Collectors 是 Java 8 加入的操作类，位于 java.util.stream 包下。它会根据不同的策略将元素收集归纳起来，比如最简单常用的是将元素装入 Map、Set、List等可变容器中

toList、toSet

Collector<T, ?, List<T>> toList() 
返回一个将输入元素累积到新List中的Collector

Collector<T, ?, Set<T>> toSet() 
返回一个将输入元素累积到新Set中的Collector

Stream<String> language = Stream.of("java", "php","java");
List<String> list = language.collect(Collectors.toList());
System.out.println(list);//[java, php, java]

Stream<String> language1 = Stream.of("java", "php","java");
Set<String> set = language1.collect(Collectors.toSet());
System.out.println(set);//[java, php]

toMap

Collector<T, ?, Map<K,U>> toMap(Function<? super T, ? extends K> keyMapper, Function<? super T, ? extends U> valueMapper)
返回一个将输入元素累积到新map中的Collector，keyMapper生成key的函数，valueMapper生成value的函数,当可以相同时报错java.lang.IllegalStateException: Duplicate key

Collector<T, ?, Map<K,U>> toMap(Function<? super T, ? extends K> keyMapper, Function<? super T, ? extends U> valueMapper, BinaryOperator<U> mergeFunction)
返回一个将输入元素累积到新map中的Collector，keyMapper生成key的函数，valueMapper生成value的函数,mergeFunction处理相同key的元素

Collector<T, ?, M> toMap(Function<? super T, ? extends K> keyMapper, Function<? super T, ? extends U> valueMapper, BinaryOperator<U> mergeFunction, Supplier<M> mapSupplier)
返回一个将输入元素累积到新map中的Collector，keyMapper生成key的函数，valueMapper生成value的函数,mergeFunction处理相同key的元素,mapSupplier一个函数，它返回一个新的指定实现类的空Map ，结果将插入其中

Stream<String> language2 = Stream.of("java", "php");
Map<String, Integer> map = language2.collect(Collectors.toMap(o -> o, o -> o.length()));
System.out.println(map);//{java=4, php=3}

Stream<String> language21 = Stream.of("java", "php","java");
Map<String, Integer> map21 = language21.collect(Collectors.toMap(o -> o, o -> o.length(), (o1, o2) ->  o1));
System.out.println(map21);//{java=4, php=3}

Stream<String> language22 = Stream.of("java", "php","java");
LinkedHashMap<String, Integer> map22 = language22.collect(Collectors.toMap(o -> o, o -> o.length(), (o1, o2) ->  o1, LinkedHashMap::new));
System.out.println(map22);//{java=4, php=3}

toConcurrentMap

和toMap相似，但是使用并发ConcurrentMap承装

Collector<T, ?, ConcurrentMap<K,U>> toConcurrentMap(Function<? super T, ? extends K> keyMapper, Function<? super T, ? extends U> valueMapper)
返回一个将输入元素累积到新并发map中的Collector，keyMapper生成key的函数，valueMapper生成value的函数,当可以相同时报错java.lang.IllegalStateException: Duplicate key

Collector<T, ?, ConcurrentMap<K,U>> toConcurrentMap(Function<? super T, ? extends K> keyMapper, Function<? super T, ? extends U> valueMapper, BinaryOperator<U> mergeFunction)
返回一个将输入元素累积到新map中的Collector，keyMapper生成key的函数，valueMapper生成value的函数,mergeFunction处理相同key的元素

Collector<T, ?, M> toConcurrentMap(Function<? super T, ? extends K> keyMapper, Function<? super T, ? extends U> valueMapper, BinaryOperator<U> mergeFunction, Supplier<M> mapSupplier)
返回一个将输入元素累积到新map中的Collector，keyMapper生成key的函数，valueMapper生成value的函数,mergeFunction处理相同key的元素,mapSupplier一个函数，它返回一个新的指定实现类的空Map ，结果将插入其中

Stream<String> language4 = Stream.of("java", "php");
ConcurrentMap<String, Integer> concurrentMap = language4.collect(Collectors.toConcurrentMap(o -> o, o -> o.length()));
System.out.println(concurrentMap);//{java=4, php=3}

Stream<String> language41 = Stream.of("java", "php","java");
ConcurrentMap<String, Integer> concurrentMap1 = language41.collect(Collectors.toConcurrentMap(o -> o, o -> o.length(), (o1, o2) ->  o1));
System.out.println(concurrentMap1);//{java=4, php=3}

Stream<String> language42 = Stream.of("java", "php","java");
ConcurrentMap<String, Integer> concurrentMap2 = language42.collect(Collectors.toConcurrentMap(o -> o, o -> o.length(), (o1, o2) ->  o1, ConcurrentHashMap::new));
System.out.println(concurrentMap2);//{java=4, php=3}

toCollection

指定承装集合类型

Collector<T, ?, C> toCollection(Supplier<C> collectionFactory)
返回一个将输入元素累积到一个新的指定实现类的Collection中

Stream<String> language3 = Stream.of("java", "php","java");
ArrayList<String> arrayList = language3.collect(Collectors.toCollection(ArrayList::new));
System.out.println(arrayList);//[java, php, java]

joining

拼接元素

Collector<CharSequence, ?, String> joining() 
返回一个Collector ，它按照遇到的顺序将输入元素连接成一个String

Collector<CharSequence, ?, String> joining(CharSequence delimiter) 
返回一个以遇到顺序连接输入元素的Collector ，由指定的分隔符分隔

Collector<CharSequence, ?, String> joining(CharSequence delimiter, CharSequence prefix, CharSequence suffix)
 返回一个Collector ，它按遇到顺序连接输入元素，由指定的分隔符分隔，具有指定的前缀和后缀

Stream<String> language = Stream.of("java", "php","java");
String joining = language.collect(Collectors.joining());
System.out.println(joining);//javaphpjava

Stream<String> language1 = Stream.of("java", "php","java");
String joining1 = language1.collect(Collectors.joining(","));
System.out.println(joining1);//java,php,java

Stream<String> language2 = Stream.of("java", "php","java");
String joining2 = language2.collect(Collectors.joining(",", "[", "]"));
System.out.println(joining2);//[java,php,java]

collectingAndThen

对Collector执行多次加工转换

Collector<T,A,RR> collectingAndThen(Collector<T,A,R> downstream, Function<R,RR> finisher) 
对Collector执行多次加工转换

Stream<String> language = Stream.of("java", "php","java");
String collectingAndThen = language.collect(Collectors.collectingAndThen(Collectors.joining(","), String::toUpperCase));
System.out.println(collectingAndThen);//JAVA,PHP,JAVA

groupingBy

对元素分组

Collector<T, ?, Map<K, List<T>>> groupingBy(Function<? super T, ? extends K> classifier)
对类型T输入元素执行“分组”操作,符合条件的元素将组成一个 List 映射到以T为key 的 Map 中

Collector<T, ?, Map<K, D>> groupingBy(Function<? super T, ? extends K> classifier, Collector<? super T, A, D> downstream)
对类型T输入元素执行“分组”操作,符合条件的元素将组成一个 集合 映射到以T为key 的 Map 中，downstream指定返回集合类型

Collector<T, ?, M> groupingBy(Function<? super T, ? extends K> classifier, Supplier<M> mapFactory, Collector<? super T, A, D> downstream)
对类型T输入元素执行“分组”操作,符合条件的元素将组成一个 集合 映射到以T为key 的 mapFactory新生成的map 中，downstream指定返回集合类型

Stream<String> language = Stream.of("java", "php","java");
Map<Integer, List<String>> map = language.collect(Collectors.groupingBy(String::length));
System.out.println(map);//{3=[php], 4=[java, java]}

Stream<String> language2 = Stream.of("java", "php","java");
Map<Integer, Set<String>> map2 = language2.collect(Collectors.groupingBy(String::length, Collectors.toSet()));
System.out.println(map2);//{3=[php], 4=[java]}

Stream<String> language3 = Stream.of("java", "php","java");
LinkedHashMap<Integer, Map<String, Integer>> map3 = language3.collect(Collectors.groupingBy(String::length, LinkedHashMap::new, Collectors.toMap(o -> o, o -> o.length(), (o1, o2) ->  o1)));
System.out.println(map3);//{4={java=4}, 3={php=3}}

groupingByConcurrent

分组后返回的是一个并发集合

Collector<T, ?, ConcurrentMap<K, List<T>>> groupingByConcurrent(Function<? super T, ? extends K> classifier)
对类型T输入元素执行“分组”操作,符合条件的元素将组成一个 List 映射到以T为key 的 并发Map 中

Collector<T, ?, ConcurrentMap<K, D>> groupingByConcurrent(Function<? super T, ? extends K> classifier, Collector<? super T, A, D> downstream)
对类型T输入元素执行“分组”操作,符合条件的元素将组成一个 集合 映射到以T为key 的 并发Map 中，downstream指定返回集合类型

Collector<T, ?, M> groupingByConcurrent(Function<? super T, ? extends K> classifier, Supplier<M> mapFactory, Collector<? super T, A, D> downstream)
对类型T输入元素执行“分组”操作,符合条件的元素将组成一个 集合 映射到以T为key 的 mapFactory新生成的并发map 中，downstream指定返回集合类型

Stream<String> language = Stream.of("java", "php","java");
Map<Integer, List<String>> map = language.collect(Collectors.groupingByConcurrent(String::length));
System.out.println(map);//{3=[php], 4=[java, java]}

Stream<String> language2 = Stream.of("java", "php","java");
Map<Integer, Set<String>> map2 = language2.collect(Collectors.groupingByConcurrent(String::length, Collectors.toSet()));
System.out.println(map2);//{3=[php], 4=[java]}

Stream<String> language3 = Stream.of("java", "php","java");
ConcurrentHashMap<Integer, Map<String, Integer>> map3 = language3.collect(Collectors.groupingByConcurrent(String::length, ConcurrentHashMap::new, Collectors.toMap(o -> o, o -> o.length(), (o1, o2) ->  o1)));
System.out.println(map3);//{3={php=3}, 4={java=4}}

partitioningBy

Collector<T, ?, Map<Boolean, List<T>>> partitioningBy(Predicate<? super T> predicate)
根据Predicate对输入元素进行判断分区，并将它们组织成Map<Boolean, List<T>>

Collector<T, ?, Map<Boolean, D>> partitioningBy(Predicate<? super T> predicate, Collector<? super T, A, D> downstream)
根据Predicate对输入元素进行判断分区，并将它们组织成Map<Boolean, ?<T>>,downstream指定返回集合类型

Stream<String> language = Stream.of("java", "php","java");
Map<Boolean, List<String>> map = language.collect(Collectors.partitioningBy(o -> o.length() > 3));
System.out.println(map);//{false=[php], true=[java, java]}

Stream<String> language2 = Stream.of("java", "php","java");
Map<Boolean, Set<String>> map2 = language2.collect(Collectors.partitioningBy(o -> o.length() > 3, Collectors.toSet()));
System.out.println(map2);//{false=[php], true=[java]}

counting

Collector<T, ?, Long> counting()
计算输入元素的数量。 如果不存在元素，则结果为 0。

Stream<String> language = Stream.of("java", "php","java");
Long count = language.collect(Collectors.counting());
System.out.println(count);//3

Stream.count() 实现相同功能

Stream<String> language = Stream.of("java", "php","java");
Long count = language.count();
System.out.println(count);//3

maxBy

Collector<T, ?, Optional<T>> maxBy(Comparator<? super T> comparator)
根据comparato比较后返回最大元素的Optional<T>

Stream<String> language = Stream.of("java", "php","java");
Optional<String> maxObj = language.collect(Collectors.maxBy(Comparator.comparingInt(String::length)));
System.out.println(maxObj.get());//java

Stream.max() 实现相同功能

Stream<String> language = Stream.of("java", "php","java");
Optional<String> max = language.max(Comparator.comparingInt(String::length));
System.out.println(max.get());//java

minBy

Collector<T, ?, Optional<T>> minBy(Comparator<? super T> comparator)
根据comparato比较后返回最小元素的Optional<T>

Stream<String> language = Stream.of("java", "php","java");
Optional<String> maxObj = language.collect(Collectors.minBy(Comparator.comparingInt(String::length)));
System.out.println(maxObj.get());//php

Stream.min() 实现相同功能

Stream<String> language = Stream.of("java", "php","java");
Optional<String> max = language.min(Comparator.comparingInt(String::length));
System.out.println(max.get());//php

summingXXX

Collector<T, ?, Double> summingDouble(ToDoubleFunction<? super T> mapper)
求输入元素的Double类型数值函数的总和
Collector<T, ?, Integer> summingInt(ToIntFunction<? super T> mapper)
求输入元素的Integer类型数值函数的总和
Collector<T, ?, Long> summingLong(ToLongFunction<? super T> mapper)
求输入元素的Long类型数值函数的总和

Stream<String> language = Stream.of("java", "php","java");
Double sumDouble = language.collect(Collectors.summingDouble(String::length));
System.out.println(sumDouble);//11.0

Stream<String> language2 = Stream.of("java", "php","java");
Integer sumInt = language2.collect(Collectors.summingInt(String::length));
System.out.println(sumInt);//11

Stream<String> language3 = Stream.of("java", "php","java");
Long sumLong = language3.collect(Collectors.summingLong(String::length));
System.out.println(sumLong);//11

Stream.sum() 实现相同功能

当Stream的类似是数字时，可以用Stream.sum()实现相同功能

DoubleStream language = DoubleStream.of(4, 3, 4);
double sumDouble = language.sum();
System.out.println(sumDouble);//11.0

IntStream language2 = IntStream.of(4, 3, 4);
Integer sumInt = language2.sum();
System.out.println(sumInt);//11

LongStream language3 = LongStream.of(4, 3, 4);
Long sumLong = language3.sum();
System.out.println(sumLong);//11

summarizingXXX

汇总统计

Collector<T, ?, DoubleSummaryStatistics> summarizingDouble(ToDoubleFunction<? super T> mapper)
以Double类型数值函数汇总统计，包含了 总数，总和，最小值，最大值，平均值 五个指标。

Collector<T, ?, IntSummaryStatistics> summarizingInt(ToIntFunction<? super T> mapper)
以Integer类型数值函数汇总统计，包含了 总数，总和，最小值，最大值，平均值 五个指标。

Collector<T, ?, LongSummaryStatistics> summarizingLong(ToLongFunction<? super T> mapper)
以Long类型数值函数汇总统计，包含了 总数，总和，最小值，最大值，平均值 五个指标。

Stream<String> language = Stream.of("java", "php","java");
DoubleSummaryStatistics doubleSummaryStatistics = language.collect(Collectors.summarizingDouble(String::length));
System.out.println(doubleSummaryStatistics);//DoubleSummaryStatistics{count=3, sum=11.000000, min=3.000000, average=3.666667, max=4.000000}

Stream<String> language2 = Stream.of("java", "php","java");
IntSummaryStatistics intSummaryStatistics = language2.collect(Collectors.summarizingInt(String::length));
System.out.println(intSummaryStatistics);//IntSummaryStatistics{count=3, sum=11, min=3, average=3.666667, max=4}

Stream<String> language3 = Stream.of("java", "php","java");
LongSummaryStatistics longSummaryStatistics = language3.collect(Collectors.summarizingLong(String::length));
System.out.println(longSummaryStatistics);//LongSummaryStatistics{count=3, sum=11, min=3, average=3.666667, max=4}

mapping

Collector<T, ?, R> mapping(Function<? super T, ? extends U> mapper, Collector<? super U, A, R> downstream)
先对元素使用 mapper 进行再加工操作，然后用另一个downstream 归纳

Stream<String> language = Stream.of("java", "php","java");
List<String> list = language.collect(Collectors.mapping(s -> s.substring(1), Collectors.toList()));
System.out.println(list);//[ava, hp, ava]

Stream.map 实现相同功能

Stream<String> language2 = Stream.of("java", "php","java");
List<String> list2 = language2.map(s -> s.substring(1)).collect(Collectors.toList());
System.out.println(list2);//[ava, hp, ava]

reducing

Collector<T, ?, Optional<T>> reducing(BinaryOperator<T> op)
根据op比较器来比较大小并分别返回最大值或者最小值,可能拿不到结果

Collector<T, ?, T> reducing(T identity, BinaryOperator<T> op)
根据op比较器来比较大小并分别返回最大值或者最小值,identity为基准值，基准值也参与比较，没有结果时返回基准值，否则返回比较结果

Collector<T, ?, U> reducing(U identity, Function<? super T, ? extends U> mapper, BinaryOperator<U> op)
先对元素进行mapper映射处理，再根据op比较器来比较大小并分别返回最大值或者最小值,identity为基准值，基准值也参与比较，没有结果时返回基准值，否则返回比较结果

Stream<String> language = Stream.of("java", "php","java");
Comparator<String> comparing = Comparator.comparing(String::length);
Optional<String> obj = language.collect(Collectors.reducing(BinaryOperator.minBy(comparing)));
System.out.println(obj);//Optional[php]

Stream<String> language2 = Stream.of("java", "php","java");
Comparator<String> comparing2 = Comparator.comparing(String::length);
String objStr = language2.collect(Collectors.reducing("androd", BinaryOperator.maxBy(comparing2)));
System.out.println(objStr);//androd

Stream<String> language3 = Stream.of("java", "php","java");
Function<String, String> mapper = str -> {return str.substring(1);};
Comparator<String> comparing3 = Comparator.comparing(String::length);
String objStr2 = language3.collect(Collectors.reducing("androd", mapper, BinaryOperator.maxBy(comparing3)));
System.out.println(objStr2);//androd

Stream.reduce 实现相同功能

Stream<String> language = Stream.of("java", "php","java");
Comparator<String> comparing = Comparator.comparing(String::length);
Optional<String> obj = language.reduce(BinaryOperator.minBy(comparing));
System.out.println(obj);//Optional[php]

Stream<String> language2 = Stream.of("java", "php","java");
Comparator<String> comparing2 = Comparator.comparing(String::length);
String objStr = language2.reduce("androd", BinaryOperator.maxBy(comparing2));
System.out.println(objStr);//androd

Stream<String> language3 = Stream.of("java", "php","java");
BiFunction<String, String, String> mapper = (str1, str2) -> { return str1.length()>str2.length()? str1:str2; };
Comparator<String> comparing3 = Comparator.comparing(String::length);
//第三个参数BinaryOperator.minBy(comparing3)无效，作用是合并各个流结果，故在parallelStream时才生效
String objStr2 = language3.reduce("androd", mapper, BinaryOperator.minBy(comparing3));
System.out.println(objStr2);//androd

averagingDouble

Collector<T, ?, Double> averagingDouble(ToDoubleFunction<? super T> mapper)
对元素进行Double类型数值函数映射处理后，求拼接值

Collector<T, ?, Double> averagingInt(ToIntFunction<? super T> mapper)
对元素进行Int类型数值函数映射处理后，求拼接值

Collector<T, ?, Double> averagingLong(ToLongFunction<? super T> mapper)
对元素进行Long类型数值函数映射处理后，求拼接值

Stream<String> language = Stream.of("java", "php","java");
Double averagingDouble = language.collect(Collectors.averagingDouble(String::length));
System.out.println(averagingDouble);//3.6666666666666665

Stream<String> language2 = Stream.of("java", "php","java");
Double averagingInt2 = language2.collect(Collectors.averagingInt(String::length));
System.out.println(averagingInt2);//3.6666666666666665

Stream<String> language3 = Stream.of("java", "php","java");
Double averagingLong3 = language3.collect(Collectors.averagingLong(String::length));
System.out.println(averagingLong3);//3.6666666666666665

Stream.average() 实现相同功能

当Stream的类似是数字时，可以用Stream.sum()实现相同功能

DoubleStream language = DoubleStream.of(4, 3, 4);
OptionalDouble sumDouble = language.average();
System.out.println(sumDouble.getAsDouble());//3.6666666666666665

IntStream language2 = IntStream.of(4, 3, 4);
OptionalDouble sumInt = language2.average();
System.out.println(sumInt.getAsDouble());//3.6666666666666665

LongStream language3 = LongStream.of(4, 3, 4);
OptionalDouble sumLong = language3.average();
System.out.println(sumLong.getAsDouble());//3.6666666666666665