Java线程池ExecutorService时让主线程等待子线程完成后继续处理与不用线程处理对比

桃扇骨 2022-01-27 22:51 386阅读 0赞

更多知识请访问  www.itkc8.com

用线程代码

package com.cowboy.service;
    
    import java.util.concurrent.*;
    
    /**
     * @ClassName CommonThreadPool
     * @Description TODO
     * @Author hux
     * @Date 2019/5/22、15:40
     * @Version 1.0
     **/
    
    public class CommonThreadPool {
    
        private static ExecutorService exec = new ThreadPoolExecutor(50, 100, 0L,
                TimeUnit.MILLISECONDS, new LinkedBlockingQueue<Runnable>(10000),
                new ThreadPoolExecutor.CallerRunsPolicy());
    
        public static void execute(Runnable command) {
            exec.execute(command);
        }
    
        /**
         * 子线程执行结束future.get()返回null，若没有执行完毕，主线程将会阻塞等待
         * @param command
         * @return
         */
        public static Future submit(Runnable command) {
            return exec.submit(command);
        }
    
        /**
         * 子线程中的返回值可以从返回的future中获取：future.get();
         * @param command
         * @return
         */
        public static Future submit(Callable command) {
            return exec.submit(command);
        }
    
        public static void shutdown(){
            exec.shutdown();
        }
    
    }

package com.cowboy.service;
    
    import com.cowboy.model.Article;
    import org.jsoup.Connection;
    import org.jsoup.Jsoup;
    import org.jsoup.nodes.Document;
    import org.jsoup.nodes.Element;
    import org.jsoup.select.Elements;
    
    import java.io.IOException;
    import java.util.ArrayList;
    import java.util.Iterator;
    import java.util.List;
    import java.util.concurrent.Callable;
    import java.util.concurrent.ExecutionException;
    import java.util.concurrent.Future;
    
    /**
     * @ClassName ThreadTest
     * @Description TODO
     * @Author hux
     * @Date 2019/5/22、15:41
     * @Version 1.0
     **/
    
    public class ThreadTest {
    
        private static final String URL = "https://blog.csdn.net/valada";//"https://blog.csdn.net/foruok";
    
    
        public static void main(String[] args) {
    
            int pageNow = 1;
            int totalPage = getTotalPage();
            System.out.println("总页数："+totalPage);
            long l1 = System.currentTimeMillis();
            List<Future> futureList = new ArrayList<>();
            for(pageNow = 1; pageNow <= totalPage; pageNow++) {
                int finalI = pageNow;
                Callable<List<Article>> task = () -> {
                    List<Article> artitcleByPage = getArtitcleByPage(finalI);
                    return artitcleByPage;
                };
                Future submit = CommonThreadPool.submit(task);
                futureList.add(submit);
            }
            //主线程处理其他工作,让子线程异步去执行.
            System.out.println("now waiting sub thread done.");
            //主线程其他工作完毕,等待子线程的结束, 调用future.get()系列的方法即可。
            List<Article> articleList = new ArrayList<>();
            try {
                for (Future future : futureList) {
                    List<Article> list = (List<Article>) future.get();
                    articleList.addAll(list);
                }
            } catch (InterruptedException | ExecutionException e) {
                e.printStackTrace();
            }
            System.out.println(System.currentTimeMillis() - l1);
            System.out.println(articleList.size() + "   "+articleList);
            CommonThreadPool.shutdown();
    
            //遍历输出博主所有的文章
            for(Article article : articleList) {
            	System.out.println("文章标题:" + article.getTitle().replaceFirst("原",""));
            	System.out.println("文章绝对路劲地址:" + article.getAddress());
            	System.out.println("文章简介:" + article.getDesption());
            	System.out.println("发表时间:" + article.getTime());
            	System.out.println("阅读数量:" + article.getReadNum());
            	System.out.println("评论数量:" + article.getCommentNum());
            }
        }
    
        /**
         * 获取总页数
         * @return
         */
        public static int getTotalPage(){
            Connection conn = Jsoup.connect(URL)
                    .userAgent("Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:50.0) Gecko/20100101 Firefox/50.0")
                    .timeout(8000)
                    .method(Connection.Method.GET);
            Document doc = null;
            try {
                doc = conn.get();
            } catch (IOException e) {
                e.printStackTrace();
            }
            Element body = doc.body();
            int totalPage = 1;
            Elements scripts = body.getElementsByTag("script");
            Iterator it = scripts.iterator();
            while(it.hasNext()) {
                Element element = (Element)it.next();
                String text = element.data();
                if (text.contains("pageSize") && text.contains("listTotal")) {
                    int i = text.indexOf("var pageSize = ");
                    int i1 = text.indexOf("var listTotal = ");
                    int i2 = text.indexOf("var pageQueryStr =");
                    String pageSize = text.substring(i+15, i1);
                    String listTotal = text.substring(i1+16, i2);
                    double i3 = Double.parseDouble(pageSize.replace(";", "").trim());
                    double i4 = Double.parseDouble(listTotal.replace(";", "").trim());
                    double number = i4 / i3;
                    totalPage = (int)Math.ceil(number);
                    break;
                }
            }
            return totalPage;
        }
    
        /**
         * 分页读取
         * @param pageNow
         * @return
         * @throws IOException
         */
        public static List<Article> getArtitcleByPage(int pageNow) throws IOException {
            Connection conn = Jsoup.connect(URL + "/article/list/" + pageNow)
                    .userAgent("Mozilla/5.0 (Windows NT 6.1; rv:47.0) Gecko/20100101 Firefox/47.")
                    .timeout(8000)
                    .method(Connection.Method.GET);
            Document doc = conn.get();
            Element body = doc.body();
            List<Article> resultList = new ArrayList<>();
            Elements articleList = body.getElementsByClass("article-item-box");
            for(Element article : articleList){
                Article articleEntity = new Article();
                Element linkNode = (article.select("div h4 a")).get(0);
                Element desptionNode = (article.getElementsByClass("content")).get(0);
                Element articleManageNode = (article.getElementsByClass("info-box")).get(0);
                articleEntity.setAddress(linkNode.attr("href"));
                articleEntity.setTitle(linkNode.text());
                articleEntity.setDesption(desptionNode.text());
                articleEntity.setTime(articleManageNode.getElementsByClass("date").text());
                articleEntity.setReadNum(articleManageNode.getElementsByClass("read-num").get(0).getElementsByClass("num").text());
                articleEntity.setCommentNum(articleManageNode.getElementsByClass("read-num").get(1).getElementsByClass("num").text());
                resultList.add(articleEntity);
            }
            return resultList;
        }
    
    
    }

结果：

处理时间：2946  
处理数量：1989

不用线程

package com.cowboy.service;
    
    import java.io.IOException;
    import java.util.ArrayList;
    import java.util.Iterator;
    import java.util.List;
    import java.util.concurrent.*;
    
    import com.cowboy.model.Article;
    import org.jsoup.*;
    import org.jsoup.nodes.*;
    import org.jsoup.select.*;
    
    
    /**
     * @author shizongger
     * @date 2017/02/09
     */
    public class Main {
    	
        private static final String URL = "https://blog.csdn.net/valada";
    
        //private final static Executor executor = Executors.newCachedThreadPool();//启用多线程
    
        public static void main(String[] args) throws IOException {
            Connection conn = Jsoup.connect(URL)
                    .userAgent("Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:50.0) Gecko/20100101 Firefox/50.0")
                    .timeout(8000)
                    .method(Connection.Method.GET);
            Document doc = conn.get();
            Element body = doc.body();
            int totalPage = 1;
            int pageNow = 1;
            Elements scripts = body.getElementsByTag("script");
            Iterator it = scripts.iterator();
            while(it.hasNext()) {
                Element element = (Element)it.next();
                String text = element.data();
                if (text.contains("pageSize") && text.contains("listTotal")) {
                    int i = text.indexOf("var pageSize = ");
                    int i1 = text.indexOf("var listTotal = ");
                    int i2 = text.indexOf("var pageQueryStr =");
                    String pageSize = text.substring(i+15, i1);
                    String listTotal = text.substring(i1+16, i2);
                    double i3 = Double.parseDouble(pageSize.replace(";", "").trim());
                    double i4 = Double.parseDouble(listTotal.replace(";", "").trim());
                    double number = i4 / i3;
                    totalPage = (int)Math.ceil(number);
                    break;
                }
            }
            System.out.println("总页数："+totalPage);
    
            List<Article> articleList = new ArrayList<>();
            long l1 = System.currentTimeMillis();
            for(pageNow = 1; pageNow <= totalPage; pageNow++){
            	articleList.addAll(getArtitcleByPage(pageNow));
            }
            System.out.println(System.currentTimeMillis() - l1);
    
            System.out.println(articleList.size() + "   "+articleList);
            //遍历输出博主所有的文章
            /*for(Article article : articleList) {
            	System.out.println("文章标题:" + article.getTitle().replaceFirst("原",""));
            	System.out.println("文章绝对路劲地址:" + article.getAddress());
            	System.out.println("文章简介:" + article.getDesption());
            	System.out.println("发表时间:" + article.getTime());
            	System.out.println("阅读数量:" + article.getReadNum());
            	System.out.println("评论数量:" + article.getCommentNum());
    
            }
            System.out.println("总文章数量："+articleList.size());*/
        }
    
    
    
        public static List<Article> getArtitcleByPage(int pageNow) throws IOException{
            Connection conn = Jsoup.connect(URL + "/article/list/" + pageNow)
                    .userAgent("Mozilla/5.0 (Windows NT 6.1; rv:47.0) Gecko/20100101 Firefox/47.")
                    .timeout(8000)
                    .method(Connection.Method.GET);
            Document doc = conn.get();
            Element body = doc.body();
            List<Article> resultList = new ArrayList<>();
            Elements articleList = body.getElementsByClass("article-item-box");
            for(Element article : articleList){
            	Article articleEntity = new Article();
                Element linkNode = (article.select("div h4 a")).get(0);
                Element desptionNode = (article.getElementsByClass("content")).get(0);
                Element articleManageNode = (article.getElementsByClass("info-box")).get(0);
                articleEntity.setAddress(linkNode.attr("href"));
                articleEntity.setTitle(linkNode.text());
                articleEntity.setDesption(desptionNode.text());
                articleEntity.setTime(articleManageNode.getElementsByClass("date").text());
                articleEntity.setReadNum(articleManageNode.getElementsByClass("read-num").get(0).getElementsByClass("num").text());
                articleEntity.setCommentNum(articleManageNode.getElementsByClass("read-num").get(1).getElementsByClass("num").text());
                resultList.add(articleEntity);
            }
            return resultList;
        }
    }

结果

处理时间：30073  
处理数量：1989

30073/2946  算算效率提升了多少倍

# **更多多线程知识访问  [www.itkc8.com][]** #

[www.itkc8.com]: http://www.itkc8.com