Java8 Collectors.groupingBy

Java8 Collectors.groupingBy 方法

1）計數

	 List<Fruit> fruitList = Lists.newArrayList(new Fruit("apple", 6),
	 	    new Fruit("apple", 6),
	        new Fruit("banana", 7), new Fruit("banana", 7),
	        new Fruit("banana", 7), new Fruit("grape",8));
	   Map<String, Long> map = fruitList.stream().
	   collect(Collectors.groupingBy(Fruit::getName,Collectors.counting()));

輸出結果是：{banana=3, apple=2, grape=1}
換一種囉嗦、複雜的寫法，但能加深理解。

	Map<String, Long> map = fruitList.stream().map(Fruit::getName).
	collect(Collectors.groupingBy(Function.identity(), Collectors.counting()));

2）排序

現在要按照水果map中value的數量逆序打印每個entry

	map.entrySet().stream().sorted(Map.Entry.<String, Long>comparingByValue().reversed())
         .forEachOrdered(System.out::println);

3）累加求和

Map<String, Integer> sumMap = fruitList.stream().collect.
(Collectors.groupingBy(Fruit::getName, Collectors.summingInt(Fruit::getPrice)));

輸出結果是：{banana=21, apple=12, grape=8}

4）分組

Map<String, List<Fruit>> groupMap = 
fruitList.stream().collect(Collectors.groupingBy(Fruit::getName));

上述代碼根據name將list分組，如果name是唯一的，那麼上述代碼就會顯得囉嗦。我們需要知道，Guava補JDK之不足，現在改Guava一顯身手了。

Map<String, Fruit> map = Maps.uniqueIndex(fruitList, Fruit::getName);

生成的Map是ImmutableMap，不可更改裏面的值。比如map.remove(“apple”)會拋出異常：java.lang.UnsupportedOperationException

根據不同的名字分爲若干組

// group by price, uses 'mapping' to convert List<Fruit> to List<String>
Map<String, List<Integer>> groupMap = 
fruitList.stream().collect(Collectors.groupingBy(Fruit::getName, 
Collectors.mapping(Fruit::getPrice, Collectors.toList())));

上面一段代碼可以用Guava代替

Multimap<String, Integer> multiMap = ArrayListMultimap.create();
fruitList.forEach(fruit -> multiMap.put(fruit.getName(), fruit.getPrice()));