Java8 stream流操作: 去重,排序,筛选,分组,聚合计算
Java8 stream流操作: 去重,排序,筛选,分组,聚合计算
流是从支持数据处理操作的源生成的元素序列，源可以是数组、文件、集合、函数。流不是集合元素，它不是数据结构并不保存数据，它的主要目的在于计算
一.List
- 测试对象:
/**
* @Description : stream流测试对象
*/
//允许链式set
@Accessors(chain = true)
@Data
public class StreamDto {
private String name;
private String addr;
private Integer age;
private Date birthDay;
private BigDecimal money;
public StreamDto(String name, String addr, Integer age, Date birthDay, BigDecimal money) {
this.name = name;
this.addr = addr;
this.age = age;
this.birthDay = birthDay;
this.money = money;
}
public StreamDto() {
}
}
- 测试集合
StreamDto d1 = new StreamDto();
d1.setName("a").setAddr("北京").setAge(20)
.setBirthDay(DateUtil.parse("2020-10-12")).setMoney(new BigDecimal(3));
StreamDto d2 = new StreamDto();
d2.setName("a").setAddr("南京").setAge(12)
.setBirthDay(DateUtil.parse("2020-10-13")).setMoney(new BigDecimal(1));
StreamDto d3 = new StreamDto();
d3.setName("b").setAddr("北京").setAge(23)
.setBirthDay(DateUtil.parse("2020-10-11")).setMoney(new BigDecimal(2));
List<StreamDto> list = CollUtil.toList(d1,d2,d3);
System.out.println("原始List=="+list.toString());
List排序(包含null字段)
- 不用stream,对自身排序,不返回新List
//age升序
list.sort(Comparator.comparing(StreamDto::getAge));
//age降序
list.sort(Comparator.comparing(StreamDto::getAge).reversed());
//正序排序，并将null放在最后(搭配reversed()就变成倒序):小->大->null
list.sort(Comparator.comparing(StreamDto::getBirthday,
Comparator.nullsLast(Date::compareTo)));
//先正序排序，并将null放在最前，搭配reversed()翻转就变成倒序：大->小->null
list.sort(Comparator.comparing(StreamDto::getBirthday,
Comparator.nullsFirst(Date::compareTo)).reversed());
//先后排序
list.sort(Comparator.comparing(StreamDto::getAge).reversed()
.thenComparing(Comparator.comparing(StreamDto::getName).reversed())
.thenComparing(Comparator.comparing(StreamDto::getAddr).reversed()));
- 使用stream
//age升序
List<StreamDto> l1 = list.stream().sorted(Comparator.comparing(StreamDto::getAge))
.collect(Collectors.toList());
//age降序
l1 = list.stream().sorted(Comparator.comparing(StreamDto::getAge).reversed())
.collect(Collectors.toList());
//正序排序，并将null放在最后(搭配reversed()就变成倒序):小->大->null
l2 = list.stream().sorted(Comparator.comparing(StreamDto::getBirthday,
Comparator.nullsLast(Date::compareTo))).collect(Collectors.toList());
//先正序排序，并将null放在最前，搭配reversed()翻转就变成倒序：大->小->null
l3 = list.stream().sorted(Comparator.comparing(StreamDto::getBirthday,
Comparator.nullsFirst(Date::compareTo)).reversed()).collect(Collectors.toList());
List聚合计算
//包含了: 计数,最大,最小,求和,平均数, 每种方式都有单独的方式实现
DoubleSummaryStatistics statistics = list.stream()
.collect(Collectors.summarizingDouble(StreamDto::getAge));
System.out.println("count:" + statistics.getCount() + ",max:" + statistics.getMax() + ",min=:" + statistics.getMin() + ",sum:" + statistics.getSum() + ",average:" + statistics.getAverage());
//count:3,max:23.0,min=:12.0,sum:55.0,average:18.333333333333332
//计数:
long count = list.stream().count();//3
//如果想直接返回整型
Integer countInt = list.stream().collect(Collectors.summingInt(l -> 1));
//最大/最小,两种方式:
//sorted: 先按年龄降序排列后取第一个,同理最小则升序取第一
Integer max = list.stream().sorted(Comparator.comparing(StreamDto::getAge).reversed()).map(StreamDto::getAge).findFirst().get();
System.out.println("max = " + max);
//min/max
Integer min = list.stream().min(Comparator.comparing(StreamDto::getAge)).get().getAge();
System.out.println("min = " + min);
//求和
Integer sum1 = list.stream().collect(Collectors.summingInt(StreamDto::getAge));
Integer sum2 = list.stream().map(StreamDto::getAge).reduce(Integer::sum).get();
//平均数
Double collect2 = list.stream().collect(Collectors.averagingInt(StreamDto::getAge));
double asDouble = list.stream().mapToLong(StreamDto::getAge).average().getAsDouble();
double asDouble1 = list.stream().mapToInt(StreamDto::getAge).average().getAsDouble();
//...
List条件筛选
List<StreamDto> l1 = null;
//条件筛选
l1 = list.stream().filter(b -> b.getAge() > 15).collect(Collectors.toList());
//多条件
l1 = list.stream()
.filter(b -> DateUtil.compare(new Date(),b.getBirthDay()) > 0
&& b.getAge() < 15 )
.collect(Collectors.toList());
l1 = list.stream()
.filter(b -> b.getMoney().compareTo(new BigDecimal(2)) > -1).collect(Collectors.toList());
List分组
//name分组
Map<String,List<StreamDto>> map = list.stream().collect(Collectors.groupingBy(StreamDto::getName));
//二级分组
Map<String,Map<Integer,List<StreamDto>>> map1 = list.stream().collect(Collectors.groupingBy(StreamDto::getName,Collectors.groupingBy(StreamDto::getAge)));
//二级分组,统计数量
Map<String,Map<Integer,Long>> map2 = list.stream().collect(Collectors.groupingBy(StreamDto::getName,Collectors.groupingBy(StreamDto::getAge,Collectors.counting())));
//多条件分组
map = list.stream().collect(Collectors.groupingBy(b -> b.getName() + "-" + b.getAge()));
//分组后又对集合进行处理,方式1
Map<String,List<String>> map3 = list.stream().collect(Collectors.groupingBy(StreamDto::getName,Collectors.groupingBy(StreamDto::getAge,Collectors.mapping(StreamDto::getAddr, Collectors.toList()))));
//方式2，只是拓展，不推荐
map3 = list.stream().collect(Collectors.groupingBy(StreamDto::getName)).entrySet().stream().collect(Collectors.toMap(b -> b.getKey(), b -> CollUtil.map(b.getValue(), StreamDto::getAddr, true)));
List去重
//根据Addr去重
List<StreamDto> list1 = list.stream()
.collect(Collectors.collectingAndThen(Collectors.toCollection(
() -> new TreeSet<>(Comparator.comparing(StreamDto::getAddr))
), ArrayList::new));
//简单去重
list1 = list.stream().distinct().collect(Collectors.toList());
List转Map
Map<String, Date> m = list.stream().collect(Collectors.toMap(b -> b.getName(), b -> b.getBirthDay()));
Map<String, StreamDto> collect3 = list.stream().collect(Collectors.toMap(b -> b.getName(), b -> b));
返回自封装的List实现类
//封装的List实现,用于链式添加:addObj()和addObjs()
ArrayListProxy<String> allPlatNames = new ArrayListProxy<>();
allPlatNames.addObjs(bidSecPlats).addObjs(projPlats);
//去重,这里要返回ArrayListProxy必须自己实现,collect(Collectors.toList())只能返回List,allPlatNames接收不了
allPlatNames = allPlatNames.stream().distinct().collect(ArrayListProxy::new, (r, x) -> {
r.addObj(x);
}, List::addAll);
集合合并返回新集合
List<String> list1 = CollUtil.toList("1");
List<String> list2 = CollUtil.toList("2");
List<List<String>> all = CollUtil.toList(list1, list2);
/**
* 第一种方式
*/
ArrayList<String> collect = all.stream().collect(ArrayList::new, (list, value) -> list.addAll(value), List::addAll);
//或者
//串行流：第一个元素时会通过ArrayList::new获得一个初始化容器a，容器a执行ArrayList::addAll方法操作流中的每一个元素，最后返回容器a，最后容器a就只有一个，所以不会用到List::addAll
ArrayList<String> collect = all.stream().collect(ArrayList::new, ArrayList::addAll, List::addAll);
//并行流：将all中的元素分片处理，每片中第一个元素时会通过ArrayList::new获得一个初始化容器a，
//容器a执行ArrayList::addAll方法操作流中的每一个元素，每片最后会返回容器a，最后多个容器a通过List::addAll方法合并后输出
ArrayList<String> collect = all.parallelStream().collect(ArrayList::new, ArrayList::addAll, List::addAll);
/**
* 第二种方式
* reduce
*/
//new ArrayList<>() 作为初始值加入计算
List<String> reduce = all.stream().reduce(CollUtil.toList("3"), (b, c) -> {
b.addAll(c);
return b;
});//返回["3","1","2"]
//不添加初始值计算
Optional<List<String>> reduce = all.stream().reduce((b, c) -> {
b.addAll(c);
return b;
});//返回["1","2"]
二.Map
Map转换
//假设有数据
Map<String, List<StreamDto>> map = new HashMap<>();
//取出key和集合中第一个对象的时间参数
Map<String, Date> m = map.entrySet().stream()
.collect(Collectors.toMap(b -> b.getKey(), b -> b.getValue().get(0).getBirthDay()));
Map排序
//前5名
Map<String, Long> top5Names = new LinkedHashMap<>();
//假设有数据
List<String> mainQ = new ArrayList<>();
//分组后统计数量
Map<String, Long> countMap = mainQ.stream()
.collect(Collectors.groupingBy(b -> b, Collectors.counting()));
//value排序,然后取出前5名
countMap.entrySet().stream().sorted(Map.Entry.<String, Long>comparingByValue().reversed()).forEachOrdered(b -> {
if (top5Names.size() < 6) {
top5Names.put(b.getKey(), b.getValue());
}
});
//取出全部
Map<String, Long> allMap = countMap.entrySet().stream()
.sorted(Map.Entry.<String, Long>comparingByValue().reversed())
.collect(Collectors.toMap(
Map.Entry::getKey, Map.Entry::getValue, (oldValue, newValue) -> newValue, LinkedHashMap::new
));
//key排序
countMap.entrySet().stream().sorted(Map.Entry.<String, Long>comparingByKey().reversed());
如果是Map
Map过滤
//假设有数据
List<String> mainQ = new ArrayList<>();
Map<String, Long> countMap = mainQ.stream()
.collect(Collectors.groupingBy(b -> b, Collectors.counting()));
//map过滤
countMap.entrySet().stream().filter(b -> b.getKey() != "").collect(HashMap::new, (m, e) -> m.put(e.getKey(), e.getValue()), HashMap::putAll);
Map合并
/**
* @Description: 合并两个map,如果key相同,那么选取时间靠后的value
*/
public static Map<String, Date> concatMap(Map<String, Date> map1, Map<String, Date> map2) {
Map<String, Date> result = Stream.concat(map1.entrySet().stream(), map2.entrySet().stream())
.collect(Collectors.toMap(Map.Entry::getKey, Map.Entry::getValue, (value1, value2) -> {
if (DateUtil.compare(value1, value2) >= 0) {
return value1;
} else {
return value2;
}
}));
return result;
}
Map遍历
/**
* 遍历Map的方式
*/
Map<String,List<ResOrgUserDto>> map =new HashMap<>();
//1.通过Map.keySet遍历key和value
for (String key : map.keySet()) {
System.out.println(map.get(key));
}
//Java8
map.keySet().forEach(k -> {
System.out.println(map.get(k));
});
//2.通过Map.entrySet使用Iterator遍历key和value
while (map.entrySet().iterator().hasNext()){
Map.Entry<String,List<ResOrgUserDto>> entry = map.entrySet().iterator().next();
System.out.println(entry.getValue());
}
//Java8
map.entrySet().iterator().forEachRemaining(m -> System.out.println(m.getValue()));
//3.通过Map.entrySet遍历key和value，在大容量时推荐使用
for (Map.Entry<String, List<ResOrgUserDto>> entry : map.entrySet()) {
System.out.println(entry.getValue());
}
//Java8
map.entrySet().forEach(entry -> {
System.out.println(entry.getValue());
});
//4.通过Map.values()遍历所有的value，但不能遍历key
for (List<ResOrgUserDto> values : map.values()) {
System.out.println(values);
}
//Java8
map.values().forEach(v -> {
System.out.println(v);
});
//5.通过k,v遍历，Java8独有的
map.forEach((k,v) -> {
System.out.println(v);
});
List,Map综合
//前5名名称
List<String> names = top5Names.entrySet().stream().map(b -> b.getKey()).collect(Collectors.toList());
//获取各表前五名平台的最后数据接收时间
Map<String, Date> bidSecLasts = bidSecQ.stream().filter(b -> names.contains(b.getTradeplat()))//筛选出前5的数据
.sorted(Comparator.comparing(BusBidsection::getSubmittimestamp).reversed())//先排序
.collect(Collectors.groupingBy(BusBidsection::getTradeplat))//在分组
//最后取出key和最后数据接收时间
.entrySet().stream().collect(Collectors.toMap(b -> b.getKey(), b -> b.getValue().get(0).getSubmittimestamp()));