xuexiangbanweb.mapper.NewsCollectMapper;
import org.apache.commons.lang3.StringUtils;
import org.springframework.stereotype.Component;
import us.codecraft.webmagic.ResultItems;
import us.codecraft.webmagic.Task;
import us.codecraft.webmagic.pipeline.Pipeline;

import javax.annotation.Resource;
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;

@Component
public class CustomerPipelinefh implements Pipeline {
    @Resource
    NewsCollectMapper newsCollectMapper;
    public static Map<String,NewsCollect> map=new ConcurrentHashMap<>();
    @Override
    public void process(ResultItems resultItems, Task task) {
            if (!ObjectUtil.isEmpty(resultItems.get("newsObject"))) {
                NewsCollect newsObject = resultItems.get("newsObject");
                if(StringUtils.isNotBlank(newsObject.getTitle())) {
                    map.put(newsObject.getTitle()+newsObject.getPublishtime(), newsObject);
                }
            }


    }


}

复制代码

PageProcessor组件逻辑

java 复制代码

package com.tjp.www.xuexiangbanweb.webmagic;

import cn.hutool.core.util.IdUtil;
import com.tjp.www.xuexiangbanweb.entity.NewsCollect;
import org.apache.commons.lang3.StringUtils;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;
import org.springframework.stereotype.Component;
import us.codecraft.webmagic.Page;
import us.codecraft.webmagic.Site;
import us.codecraft.webmagic.processor.PageProcessor;

@Component
public class NewsPageProcessor2 implements PageProcessor {

    @Override
    public void process(Page page) {// process是定制爬虫逻辑的核心接口，在这里编写抽取逻辑

            detailPages(page);

    }

    public  void detailPages(Page page){
        String keyword = page.getRequest().getExtra("keyword").toString();
        Document document = page.getHtml().getDocument();
        String url = page.getRequest().getUrl();
        Elements otherBody = document.getElementsByClass("atlas-4Dz3AOM8");
        String title = document.getElementsByClass("title-14yWv8ay").text();
        String person = document.getElementsByClass("source-3cecBclA").text();
        String time = document.getElementsByClass("time-M6w87NaQ").text();
        Elements p1 = document.getElementsByClass("smallFont-Z_OfA44W\n" +
                "                            text-20BABGxP\n" +
                "                            ").select("p");
        String text = document.getElementsByClass("videoArea-3bC47kpL").text();
        //不考虑视频 08/14 41:12    2021/18/01 17:52
        if(!StringUtils.isNotBlank(text)&&StringUtils.isNotBlank(time)&&StringUtils.isNotBlank(p1.toString())) {
            String[] splitTime = time.split("/");
            if(splitTime.length<3){
                time="2021/"+time;
            }
            if("2022".equals(time.split("/")[0])||"2021".equals(time.split("/")[0])) {
                String body = "";
                for (Element element : p1) {
                    body += element.text();
                }
                if (!StringUtils.isNotBlank(body)) {

                    for (Element element : otherBody) {
                        body += element.getElementsByClass("atlasBox-2CqWytIX").next().text();
                    }
                }
                if (body.length() > 32767) {
                    body = body.substring(0, 32767);
                }
                String imgUrl = "";
                Elements img = p1.select("img");
                if (!img.isEmpty()) {
                    for (Element element : img) {
                        String src = element.attr("src");
                        imgUrl = src;
                        break;
                    }
                }
                NewsCollect newsCollect = new NewsCollect();
                newsCollect.setId(IdUtil.getSnowflakeNextId());
                newsCollect.setTitle(title.replaceAll("\\s*", ""));
                newsCollect.setPublishpersoninfo(person);// 01/17 18:30
                newsCollect.setPublishtime(time.replaceAll("/","-"));
                newsCollect.setNewsinfo(body);
                newsCollect.setAttachments(imgUrl);
                newsCollect.setNewssource("凤凰网");
                newsCollect.setLinksnum("0");
                newsCollect.setForwardnum("0");
                newsCollect.setCommentnum("0");
                newsCollect.setEventtype("文章");
                newsCollect.setKeyword(keyword);
                newsCollect.setUrl(url);
                page.putField("newsObject", newsCollect);
            }
        }

    }

    @Override
    public Site getSite() {  // 部分一：抓取网站的相关配置，包括编码、抓取间隔、重试次数等
        return Site.me().setRetryTimes(10)
                .setCharset("utf-8").setTimeOut(6000)
                .setUserAgent("Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/100.0.4896.88 Safari/537.36");
    }




}