Java 8 Stream流API解析

拼搏现实的明天。 2022-10-07 12:54 220阅读 0赞

## 1. Stream流简介 ##

Java 8 API添加了一个新的抽象称为流Stream，可以以一种声明的方式处理数据。

Stream流使用一种类似用 SQL 语句从数据库查询数据的直观方式来提供一种对 Java 集合运算和表达的高阶抽象。

**Stream流特点：**

*  不是数据结构，不会保存数据，只是在原数据集上定义了一组操作
 *  惰性求值，流在中间处理过程中，只是对操作进行了记录，并不会立即执行，需要等到执行终止操作的时候才会进行实际的计算
 *  Stream不保存数据，它会将操作后的数据保存到另外一个对象中

**一般使用步骤：**

1.  获取数据源，创建流
2.  中间操作(Intermediate)：可以有多个，每次返回一个新的流，可进行链式操作
3.  终端操作(Terminal)：只能有一个，每次执行完，这个流也就用光光了，无法执行下一个操作

## 2. Stream流的创建方式 ##

package com.jbp.controller;
    
    import java.io.BufferedReader;
    import java.io.FileNotFoundException;
    import java.io.FileReader;
    import java.util.Arrays;
    import java.util.List;
    import java.util.Random;
    import java.util.regex.Pattern;
    import java.util.stream.Stream;
    
    /** * @ClassName: StreamAPI * @description: * @author: JiangBeiPing * @create: 2021-06-18 11:07 * @Version: 1.0 **/
    public class StreamAPI { 
        public static void main(String[] args) throws FileNotFoundException { 
            // 1.Stream创建
            // 数组 -----------------------------------------
            // 数组：Arrays.stream(T array)
            Integer[] array = { 1, 2, 3};
            Stream<Integer> integerStream1 = Arrays.stream(array);
    
            // 集合 Collection ------------------------------
            List<Integer> list = Arrays.asList(array);
            // 集合: Collection.stream()
            Stream<Integer> listStream1 = list.stream();
            // 集合: Collection.parallelStream() 默认使用的是 ForkJoinPool.commonPool()线程池
            Stream<Integer> listStream2 = list.parallelStream();
    
            // BufferedReader -------------------------------
            BufferedReader reader = new BufferedReader(new FileReader("xxx"));
            Stream<String> linesStream = reader.lines();
    
            // 字符串处理 String ------------------------------
            String str = "1,2,3";
            Pattern pattern = Pattern.compile(",");
            Stream<String> stringStream = pattern.splitAsStream(str);
    
            // Stream自带的静态方法 ----------------------------
            // Stream.of()
            Stream<Integer> integerStream = Stream.of(1, 2, 3);
            // Stream.iterate()
            Stream<Integer> integerStream2 = Stream.iterate(1, s -> (s + 2));
            // Stream.generate()
            Stream<Double> generate = Stream.generate(Math::random);
            
        }
    
    }

## 3. Stream流API分类 ##

<table> 
 <thead> 
  <tr> 
   <th>API</th> 
   <th>类型</th> 
   <th>返回类型</th> 
   <th>使用的类型/函数式接口</th> 
   <th>函数描述符</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>filter</td> 
   <td>中间操作(Intermediate)</td> 
   <td>Stream &lt; T &gt;</td> 
   <td>Predicate &lt; T &gt;</td> 
   <td>T - &gt; boolean</td> 
  </tr> 
  <tr> 
   <td>distinct</td> 
   <td>中间操作(Intermediate，有状态 - 无界)</td> 
   <td>Stream &lt; T &gt;</td> 
   <td>—</td> 
   <td>—</td> 
  </tr> 
  <tr> 
   <td>skip</td> 
   <td>中间操作(Intermediate，有状态 - 有界)</td> 
   <td>Stream &lt; T &gt;</td> 
   <td>long</td> 
   <td>—</td> 
  </tr> 
  <tr> 
   <td>map</td> 
   <td>中间操作(Intermediate)</td> 
   <td>Stream &lt; T &gt;</td> 
   <td>Function &lt; T,R &gt;</td> 
   <td>T - &gt; R</td> 
  </tr> 
  <tr> 
   <td>flatMap</td> 
   <td>中间操作(Intermediate)</td> 
   <td>Stream &lt; T &gt;</td> 
   <td>Function &lt; T,Stream &lt; R &gt; &gt;</td> 
   <td>T - &gt; Stream &lt; R &gt;</td> 
  </tr> 
  <tr> 
   <td>sorted</td> 
   <td>中间操作(Intermediate，有状态 - 无界)</td> 
   <td>Stream &lt; T &gt;</td> 
   <td>Compartor &lt; T &gt;</td> 
   <td>(T,T) - &gt; int</td> 
  </tr> 
  <tr> 
   <td>anyMatch</td> 
   <td>终端操作(Terminal)</td> 
   <td>boolean</td> 
   <td>Predicate &lt; T &gt;</td> 
   <td>T - &gt; boolean</td> 
  </tr> 
  <tr> 
   <td>noneMatch</td> 
   <td>终端操作(Terminal)</td> 
   <td>boolean</td> 
   <td>Predicate &lt; T &gt;</td> 
   <td>T - &gt; boolean</td> 
  </tr> 
  <tr> 
   <td>allMatch</td> 
   <td>终端操作(Terminal)</td> 
   <td>boolean</td> 
   <td>Predicate &lt; T &gt;</td> 
   <td>T - &gt; boolean</td> 
  </tr> 
  <tr> 
   <td>findAny</td> 
   <td>终端操作(Terminal)</td> 
   <td>Optional &lt; T &gt;</td> 
   <td>—</td> 
   <td>—</td> 
  </tr> 
  <tr> 
   <td>findFirst</td> 
   <td>终端操作(Terminal)</td> 
   <td>Optional &lt; T &gt;</td> 
   <td>—</td> 
   <td>—</td> 
  </tr> 
  <tr> 
   <td>forEach</td> 
   <td>终端操作(Terminal)</td> 
   <td>void</td> 
   <td>Consumer &lt; T &gt;</td> 
   <td>T - &gt; void</td> 
  </tr> 
  <tr> 
   <td>collect</td> 
   <td>终端操作(Terminal)</td> 
   <td>R</td> 
   <td>Collector &lt;T,A,R&gt;</td> 
   <td>—</td> 
  </tr> 
  <tr> 
   <td>reduce</td> 
   <td>终端操作(Terminal，有状态 - 有界)</td> 
   <td>Optional &lt; T &gt;</td> 
   <td>BinaryOperator &lt; T &gt;</td> 
   <td>(T,T) - &gt; T</td> 
  </tr> 
  <tr> 
   <td>count</td> 
   <td>终端操作(Terminal)</td> 
   <td>long</td> 
   <td>—</td> 
   <td>—</td> 
  </tr> 
 </tbody> 
</table>

*  **有状态**：指该操作只有拿到所有元素之后才能继续下去
 *  **无状态**：指元素的处理不受之前元素的影响

## 4. Stream流中间操作API ##

#### 4.1 filter 过滤 ####

将结果为false的元素过滤掉，为true会把当前元素会保留下来。

数值型过滤：

Integer[] array = { 1, 2, 3};
     Stream<Integer> integerStream1 = Arrays.stream(array);
     Stream<Integer> filterStream = integerStream1.filter(i -> i == 2);
     System.out.println(Arrays.toString(filterStream.toArray())); // [2]

对象型过滤：

User user1 = new User(1,"张三");
    User user2 = new User(2,"李四");
    User user3 = new User(3,"王五");
    ArrayList<User> users = new ArrayList<>();
    users.add(user1);
    users.add(user2);
    users.add(user3);
    Stream<User> userStream = users.stream().filter(user -> "王五".equals(user.getName()));
    System.out.println(Arrays.toString(userStream.toArray())); // [StreamAPI.User(id=3, name=王五)]

#### 4.2 map 映射 ####

接收一个函数作为参数，该函数会被应用到每个元素上，并将其映射成一个新的元素

数组类型映射：

Integer[] array = { 1, 2, 3};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    Stream<Integer> mapStream = integerStream1.map(i -> { 
                i++;
                return i;
            });
    System.out.println(Arrays.toString(mapStream.toArray())); // [2, 3, 4]

对象类型映射：

User user1 = new User(1,"张三");
    User user2 = new User(2,"李四");
    User user3 = new User(3,"王五");
    ArrayList<User> users = new ArrayList<>();
    users.add(user1);
    users.add(user2);
    users.add(user3);
    
    Stream<User> mapStream2 = users.stream().map(user -> { 
                String name = user.getName();
                user.setName(name + "xx");
                return user;
            });
    System.out.println(Arrays.toString(mapStream2.toArray())); // [StreamAPI.User(id=1, name=张三xx), StreamAPI.User(id=2, name=李四xx), StreamAPI.User(id=3, name=王五xx)]

#### 4.3 flatMap 映射 ####

接收一个函数作为参数，将流中的每个值都换成另一个流，然后把所有流连接成一个流

数组类型映射：

Integer[] array = { 1, 2, 3};
    Integer[] array2 = { 8, 9, 10};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    // 每个元素 i 转换为一次新的stream流 并返回
    Stream<Integer> mapStream = integerStream1.flatMap(i -> { 
                Stream<Integer> stream = Arrays.stream(array2);
                return stream;
            });
    System.out.println(Arrays.toString(mapStream.toArray())); // [8, 9, 10, 8, 9, 10, 8, 9, 10]

对象类型映射：

User user1 = new User(1,"张三");
    User user2 = new User(2,"李四");
    User user3 = new User(3,"王五");
    ArrayList<User> users = new ArrayList<>();
    users.add(user1);
    users.add(user2);
    users.add(user3);
    
    Stream<String> stringStream = users.stream().flatMap(user -> { 
                String name = user.getName();
                String[] split = name.split("");
                Stream<String> stream = Arrays.stream(split);
                return stream;
            });
    System.out.println(Arrays.toString(stringStream.toArray())); // [张, 三, 李, 四, 王, 五]

#### 4.4 forEach 遍历 ####

Integer[] array = { 1, 2, 3};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    // 遍历array数组 ---> 1,2,3
    integerStream1.forEach(i -> { 
                System.out.println(i);
            });

#### 4.5 limit 截取 ####

保留前n个元素

Integer[] array = { 1, 2, 3};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    Stream<Integer> integerStream = integerStream1.limit(2);
    System.out.println(Arrays.toString(integerStream.toArray())); // [1, 2]

#### 4.6 skip 跳过 ####

跳过前n个元素

Integer[] array = { 1, 2, 3};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    Stream<Integer> integerStream = integerStream1.skip(2);
    System.out.println(Arrays.toString(integerStream.toArray())); // [3]

#### 4.7 distinct 去重 ####

剔除重复元素

Integer[] array = { 1, 2, 3,3};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    Stream<Integer> integerStream = integerStream1.distinct();
    System.out.println(Arrays.toString(integerStream.toArray())); // [1, 2, 3]

#### 4.8 sorted 排序 ####

*  sorted()：自然排序，流中元素需实现Comparable接口
 *  sorted(Comparator com)：定制排序，自定义Comparator排序器

Integer[] array = { 1, 2, 3,8,5};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    Stream<Integer> integerStream = integerStream1.sorted();
    System.out.println(Arrays.toString(integerStream.toArray())); // [1, 2, 3, 5, 8]

Integer[] array = { 1, 2, 3,8,5};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    // 倒序排列
    Stream<Integer> integerStream = integerStream1.sorted((i1,i2) -> { 
               return i2 - i1;
            });
    System.out.println(Arrays.toString(integerStream.toArray())); // [8, 5, 3, 2, 1]

#### 4.9 peek 消费 ####

如同于map，能得到流中的每一个元素。但map接收的是一个Function表达式，有返回值；而peek接收的是Consumer表达式，没有返回值

User user1 = new User(1,"张三");
    User user2 = new User(2,"李四");
    User user3 = new User(3,"王五");
    ArrayList<User> users = new ArrayList<>();
    users.add(user1);
    users.add(user2);
    users.add(user3);
    Stream<User> peek = users.stream().peek(user -> { 
                String name = user.getName();
                user.setName(name + "v");
            });
    System.out.println(Arrays.toString(peek.toArray())); // [StreamAPI.User(id=1, name=张三v), StreamAPI.User(id=2, name=李四v), StreamAPI.User(id=3, name=王五v)]

## 5. Stream流终止操作API ##

#### 5.1 匹配、聚合操作 ####

**匹配：**

*  anyMatch()，只要有一个元素匹配传入的条件，就返回 true
 *  allMatch()，只有有一个元素不匹配传入的条件，就返回 false；如果全部匹配，则返回true
 *  noneMatch()，只要有一个元素匹配传入的条件，就返回 false；如果全部匹配，则返回 true

Integer[] array = { 1, 2, 3,8,5};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    boolean anyMatch = integerStream1.anyMatch(i -> i > 6); // true
    System.out.println(anyMatch);
    boolean allMatch = integerStream1.allMatch(i -> i > 6); // false
    System.out.println(allMatch);
    boolean noneMatch = integerStream1.noneMatch(i -> i > 6); // false
    System.out.println(noneMatch);

**聚合：**

*  findFirst：返回流中第一个元素
 *  findAny：返回流中的任意元素
 *  count：返回流中元素的总个数
 *  max：返回流中元素最大值
 *  min：返回流中元素最小值

Integer[] array = { 1, 2, 3,8,5};
    Stream<Integer> integerStream1 = Arrays.stream(array);
    Optional<Integer> first = integerStream1.findFirst();// 1
    Optional<Integer> any = integerStream1.findAny();
    long count = integerStream1.count(); // 5
    Optional<Integer> max = integerStream1.max(Integer::compareTo);// Optional[8]
    Optional<Integer> min = integerStream1.min(Integer::compareTo);//Optional[1]

#### 5.2 归并 ####

合并流的元素并产生单个值

*  Optional reduce(BinaryOperator accumulator)
 *  T reduce(T identity, BinaryOperator accumulator)
 *  < U > U reduce(U identity,BiFunction<U, ? super T, U> accumulator,BinaryOperator combiner)

identity = 默认值或初始值

BinaryOperator = 函数式接口，取两个值并产生一个新值。（注： java Function 函数中的 BinaryOperator 接口用于执行 lambda 表达式并返回一个 T 类型的返回值）

Integer[] array = { 1, 2, 3,8,5};
    int sum = 0;
    for (Integer i : array) { 
         sum += i;
    }
    System.out.println(sum); // 19
    
    Stream<Integer> integerStream1 = Arrays.stream(array);
    Integer reduce = integerStream1.reduce(0, (a, b) -> a + b);
    System.out.println(reduce);// 19

## 6. Stream流收集器collect ##

**集合、聚合收集：**

User user1 = new User(1,"张三");
            User user2 = new User(2,"李四");
            User user3 = new User(3,"王五");
            ArrayList<User> users = new ArrayList<>();
            users.add(user1);
            users.add(user2);
            users.add(user3);
    
            // list
            List<Integer> list = integerStream1.collect(Collectors.toList());
            // set
            Set<Integer> set = integerStream1.collect(Collectors.toSet());
            // map key不能相同
            Map<Integer, String> map = users.stream().collect(Collectors.toMap(User::getId, User::getName));
            // 获取users元素个数
            Long count = users.stream().collect(Collectors.counting());
            // 获取users集合id最大值
            Optional<Integer> max = users.stream().map(User::getId).collect(Collectors.maxBy(Integer::compare));
            // 获取users集合id最小值
            Optional<Integer> min = users.stream().map(User::getId).collect(Collectors.minBy(Integer::compare));
            // 获取users集合所有id的和
            Integer id = users.stream().collect(Collectors.summingInt(User::getId));
            // 获取users集合所有id的平均值
            Double average = users.stream().collect(Collectors.averagingDouble(User::getId));

**字符串拼接：**

String s = users.stream().map(User::getName).collect(Collectors.joining(",", "{", "}"));// {张三,李四,王五}

**分组、分区和归并：**

User user1 = new User(1,"张三");
            User user2 = new User(2,"李四");
            User user3 = new User(2,"王五");
            ArrayList<User> users = new ArrayList<>();
            users.add(user1);
            users.add(user2);
            users.add(user3);
            // 按照id分组
            Map<Integer, List<User>> group = users.stream().collect(Collectors.groupingBy(User::getId));// {1=[StreamAPI.User(id=1, name=张三)], 2=[StreamAPI.User(id=2, name=李四), StreamAPI.User(id=2, name=王五)]}
            // 多重分组：先按照名字分，再按照id分
            Map<String, Map<Integer, List<User>>> map = users.stream().collect(Collectors.groupingBy(User::getName, Collectors.groupingBy(User::getId)));//{李四={2=[StreamAPI.User(id=2, name=李四)]}, 张三={1=[StreamAPI.User(id=1, name=张三)]}, 王五={2=[StreamAPI.User(id=2, name=王五)]}}
    
            // 按照id分区：大于1为一部分，小于等于1为一部分
            Map<Boolean, List<User>> collect = users.stream().collect(Collectors.partitioningBy(user -> user.getId() > 1));//{false=[StreamAPI.User(id=1, name=张三)], true=[StreamAPI.User(id=2, name=李四), StreamAPI.User(id=2, name=王五)]}
    
            // 归并：计算users集合的id之和
            Optional<Integer> sum = users.stream().map(User::getId).collect(Collectors.reducing(Integer::sum));//Optional[5]