How to write the implementation code of java collection

1. HashMap

public class HashMapDemo {
    private Map map = null;
    public void init() {
        map = new HashMap();
        map.put("a", "aaa");
        map.put("b", "bbb");
        map.put("c", "ccc");
        System.out.println(map);
    }
    // 添加元素
    public void testPut() {
        // V put(K key, V value) ：把指定的key和value添加到集合中
        map.put("a1", "aaa");
        map.put("b1", "bbb");
        map.put("c1", "ccc");
        System.out.println(map);
        // void putAll(Map<? extends K,? extends V>m) ：把指定集合添加集合中
        Map map1 = new HashMap();
        map1.put("e", "eee");
        map1.put("f", "fff");
        map.putAll(map1);
        System.out.println(map);
        // default V putIfAbsent(K key, V value) ：如果key不存在就添加
        map.putIfAbsent("a", "hello");
        System.out.println(map);
        map.putIfAbsent("g", "ggg");
        System.out.println(map);
    }
    // 修改元素
    public void testModify() {
        // V put(K key, V value) ：把集合中指定key的值修改为指定的值
        map.put("a", "hello");
        map.put("a", "world");
        System.out.println(map);
        // 说明，当key相同时，后面的值会覆盖前面的值。
        // default V replace(K key, V value) ：根据key来替换值，而不做增加操作
        Object replace = map.replace("b1", "java");
        System.out.println(replace);
        System.out.println(map);
        //default boolean replace(K key, V oldValue,V newValue)
    }
    // 删除元素
    public void testRemove() {
        // V remove(Object key) ：根据指定key删除集合中对应的值
        Object c = map.remove("c");
        System.out.println(c);
        System.out.println(map);
        // default boolean remove(Object key, Objectvalue) ：根据key和value进行删除
        map.remove("b", "bbb1");
        System.out.println(map);
        // void clear() ：清空集合中所有元素
        map.clear();
        System.out.println(map);
    }
    // 判断元素
    public void testJudge() {
        // boolean isEmpty() ：判断集合是否为空，如果是返回true，否则返回false
        System.out.println(map.isEmpty());
        // boolean containsKey(Object key) ：判断集合中是否包含指定的key，包含返回true，否则返回false
        boolean flag = map.containsKey("a");
        System.out.println(flag); // true
        flag = map.containsKey("a1");
        System.out.println(flag); // false
        // boolean containsValue(Object value) ：判断集合中是否包含指定的value，包含返回true，否则返回false
        flag = map.containsValue("aaa");
        System.out.println(flag); // true
        flag = map.containsValue("aaa1");
        System.out.println(flag); // false
    }
    // 获取元素
    public void testGet() {
        // int size() ：返回集合的元素个数
        int size = map.size();
        System.out.println(size);
        // V get(Object key) ：根据Key获取值，如果找到就返回对应的值，否则返回null
        Object val = map.get("a");
        System.out.println(val);
        val = map.get("a1");
        System.out.println(val); // null
        // default V getOrDefault(Object key, VdefaultValue) ：根据Key获取值，如果key不存在，则返回默认值
        val = map.getOrDefault("a1", "hello");
        System.out.println(val);
        // Collection<V> values() ：返回集合中所有的Value
        Collection values = map.values();
        for (Object value : values) {
            System.out.println(value);
        }
        // Set<K> keySet() ：返回集合中所有的Key
        Set set = map.keySet();
        for (Object o : set) {
            System.out.println(o);
        }
    }
    // 迭代元素
    public void testIterator() {
        // 第一种：通过key获取值的方式
        Set keySet = map.keySet();
        Iterator it = keySet.iterator();
        while (it.hasNext()) {
            Object key = it.next();
            Object val = map.get(key);
            System.out.println(key + "=" + val);
        }
        System.out.println("------------------------ ");
        // 第二种：使用for循环
        for (Object key : map.keySet()) {
            System.out.println(key + "=" +
                               map.get(key));
        }
        System.out.println("------------------------ ");
        // 第三种：使用Map接口中的内部类来完成，在框架中大量使用
        Set entrySet = map.entrySet();
        for (Object obj : entrySet) {
            Map.Entry entry = (Map.Entry) obj;
            System.out.println(entry.getKey() + "=" +
                               entry.getValue());
        }
    }
}

Note: In HashMap, the key-value is allowed to be empty, but the key is unique and the value can be repeated. hashMap is not thread-safe.

2. TreeMap

is an ordered collection, and the natural sorting method is used by default.

public class Person implements Comparable {
    private String name;
    private int age;
    @Override
    public int compareTo(Object o) {
    if (o instanceof Person) {
    Person p = (Person) o;
    return this.age - p.age;
    }
    return 0;
}
    public Person() {}
    public Person(String name, int age) {
        this.name = name;
        this.age = age;
    }
    public int getAge() {
        return age;
    }
    public void setAge(int age) {
        this.age = age;
    }
    @Override
    public String toString() {
        return "Person{" +
            "name='" + name + '\'' +
            ", age=" + age +
            '}';
    }
}

Test

public class TeeMapDemo {
    @Test
    public void testInteger() {
        TreeMap tm = new TreeMap();
        tm.put(3, 333);
        tm.put(2, 222);
        tm.put(11, 111);
        tm.put(2, 222);
        System.out.println(tm);
    }
    @Test
    public void testString() {
        TreeMap tm = new TreeMap();
        tm.put("hello", "hello");
        tm.put("world", "world");
        tm.put("about", "");
        tm.put("abstract", "");
        System.out.println(tm);
    }
    @Test
    public void testPerson() {
        TreeMap tm = new TreeMap(new Comparator(){
            @Override
            public int compare(Object o1, Object o2) {
                if (o1 instanceof Person && o2
                    instanceof Person) {
                    Person p1 = (Person) o1;
                    Person p2 = (Person) o2;
                    return p1.getAge() - p2.getAge();
                }
                return 0;
            }
        });
        tm.put(new Person("张三",18), null);
        tm.put(new Person("李四",17), null);
        System.out.println(tm);
    }
}

Description: From the above code, we can find that the use of TreeMap is very similar to the use of TreeSet. Observing the source code of the HashSet collection, we can see that when creating a HashSet collection, In fact, the bottom layer uses HashMap.

public HashSet() {
	map = new HashMap<>();
}

HashSet actually stores the Key of HashMap.

3.ConcurrentHashMap

In the Map collection we introduced HashMap, TreeMap, these collections are not thread-safe in the case of multi-threading , so thread safety issues may arise.

In Java, Hashtable is a thread-safe HashMap. There is no difference between Hashtable and HashMap in terms of methods. synchronized is used to achieve thread safety. We observe the source code of Hashtable.

public synchronized V get(Object key) {
        Entry<?,?> tab[] = table;
        int hash = key.hashCode();
        int index = (hash & 0x7FFFFFFF) % tab.length;
        for (Entry<?,?> e = tab[index] ; e != null ; e = e.next) {
            if ((e.hash == hash) && e.key.equals(key)) {
                return (V)e.value;
            }
        }
        return null;
    }

The above is the get source code of Hashtable. It can be seen that it only adds a lock to the method, which greatly reduces the execution efficiency of the thread and achieves the goal at the expense of efficiency. This is obviously not what we are doing. What we want in practice, so we need a method that is both thread-safe and efficient.

ConcurrentHashMap uses the principle of segmented lock. Let’s observe the source code.

 public V put(K key, V value) {
        return putVal(key, value, false);
    }
final V putVal(K key, V value, boolean onlyIfAbsent) {
        if (key == null || value == null) throw new NullPointerException();
        int hash = spread(key.hashCode());
        int binCount = 0;
        for (Node<K,V>[] tab = table;;) {
            Node<K,V> f; int n, i, fh;
            if (tab == null || (n = tab.length) == 0)
                tab = initTable();
            else if ((f = tabAt(tab, i = (n - 1) & hash)) == null) {
                if (casTabAt(tab, i, null,
                             new Node<K,V>(hash, key, value, null)))
                    break;                   // no lock when adding to empty bin
            }
            else if ((fh = f.hash) == MOVED)
                tab = helpTransfer(tab, f);
            else {
                V oldVal = null;
                synchronized (f) {
                    if (tabAt(tab, i) == f) {
                        if (fh >= 0) {
                            binCount = 1;
                            for (Node<K,V> e = f;; ++binCount) {
                                K ek;
                                if (e.hash == hash &&
                                    ((ek = e.key) == key ||
                                     (ek != null && key.equals(ek)))) {
                                    oldVal = e.val;
                                    if (!onlyIfAbsent)
                                        e.val = value;
                                    break;
                                }
                                Node<K,V> pred = e;
                                if ((e = e.next) == null) {
                                    pred.next = new Node<K,V>(hash, key,
                                                              value, null);
                                    break;
                                }
                            }
                        }
                        else if (f instanceof TreeBin) {
                            Node<K,V> p;
                            binCount = 2;
                            if ((p = ((TreeBin<K,V>)f).putTreeVal(hash, key,
                                                           value)) != null) {
                                oldVal = p.val;
                                if (!onlyIfAbsent)
                                    p.val = value;
                            }
                        }
                    }
                }
                if (binCount != 0) {
                    if (binCount >= TREEIFY_THRESHOLD)
                        treeifyBin(tab, i);
                    if (oldVal != null)
                        return oldVal;
                    break;
                }
            }
        }
        addCount(1L, binCount);
        return null;
    }

It can be seen from the source code that ConcurrentHashMap only adds a lock when a thread operates the current data, so the efficiency is greatly improved.

Improved efficiency while being thread-safe.

The above is the detailed content of How to write the implementation code of java collection. For more information, please follow other related articles on the PHP Chinese website!