Java 多线程进阶-并发数据结构

Posted 2021-02-28 Sweet小马

tags:

篇首语：本文由小常识网(cha138.com)小编为大家整理，主要介绍了Java 多线程进阶-并发数据结构相关的知识，希望对你有一定的参考价值。

Java 多线程进阶-并发数据结构

并发数据结构

常用的数据结构是线程不安全的
- ArrayList/HashMap/HashSet 非同步的
- 多个线程同时独写, 可能会抛出异常或数据错误
传统Vector/HashTable等同步数据集合性能过差
并发数据结构: 数据添加或删除
- 阻塞式集合: 当集合为空或者满时, 等待
- 非阻塞式集合: 当集合为空或者满时, 不等待, 返回null或异常

List

Vector 同步安全, 写多读少, 效率差
ArrayList 不安全
通过基于 synchronized的Collections.synchronizedList(List list) 方法将 List 变成线程安全的, 效率差
CopyOnWriteArrayList (JDK5提供的基于复制机制的并发列表类), 非阻塞的, 适用于读多写少即填入数据之后大部分操作是读取和遍历的. 效率好, 适合在多线程下使用.

    package thread0418;

    import java.time.LocalDateTime;
    import java.util.ArrayList;
    import java.util.Collections;
    import java.util.List;
    import java.util.concurrent.CopyOnWriteArrayList;

    /**
     * 并发数据结构对比
     */
    public class ThreadListDemo1 {
        public static void main(String[] args) throws InterruptedException {
            // 线程不安全
            List<String> unsafeList = new ArrayList<String>();
            // 线程安全 将一个不安全的转成安全的
            List<String> safeList1 = Collections.synchronizedList(new ArrayList<String>());
            // 线程安全
            CopyOnWriteArrayList<String> safeList2 = new CopyOnWriteArrayList<String>();

            ListThread111 t1 = new ListThread111(unsafeList);
            ListThread111 t2 = new ListThread111(safeList1);
            ListThread111 t3 = new ListThread111(safeList2);

            // 分别启动十个线程, 运行测试
            for (int i = 0; i < 10; i++) {
                Thread t = new Thread(t1, String.valueOf(i));
                t.start();
            }
            for (int i = 0; i < 10; i++) {
                Thread t = new Thread(t2, String.valueOf(i));
                t.start();
            }
            for (int i = 0; i < 10; i++) {
                Thread t = new Thread(t3, String.valueOf(i));
                t.start();
            }

            // 等待子线程执行完
            Thread.sleep(2000);

            System.out.println(LocalDateTime.now() + " => " + t1.list.size());
            System.out.println(LocalDateTime.now() + " => " + t2.list.size());
            System.out.println(LocalDateTime.now() + " => " + t3.list.size());

            // 输出list中的值
            System.out.println(LocalDateTime.now() + " => " + "unsafeList:");
            for (String s : t1.list) {
                if (s == null) {
                    System.out.print("null ");
                } else {
                    System.out.print(s + " ");
                }
            }
            System.out.println();

            System.out.println(LocalDateTime.now() + " => " + "safeList1: ");
            for (String s : t2.list) {
                if (s == null) {
                    System.out.print("null ");
                } else {
                    System.out.print(s + " ");
                }
            }
            System.out.println();

            System.out.println(LocalDateTime.now() + " => " + "safeList2: ");
            for (String s : t3.list) {
                if (s == null) {
                    System.out.print("null ");
                } else {
                    System.out.print(s + " ");
                }
            }


        }
    }

    class ListThread111 implements Runnable {
        public List<String> list;

        public ListThread111(List<String> list) {
            this.list = list;
        }

        @Override
        public void run() {
            int i = 0;
            while (i < 10) {
                try {
                    Thread.sleep(10);
                } catch (Exception e) {
                    e.printStackTrace();
                }
                // 把当前线程名字加到 list 中
                list.add(Thread.currentThread().getName());
                i++;
            }
        }
    }

Set

HashSet 不安全
Collections.synchronizedSet(Set set) 基于 synchronized, 效率差
CopyOnWriteArraySet (JDK5提供的基于 CopyOnWriteArrayList 实现), 非阻塞的, 适用于读多写少，效率好, 适合在多线程下使用.

    package thread0418;

    import java.time.LocalDateTime;
    import java.util.Collections;
    import java.util.HashSet;
    import java.util.Set;
    import java.util.concurrent.CopyOnWriteArraySet;

    public class ThreadSetDemo1 {
        public static void main(String[] args) throws InterruptedException {
            // 线程不安全
            Set<String> unsafeSet = new HashSet<String>();
            // 线程安全
            Set<String> safeSet1 = Collections.synchronizedSet(new HashSet<>());
            // 线程安全
            CopyOnWriteArraySet<String> safeSet2 = new CopyOnWriteArraySet<>();

            SetThread111 t1 = new SetThread111(unsafeSet);
            SetThread111 t2 = new SetThread111(safeSet1);
            SetThread111 t3 = new SetThread111(safeSet2);

            // 分别启动十个线程, 运行测试
            for (int i = 0; i < 10; i++) {
                Thread t = new Thread(t1, String.valueOf(i));
                t.start();
            }
            for (int i = 0; i < 10; i++) {
                Thread t = new Thread(t2, String.valueOf(i));
                t.start();
            }
            for (int i = 0; i < 10; i++) {
                Thread t = new Thread(t3, String.valueOf(i));
                t.start();
            }

            // 等待子线程执行完成
            Thread.sleep(2000);


            System.out.println(LocalDateTime.now() + " => " + t1.set.size());
            System.out.println(LocalDateTime.now() + " => " + t2.set.size());
            System.out.println(LocalDateTime.now() + " => " + t3.set.size());

            // 输出 Set 中的值
            System.out.println(LocalDateTime.now() + " => " + "unsafeSet: ");
            for (String ele : t1.set) {
                if (ele == null) {
                    System.out.print("null ");
                } else {
                    System.out.print(ele + " ");
                }
            }
            System.out.println();
            System.out.println(LocalDateTime.now() + " => " + "safeSet1: ");
            for (String ele : t2.set) {
                if (ele == null) {
                    System.out.print("null ");
                } else {
                    System.out.print(ele + " ");
                }
            }
            System.out.println();
            System.out.println(LocalDateTime.now() + " => " + "safeSet2: ");
            for (String ele : t3.set) {
                if (ele == null) {
                    System.out.print("null ");
                } else {
                    System.out.print(ele + " ");
                }
            }
        }
    }

    class SetThread111 implements Runnable {
        public Set<String> set;

        public SetThread111(Set<String> set) {
            this.set = set;
        }

        @Override
        public void run() {
            int i = 0;
            while (i < 10) {
                try {
                    Thread.sleep(10);
                } catch (InterruptedException e) {
                    e.printStackTrace();
                }
                // 把当前线程名称加入 set里
                set.add(Thread.currentThread().getName() + i);
                i++;
            }
        }
    }

Map

Hashtable 同步安全, 写多读少, 效率差
HashMap 不安全
Collections.synchronizedMap(Map map) 基于 synchronized, 效率差
ConcurrentHashMap 读多写少, 非阻塞

    package thread0418;

    import java.time.LocalDateTime;
    import java.util.Collections;
    import java.util.HashMap;
    import java.util.Iterator;
    import java.util.Map;
    import java.util.concurrent.ConcurrentHashMap;

    public class ThreadMapDemo1 {
        public static void main(String[] args) throws InterruptedException {
            // 线程不安全
            MapThread111 t1 = new MapThread111(new HashMap<>());
            // 线程安全
            MapThread111 t2 = new MapThread111(Collections.synchronizedMap(new HashMap<>()));
            // 线程安全
            MapThread111 t3 = new MapThread111(new ConcurrentHashMap<>());

            // 分别启动十个线程, 运行测试
            for (int i = 0; i < 10; i++) {
                Thread t = new Thread(t1);
                t.start();
            }
            for (int i = 0; i < 10; i++) {
                Thread t = new Thread(t2);
                t.start();
            }
            for (int i = 0; i < 10; i++) {
                Thread t = new Thread(t3);
                t.start();
            }

            // 等待子线程执行完
            Thread.sleep(2000);
            System.out.println(LocalDateTime.now() + " => " + t1.map.size());
            System.out.println(LocalDateTime.now() + " => " + t2.map.size());
            System.out.println(LocalDateTime.now() + " => " + t3.map.size());

            // 输出 map 中的值
            System.out.println(LocalDateTime.now() + " => " + "unsafeMap:");
            Iterator<Map.Entry<Integer, String>> iterator = t1.map.entrySet().iterator();
            while (iterator.hasNext()) {
                Map.Entry<Integer, String> entry = iterator.next();
                System.out.print(entry.getKey() + ":" + entry.getValue() + "  ");
            }
            System.out.println();

            System.out.println(LocalDateTime.now() + " => " + "safeMap:");
            iterator = t2.map.entrySet().iterator();
            while (iterator.hasNext()) {
                Map.Entry<Integer, String> entry = iterator.next();
                System.out.print(entry.getKey() + ":" + entry.getValue() + "  ");
            }
            System.out.println();

            System.out.println(LocalDateTime.now() + " => " + "safeMap2:");
            iterator = t3.map.entrySet().iterator();
            while (iterator.hasNext()) {
                Map.Entry<Integer, String> entry = iterator.next();
                System.out.print(entry.getKey() + ":" + entry.getValue() + "  ");
            }
            System.out.println();
            System.out.println(LocalDateTime.now() + " => " + "mapThread1.map.size()" + t1.map.size());
            System.out.println(LocalDateTime.now() + " => " + "mapThread2.map.size()" + t2.map.size());
            System.out.println(LocalDateTime.now() + " => " + "mapThread3.map.size()" + t3.map.size());


        }
    }

    class MapThread111 implements Runnable {
        public Map<Integer, String> map;

        public MapThread111(Map<Integer, String> map) {
            this.map = map;
        }

        @Override
        public void run() {
            int i = 0;
            while (i < 100) {
                try {
                    Thread.sleep(10);
                } catch (InterruptedException e) {
                    e.printStackTrace();
                }
                // 把当前线程名称放入map中
                map.put(++i, Thread.currentThread().getName());
            }
        }
    }

Queue(单向队列) & Deque(双向队列)(JDK1.5)

ConcurrentLinkedQueue 非阻塞
ArrayBlockingQueue/LinkedBlockingQueue 阻塞

    package thread0418;

    import java.time.LocalDateTime;
    import java.util.ArrayDeque;
    import java.util.Queue;
    import java.util.concurrent.ArrayBlockingQueue;
    import java.util.concurrent.ConcurrentLinkedDeque;

    public class ThreadQueueDemo1 {
        public static void main(String[] args) throws InterruptedException {
            // 线程不安全
            QueueThread111 t1 = new QueueThread111(new ArrayDeque<>());
            // 线程安全
            QueueThread111 t2 = new QueueThread111(new ConcurrentLinkedDeque<>());
            QueueThread111 t3 = new QueueThread111(new ArrayBlockingQueue<>(100));

            for (int i = 0; i < 10; i++) {
                Thread thread = new Thread(t1, String.valueOf(i));
                thread.start();
            }
            for (int i = 0; i < 10; i++) {
                Thread thread = new Thread(t2, String.valueOf(i));
                thread.start();
            }
            for (int i = 0; i < 10; i++) {
                Thread thread = new Thread(t3, String.valueOf(i));
                thread.start();
            }

            // 等待子线程执行完
            Thread.sleep(2000);

            System.out.println(LocalDateTime.now() + " => " + t1.queue.size());
            System.out.println(LocalDateTime.now() + " => " + t2.queue.size());
            System.out.println(LocalDateTime.now() + " => " + t3.queue.size());

            // 输出 Queue 中的值
            System.out.println("unsafeQueue: ");
            for (String s : t1.queue) {
                System.out.print(s + " ");
            }
            System.out.println();
            System.out.println("safeQueue1: ");
            for (String s : t2.queue) {
                System.out.print(s + " ");
            }
            System.out.println();
            System.out.println("safeQueue2: ");
            for (String s : t3.queue) {
                System.out.print(s + " ");
            }
        }
    }

    class QueueThread111 implements Runnable {
        public Queue<String> queue;

        public QueueThread111(Queue<String> queue) {
            this.queue = queue;
        }

        @Override
        public void run() {
            int i = 0;
            while (i < 10) {
                i++;
                try {
                    Thread.sleep(10);
                } catch (InterruptedException e) {
                    e.printStackTrace();
                }
                // 把当前线程名称加入list中
                queue.add(Thread.currentThread().getName());
            }
        }
    }

总结
- 了解数据结构并发独写的问题
- 根据业务特点, 使用正确的并发数据结构

以上是关于Java 多线程进阶-并发数据结构的主要内容，如果未能解决你的问题，请参考以下文章

Java 多线程进阶-并发协作控制

Java 多线程进阶-并发编程线程组ThreadGroup

Java进阶 - 并发(多线程基本概念)

Java进阶 - 并发(多线程锁类)

Java进阶 - 并发(关键字)

Java知识体系Java并发编程进阶，多线程和锁底层原理探究