Java8的stream API与 C#的 LINQ 拓展方法对比

Posted by Zeusro on March 8, 2018
👈🏻 Select language
  1. 前期准备
    • 定义实体
    • 定义集合
  2. 单集合
    1. 分类筛选
      • 计数(Count)
      • 分组(GroupBy)
      • 匹配的第一项(findFirst/First,FirstOrDefault)
      • 遍历(ForEach)
      • 极值Max/Min
      • 跳过(skip/Skip),截取(limit/Take)
    2. 排序
      • 去重复(Distinct)
      • 升序(sort/OrderBy)
      • 降序(sort/OrderByDescending)
  3. 多集合
    • 交集 list1 ∩ list2
    • 并集list1 ∪ list2
    • 差集list1 - list2
  4. 数据结构转换

为方便初学 Java8/C# 集合操作的人,特意写下这篇文章.

前期准备

C#版

java版

单集合

分类筛选

  • 计数(Count)
1
2
3
            Date time1 = convertLocalDateToTimeZone(LocalDate.of(1990, 1, 1));
            //0
            Long count1 = list1.stream().filter(o -> o.getBirthday().equals(time1)).count();
1
2
3
4
5
6
            int count1 = list1.Where(o => o.Birthday.Equals(new DateTime(1990, 1, 1)) && o.Sex == Sex.Male).Count();
            long count2 = list1.Where(o => o.Birthday.Equals(new DateTime(1990, 1, 1)) && o.Sex == Sex.Male).LongCount();
            /*              
             0
             0
             */
  • 分组(GroupBy)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
            Map<Sex, List<Person>> group1 = list1.stream().collect(Collectors.groupingBy(Person::getSex));
            Iterator it = group1.entrySet().iterator();
            while (it.hasNext()) {
                Map.Entry<Sex, List<Person>> groupByItem = (Map.Entry) it.next();
                Sex sex = groupByItem.getKey();
                out.println(sex);
                groupByItem.getValue().forEach(person -> {
                    out.println(new Gson().toJson(person));
                });
            }
/*
输出结果:
Male
{"height":170,"weight":50,"identifier":"2","address":"北京","birthday":"Feb 1, 1982 12:00:00 AM","hobbies":["吃飯","看電影"],"sex":"Male"}
Female
{"height":165,"weight":50,"identifier":"1","address":"北京","birthday":"Jan 1, 1981 12:00:00 AM","hobbies":["吃飯","逛街"],"sex":"Female"}
X
{"height":170,"weight":50,"identifier":"3","address":"北京","birthday":"Mar 1, 1983 12:00:00 AM","hobbies":["吃飯","上網"],"sex":"X"}
 */
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
var group1 = list1.GroupBy(o => o.Sex);
            //当我们使用 GroupBy() 扩展方法时,使用了延迟执行。 这意味着,当你遍历集合的时候,下一个要出现的项目可能会或者可能不会被加载。 这是一个很大的性能改进,但它会引起有趣的副作用。
            list1.RemoveAll(o => o.Sex == Sex.X);//定义 groupby 集合后对原集合进行修改,会发现group1里面已经没了 Sex=X的分组
            foreach (var groupByItem in group1)
            {
                Sex sex = groupByItem.Key;
                System.Console.WriteLine(sex);
                foreach (Person person in groupByItem)
                {                    
                    System.Console.WriteLine(JsonConvert.SerializeObject(person));
                }
            }
            /*
            输出结果:
            {"Height":165,"Weight":50,"Birthday":"1981-01-01T00:00:00","Hobbies":["吃飯","逛街"],"Identifier":"1","Address":"北京","Sex":2}
            Male
            {"Height":170,"Weight":50,"Birthday":"1982-02-01T00:00:00","Hobbies":["吃飯","看電影"],"Identifier":"2","Address":"北京","Sex":1}
            Female
            {"Height":165,"Weight":50,"Birthday":"1981-01-01T00:00:00","Hobbies":["吃飯","逛街"],"Identifier":"1","Address":"北京","Sex":2}
            Male
            {"Height":170,"Weight":50,"Birthday":"1982-02-01T00:00:00","Hobbies":["吃飯","看電影"],"Identifier":"2","Address":"北京","Sex":1}
             */
            //该 ToLookup() 方法创建一个类似 字典(Dictionary ) 的列表List, 但是它是一个新的 .NET Collection 叫做 lookup。 Lookup,不像Dictionary, 是不可改变的。 这意味着一旦你创建一个lookup, 你不能添加或删除元素。
            var group2 = list1.ToLookup(o => o.Sex);
            foreach (var groupByItem in group2)
            {
                Sex sex = groupByItem.Key;
                foreach (Person person in groupByItem)
                {
                    System.Console.WriteLine(sex);
                    System.Console.WriteLine(JsonConvert.SerializeObject(person));
                }

            }
            /*
            输出结果:            
            {"Height":165,"Weight":50,"Birthday":"1981-01-01T00:00:00","Hobbies":["吃飯","逛街"],"Identifier":"1","Address":"北京","Sex":3}
            {"Height":170,"Weight":50,"Birthday":"1982-02-01T00:00:00","Hobbies":["吃飯","看電影"],"Identifier":"2","Address":"北京","Sex":3}
             */

与此对比,stream没有RemoveAll的操作

  • 匹配的第一项(findFirst/First,FirstOrDefault)
1
2
3
4
5
 Person after90 = list1.stream()
                    .filter(o -> o.getBirthday().after(convertLocalDateToTimeZone(LocalDate.of(1990, 1, 1))))
                    .findFirst()
                    .orElse(null);
            // null
1
2
3
4
5
            var after90 = list1.Where(o => o.Birthday >= new DateTime(1990, 1, 1)).First();//如果结果为空,将会导致异常,所以一般极少使用该方法
            //An unhandled exception of type 'System.InvalidOperationException' occurred in System.Linq.dll: 'Sequence contains no elements'
            after90 = list1.Where(o => o.Birthday >= new DateTime(1990, 1, 1)).FirstOrDefault();
            var after00 = list1.Where(o => o.Birthday >= new DateTime(2000, 1, 1)).FirstOrDefault();
  • 遍历(ForEach)
1
2
3
4
5
6
7
8
9
10
11
12
13
            list1.stream().forEach(o -> {
                //在ForEach當中可對集合進行操作
                o.setSex(Sex.X);
            });
            list1.forEach(o -> {
                out.println(new Gson().toJson(o));
            });
/*
{"height":165,"weight":50,"identifier":"1","address":"北京","birthday":"Jan 1, 1981 12:00:00 AM","hobbies":["吃飯","逛街"],"sex":"X"}
{"height":170,"weight":50,"identifier":"2","address":"北京","birthday":"Feb 1, 1982 12:00:00 AM","hobbies":["吃飯","看電影"],"sex":"X"}
{"height":170,"weight":50,"identifier":"3","address":"北京","birthday":"Mar 1, 1983 12:00:00 AM","hobbies":["吃飯","上網"],"sex":"X"}
 */
1
2
3
4
5
6
7
8
9
10
            list1.ForEach(item =>
            {
                //在ForEach當中可對集合進行操作
                item.Sex = Sex.X;
            });
            list1.ForEach(item =>
           {
               System.Console.WriteLine(JsonConvert.SerializeObject(item));
           });
  • 极值Max/Min
1
2
3
4
5
6
7
8
9
10
11
12
13
14
            //IntStream的max方法返回的是OptionalInt,要先判断有没有值再读取值.isPresent=false 时直接getAsInt会报错.mapToLong,mapToDouble同理
            OptionalInt maxHeightOption = list1.stream().mapToInt(Person::getHeight).max();
            //字符串拼接、数值的 sum、min、max、average 都是特殊的 reduce。
            //当集合为长度0的集合时会返回起始值Integer.MIN_VALUE,起始值也不能乱传,个中缘由我暂不清楚
            int maxHeight = list1.stream().mapToInt(Person::getHeight).reduce(Integer.MIN_VALUE, Integer::max);
            out.println(maxHeight);
            if (maxHeightOption.isPresent()) {
                maxHeight = maxHeightOption.getAsInt();
                out.println(maxHeight);
            }
            //mapToInt参数的2种写法都一样,我比较喜欢以下写法,但是 idea 会报 warning
            OptionalInt minWeightOption = list1.stream().mapToInt(o -> o.getHeight()).min();
            int minWeight = list1.stream().mapToInt(o -> o.getHeight()).reduce(Integer.MAX_VALUE, Integer::min);
1
2
3
4
            int maxHeight = list1.Select(o => o.Height).Max();
            //同 list1.Max(o => o.Height);
            int minWeight = list1.Min(o => o.Weight);            
  • 跳过(skip/Skip),截取(limit/Take)
1
2
3
            //skip和 limit参数都是long, 这个要注意
            list1.stream().skip(1L).limit(2L);

排序

  • 去重复(Distinct)
1
list1.stream().map(Person::getIdentifier).distinct();
1
list1.Select(o=>o.Identifier).Distinct();
1
 list1.Skip(1).Take(2);
  • 升序(sort/OrderBy)
1
2
3
4
5
6
            out.println("------------------------------------|升序|------------------------------------");
            list1 = list1.stream().sorted(Comparator.comparing(Person::getBirthday)).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));
            list1 = list1.stream().sorted((left, right) -> left.getBirthday().compareTo(right.getBirthday())).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));
1
2
            //升序
            list1 = list1.OrderBy(o => o.Birthday).ToList();
  • 降序(sort/OrderByDescending)
1
2
3
4
5
6
7
            out.println("------------------------------------|降序|------------------------------------");
            list1 = list1.stream().sorted(Comparator.comparing(Person::getBirthday).reversed()).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));
            list1 = list1.stream().sorted((left, right) -> right.getBirthday().compareTo(left.getBirthday())).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));

1
2
3
            //降序
            list1 = list1.OrderByDescending(o => o.Birthday).ToList();

多集合

  • 交集 list1 ∩ list2
1
2
3
            out.println("------------------------------------|交集 list1 ∩ list2|------------------------------------");
            list1.stream().filter(o -> list2.contains(o)).collect(Collectors.toList());
1
2
3
4
5
//连接,下面表示把 list1和 list2当中相同身份证号的取出来,生成一个新的集合            
            //实际上, join 有另外的用法,类似 sqlserver 里面的多表连接,将不同数据源结合到一起,生成新的数据结构
            var intersect = list1.Join(list2, o => o.Identifier, o => o.Identifier, (a, b) => a).ToList();
            //交集 list1 ∩ list2                        
            intersect = list1.Intersect(list2).ToList();
  • 并集list1 ∪ list2
1
2
3
            out.println("------------------------------------|并集list1 ∪ list2 |------------------------------------");
            list1.addAll(list2);
1
2
3
            //并集list1 ∪ list2 
            var union = list1.Union(list2).ToList();
  • 差集list1 - list2
1
2
3
            out.println("------------------------------------|差集list1 - list2|------------------------------------");
            list1.stream().filter(item1 -> !list2.contains(item1)).collect(Collectors.toList());
1
2
            //差集list1 - list2
            var except = list1.Except(list2).ToList();

数据结构转换

1
2
3
4
5
6
7
            out.println("------------------------------------|数据结构转换|------------------------------------");
            List<Person> list3 = list1.stream().filter(o -> true).collect(Collectors.toList());
            ArrayList<Person> list4 = list1.stream().filter(o -> true).collect(Collectors.toCollection(ArrayList::new));
            Set<Person> list5 = list1.stream().filter(o -> true).collect(Collectors.toSet());
            Object[] list6 = list1.stream().toArray();
            Person[] list7 = list1.stream().toArray(Person[]::new);
1
2
3
4
5
            //数据结构转换
            list1.ToArray();
            //注意如果 key 重复,ToDictionary会导致出错
            list1.ToDictionary(o => o.Identifier, o => o);
            list1.ToHashSet();
  1. Preparation
    • Define Entity
    • Define Collection
  2. Single Collection
    1. Filtering and Classification
      • Count
      • GroupBy
      • First Matching Item (findFirst/First, FirstOrDefault)
      • ForEach
      • Max/Min
      • Skip, Take (limit)
    2. Sorting
      • Distinct
      • Ascending (sort/OrderBy)
      • Descending (sort/OrderByDescending)
  3. Multiple Collections
    • Intersection list1 ∩ list2
    • Union list1 ∪ list2
    • Difference list1 - list2
  4. Data Structure Conversion

This article is written specifically to help beginners learning Java8/C# collection operations.

Preparation

C# Version

Java Version

Single Collection

Filtering and Classification

  • Count
1
2
3
            Date time1 = convertLocalDateToTimeZone(LocalDate.of(1990, 1, 1));
            //0
            Long count1 = list1.stream().filter(o -> o.getBirthday().equals(time1)).count();
1
2
3
4
5
6
            int count1 = list1.Where(o => o.Birthday.Equals(new DateTime(1990, 1, 1)) && o.Sex == Sex.Male).Count();
            long count2 = list1.Where(o => o.Birthday.Equals(new DateTime(1990, 1, 1)) && o.Sex == Sex.Male).LongCount();
            /*              
             0
             0
             */
  • GroupBy
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
            Map<Sex, List<Person>> group1 = list1.stream().collect(Collectors.groupingBy(Person::getSex));
            Iterator it = group1.entrySet().iterator();
            while (it.hasNext()) {
                Map.Entry<Sex, List<Person>> groupByItem = (Map.Entry) it.next();
                Sex sex = groupByItem.getKey();
                out.println(sex);
                groupByItem.getValue().forEach(person -> {
                    out.println(new Gson().toJson(person));
                });
            }
/*
Output:
Male
{"height":170,"weight":50,"identifier":"2","address":"北京","birthday":"Feb 1, 1982 12:00:00 AM","hobbies":["吃飯","看電影"],"sex":"Male"}
Female
{"height":165,"weight":50,"identifier":"1","address":"北京","birthday":"Jan 1, 1981 12:00:00 AM","hobbies":["吃飯","逛街"],"sex":"Female"}
X
{"height":170,"weight":50,"identifier":"3","address":"北京","birthday":"Mar 1, 1983 12:00:00 AM","hobbies":["吃飯","上網"],"sex":"X"}
 */
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
var group1 = list1.GroupBy(o => o.Sex);
            //When we use the GroupBy() extension method, deferred execution is used. This means that when you iterate through the collection, the next item to appear may or may not be loaded. This is a big performance improvement, but it can cause interesting side effects.
            list1.RemoveAll(o => o.Sex == Sex.X);//After defining the groupby collection, modifying the original collection will find that the Sex=X group is already gone from group1
            foreach (var groupByItem in group1)
            {
                Sex sex = groupByItem.Key;
                System.Console.WriteLine(sex);
                foreach (Person person in groupByItem)
                {                    
                    System.Console.WriteLine(JsonConvert.SerializeObject(person));
                }
            }
            /*
            Output:
            {"Height":165,"Weight":50,"Birthday":"1981-01-01T00:00:00","Hobbies":["吃飯","逛街"],"Identifier":"1","Address":"北京","Sex":2}
            Male
            {"Height":170,"Weight":50,"Birthday":"1982-02-01T00:00:00","Hobbies":["吃飯","看電影"],"Identifier":"2","Address":"北京","Sex":1}
            Female
            {"Height":165,"Weight":50,"Birthday":"1981-01-01T00:00:00","Hobbies":["吃飯","逛街"],"Identifier":"1","Address":"北京","Sex":2}
            Male
            {"Height":170,"Weight":50,"Birthday":"1982-02-01T00:00:00","Hobbies":["吃飯","看電影"],"Identifier":"2","Address":"北京","Sex":1}
             */
            //The ToLookup() method creates a Dictionary-like List, but it's a new .NET Collection called lookup. Lookup, unlike Dictionary, is immutable. This means once you create a lookup, you cannot add or remove elements.
            var group2 = list1.ToLookup(o => o.Sex);
            foreach (var groupByItem in group2)
            {
                Sex sex = groupByItem.Key;
                foreach (Person person in groupByItem)
                {
                    System.Console.WriteLine(sex);
                    System.Console.WriteLine(JsonConvert.SerializeObject(person));
                }

            }
            /*
            Output:            
            {"Height":165,"Weight":50,"Birthday":"1981-01-01T00:00:00","Hobbies":["吃飯","逛街"],"Identifier":"1","Address":"北京","Sex":3}
            {"Height":170,"Weight":50,"Birthday":"1982-02-01T00:00:00","Hobbies":["吃飯","看電影"],"Identifier":"2","Address":"北京","Sex":3}
             */

In contrast, stream doesn’t have a RemoveAll operation

  • First Matching Item (findFirst/First, FirstOrDefault)
1
2
3
4
5
 Person after90 = list1.stream()
                    .filter(o -> o.getBirthday().after(convertLocalDateToTimeZone(LocalDate.of(1990, 1, 1))))
                    .findFirst()
                    .orElse(null);
            // null
1
2
3
4
5
            var after90 = list1.Where(o => o.Birthday >= new DateTime(1990, 1, 1)).First();//If the result is empty, it will cause an exception, so this method is rarely used
            //An unhandled exception of type 'System.InvalidOperationException' occurred in System.Linq.dll: 'Sequence contains no elements'
            after90 = list1.Where(o => o.Birthday >= new DateTime(1990, 1, 1)).FirstOrDefault();
            var after00 = list1.Where(o => o.Birthday >= new DateTime(2000, 1, 1)).FirstOrDefault();
  • ForEach
1
2
3
4
5
6
7
8
9
10
11
12
13
            list1.stream().forEach(o -> {
                //Can operate on the collection within ForEach
                o.setSex(Sex.X);
            });
            list1.forEach(o -> {
                out.println(new Gson().toJson(o));
            });
/*
{"height":165,"weight":50,"identifier":"1","address":"北京","birthday":"Jan 1, 1981 12:00:00 AM","hobbies":["吃飯","逛街"],"sex":"X"}
{"height":170,"weight":50,"identifier":"2","address":"北京","birthday":"Feb 1, 1982 12:00:00 AM","hobbies":["吃飯","看電影"],"sex":"X"}
{"height":170,"weight":50,"identifier":"3","address":"北京","birthday":"Mar 1, 1983 12:00:00 AM","hobbies":["吃飯","上網"],"sex":"X"}
 */
1
2
3
4
5
6
7
8
9
10
            list1.ForEach(item =>
            {
                //Can operate on the collection within ForEach
                item.Sex = Sex.X;
            });
            list1.ForEach(item =>
           {
               System.Console.WriteLine(JsonConvert.SerializeObject(item));
           });
  • Max/Min
1
2
3
4
5
6
7
8
9
10
11
12
13
14
            //IntStream's max method returns OptionalInt, must check if there's a value before reading. getAsInt will error when isPresent=false. mapToLong, mapToDouble are similar
            OptionalInt maxHeightOption = list1.stream().mapToInt(Person::getHeight).max();
            //String concatenation, sum, min, max, average of numbers are all special reduces.
            //When the collection is a length 0 collection, it returns the starting value Integer.MIN_VALUE. The starting value cannot be passed randomly, I'm not clear on the reason yet
            int maxHeight = list1.stream().mapToInt(Person::getHeight).reduce(Integer.MIN_VALUE, Integer::max);
            out.println(maxHeight);
            if (maxHeightOption.isPresent()) {
                maxHeight = maxHeightOption.getAsInt();
                out.println(maxHeight);
            }
            //Both ways of writing mapToInt parameters are the same. I prefer the following, but idea will report a warning
            OptionalInt minWeightOption = list1.stream().mapToInt(o -> o.getHeight()).min();
            int minWeight = list1.stream().mapToInt(o -> o.getHeight()).reduce(Integer.MAX_VALUE, Integer::min);
1
2
3
4
            int maxHeight = list1.Select(o => o.Height).Max();
            //Same as list1.Max(o => o.Height);
            int minWeight = list1.Min(o => o.Weight);            
  • Skip, Take (limit)
1
2
3
            //Both skip and limit parameters are long, note this
            list1.stream().skip(1L).limit(2L);

Sorting

  • Distinct
1
list1.stream().map(Person::getIdentifier).distinct();
1
list1.Select(o=>o.Identifier).Distinct();
1
 list1.Skip(1).Take(2);
  • Ascending (sort/OrderBy)
1
2
3
4
5
6
            out.println("------------------------------------|Ascending|------------------------------------");
            list1 = list1.stream().sorted(Comparator.comparing(Person::getBirthday)).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));
            list1 = list1.stream().sorted((left, right) -> left.getBirthday().compareTo(right.getBirthday())).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));
1
2
            //Ascending
            list1 = list1.OrderBy(o => o.Birthday).ToList();
  • Descending (sort/OrderByDescending)
1
2
3
4
5
6
7
            out.println("------------------------------------|Descending|------------------------------------");
            list1 = list1.stream().sorted(Comparator.comparing(Person::getBirthday).reversed()).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));
            list1 = list1.stream().sorted((left, right) -> right.getBirthday().compareTo(left.getBirthday())).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));

1
2
3
            //Descending
            list1 = list1.OrderByDescending(o => o.Birthday).ToList();

Multiple Collections

  • Intersection list1 ∩ list2
1
2
3
            out.println("------------------------------------|Intersection list1 ∩ list2|------------------------------------");
            list1.stream().filter(o -> list2.contains(o)).collect(Collectors.toList());
1
2
3
4
5
//Join, below means take out items with the same ID number from list1 and list2, generate a new collection            
            //Actually, join has other uses, similar to multi-table joins in sqlserver, combining different data sources together to generate new data structures
            var intersect = list1.Join(list2, o => o.Identifier, o => o.Identifier, (a, b) => a).ToList();
            //Intersection list1 ∩ list2                        
            intersect = list1.Intersect(list2).ToList();
  • Union list1 ∪ list2
1
2
3
            out.println("------------------------------------|Union list1 ∪ list2 |------------------------------------");
            list1.addAll(list2);
1
2
3
            //Union list1 ∪ list2 
            var union = list1.Union(list2).ToList();
  • Difference list1 - list2
1
2
3
            out.println("------------------------------------|Difference list1 - list2|------------------------------------");
            list1.stream().filter(item1 -> !list2.contains(item1)).collect(Collectors.toList());
1
2
            //Difference list1 - list2
            var except = list1.Except(list2).ToList();

Data Structure Conversion

1
2
3
4
5
6
7
            out.println("------------------------------------|Data Structure Conversion|------------------------------------");
            List<Person> list3 = list1.stream().filter(o -> true).collect(Collectors.toList());
            ArrayList<Person> list4 = list1.stream().filter(o -> true).collect(Collectors.toCollection(ArrayList::new));
            Set<Person> list5 = list1.stream().filter(o -> true).collect(Collectors.toSet());
            Object[] list6 = list1.stream().toArray();
            Person[] list7 = list1.stream().toArray(Person[]::new);
1
2
3
4
5
            //Data Structure Conversion
            list1.ToArray();
            //Note if key is duplicate, ToDictionary will cause an error
            list1.ToDictionary(o => o.Identifier, o => o);
            list1.ToHashSet();
  1. Подготовка
    • Определение сущности
    • Определение коллекции
  2. Одна коллекция
    1. Классификация и фильтрация
      • Подсчет (Count)
      • Группировка (GroupBy)
      • Первый совпадающий элемент (findFirst/First, FirstOrDefault)
      • Обход (ForEach)
      • Экстремальные значения Max/Min
      • Пропуск (skip/Skip), извлечение (limit/Take)
    2. Сортировка
      • Удаление дубликатов (Distinct)
      • По возрастанию (sort/OrderBy)
      • По убыванию (sort/OrderByDescending)
  3. Несколько коллекций
    • Пересечение list1 ∩ list2
    • Объединение list1 ∪ list2
    • Разность list1 - list2
  4. Преобразование структуры данных

Эта статья написана специально для помощи начинающим изучать операции с коллекциями Java8/C#.

Подготовка

Версия C#

Версия Java

Одна коллекция

Классификация и фильтрация

  • Подсчет (Count)
1
2
3
            Date time1 = convertLocalDateToTimeZone(LocalDate.of(1990, 1, 1));
            //0
            Long count1 = list1.stream().filter(o -> o.getBirthday().equals(time1)).count();
1
2
3
4
5
6
            int count1 = list1.Where(o => o.Birthday.Equals(new DateTime(1990, 1, 1)) && o.Sex == Sex.Male).Count();
            long count2 = list1.Where(o => o.Birthday.Equals(new DateTime(1990, 1, 1)) && o.Sex == Sex.Male).LongCount();
            /*              
             0
             0
             */
  • Группировка (GroupBy)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
            Map<Sex, List<Person>> group1 = list1.stream().collect(Collectors.groupingBy(Person::getSex));
            Iterator it = group1.entrySet().iterator();
            while (it.hasNext()) {
                Map.Entry<Sex, List<Person>> groupByItem = (Map.Entry) it.next();
                Sex sex = groupByItem.getKey();
                out.println(sex);
                groupByItem.getValue().forEach(person -> {
                    out.println(new Gson().toJson(person));
                });
            }
/*
Вывод:
Male
{"height":170,"weight":50,"identifier":"2","address":"北京","birthday":"Feb 1, 1982 12:00:00 AM","hobbies":["吃飯","看電影"],"sex":"Male"}
Female
{"height":165,"weight":50,"identifier":"1","address":"北京","birthday":"Jan 1, 1981 12:00:00 AM","hobbies":["吃飯","逛街"],"sex":"Female"}
X
{"height":170,"weight":50,"identifier":"3","address":"北京","birthday":"Mar 1, 1983 12:00:00 AM","hobbies":["吃飯","上網"],"sex":"X"}
 */
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
var group1 = list1.GroupBy(o => o.Sex);
            //Когда мы используем метод расширения GroupBy(), используется отложенное выполнение. Это означает, что при итерации по коллекции следующий элемент может быть или не быть загружен. Это большое улучшение производительности, но может вызвать интересные побочные эффекты.
            list1.RemoveAll(o => o.Sex == Sex.X);//После определения коллекции groupby, изменение исходной коллекции обнаружит, что группа Sex=X уже отсутствует в group1
            foreach (var groupByItem in group1)
            {
                Sex sex = groupByItem.Key;
                System.Console.WriteLine(sex);
                foreach (Person person in groupByItem)
                {                    
                    System.Console.WriteLine(JsonConvert.SerializeObject(person));
                }
            }
            /*
            Вывод:
            {"Height":165,"Weight":50,"Birthday":"1981-01-01T00:00:00","Hobbies":["吃飯","逛街"],"Identifier":"1","Address":"北京","Sex":2}
            Male
            {"Height":170,"Weight":50,"Birthday":"1982-02-01T00:00:00","Hobbies":["吃飯","看電影"],"Identifier":"2","Address":"北京","Sex":1}
            Female
            {"Height":165,"Weight":50,"Birthday":"1981-01-01T00:00:00","Hobbies":["吃飯","逛街"],"Identifier":"1","Address":"北京","Sex":2}
            Male
            {"Height":170,"Weight":50,"Birthday":"1982-02-01T00:00:00","Hobbies":["吃飯","看電影"],"Identifier":"2","Address":"北京","Sex":1}
             */
            //Метод ToLookup() создает список, похожий на Dictionary, но это новая коллекция .NET под названием lookup. Lookup, в отличие от Dictionary, неизменяем. Это означает, что после создания lookup вы не можете добавлять или удалять элементы.
            var group2 = list1.ToLookup(o => o.Sex);
            foreach (var groupByItem in group2)
            {
                Sex sex = groupByItem.Key;
                foreach (Person person in groupByItem)
                {
                    System.Console.WriteLine(sex);
                    System.Console.WriteLine(JsonConvert.SerializeObject(person));
                }

            }
            /*
            Вывод:            
            {"Height":165,"Weight":50,"Birthday":"1981-01-01T00:00:00","Hobbies":["吃飯","逛街"],"Identifier":"1","Address":"北京","Sex":3}
            {"Height":170,"Weight":50,"Birthday":"1982-02-01T00:00:00","Hobbies":["吃飯","看電影"],"Identifier":"2","Address":"北京","Sex":3}
             */

В отличие от этого, stream не имеет операции RemoveAll

  • Первый совпадающий элемент (findFirst/First, FirstOrDefault)
1
2
3
4
5
 Person after90 = list1.stream()
                    .filter(o -> o.getBirthday().after(convertLocalDateToTimeZone(LocalDate.of(1990, 1, 1))))
                    .findFirst()
                    .orElse(null);
            // null
1
2
3
4
5
            var after90 = list1.Where(o => o.Birthday >= new DateTime(1990, 1, 1)).First();//Если результат пуст, это вызовет исключение, поэтому этот метод редко используется
            //An unhandled exception of type 'System.InvalidOperationException' occurred in System.Linq.dll: 'Sequence contains no elements'
            after90 = list1.Where(o => o.Birthday >= new DateTime(1990, 1, 1)).FirstOrDefault();
            var after00 = list1.Where(o => o.Birthday >= new DateTime(2000, 1, 1)).FirstOrDefault();
  • Обход (ForEach)
1
2
3
4
5
6
7
8
9
10
11
12
13
            list1.stream().forEach(o -> {
                //Можно работать с коллекцией внутри ForEach
                o.setSex(Sex.X);
            });
            list1.forEach(o -> {
                out.println(new Gson().toJson(o));
            });
/*
{"height":165,"weight":50,"identifier":"1","address":"北京","birthday":"Jan 1, 1981 12:00:00 AM","hobbies":["吃飯","逛街"],"sex":"X"}
{"height":170,"weight":50,"identifier":"2","address":"北京","birthday":"Feb 1, 1982 12:00:00 AM","hobbies":["吃飯","看電影"],"sex":"X"}
{"height":170,"weight":50,"identifier":"3","address":"北京","birthday":"Mar 1, 1983 12:00:00 AM","hobbies":["吃飯","上網"],"sex":"X"}
 */
1
2
3
4
5
6
7
8
9
10
            list1.ForEach(item =>
            {
                //Можно работать с коллекцией внутри ForEach
                item.Sex = Sex.X;
            });
            list1.ForEach(item =>
           {
               System.Console.WriteLine(JsonConvert.SerializeObject(item));
           });
  • Экстремальные значения Max/Min
1
2
3
4
5
6
7
8
9
10
11
12
13
14
            //Метод max IntStream возвращает OptionalInt, нужно проверить, есть ли значение, прежде чем читать. getAsInt выдаст ошибку, когда isPresent=false. mapToLong, mapToDouble аналогичны
            OptionalInt maxHeightOption = list1.stream().mapToInt(Person::getHeight).max();
            //Конкатенация строк, sum, min, max, average чисел — все это специальные reduce.
            //Когда коллекция имеет длину 0, возвращается начальное значение Integer.MIN_VALUE. Начальное значение нельзя передавать случайно, причина мне пока не ясна
            int maxHeight = list1.stream().mapToInt(Person::getHeight).reduce(Integer.MIN_VALUE, Integer::max);
            out.println(maxHeight);
            if (maxHeightOption.isPresent()) {
                maxHeight = maxHeightOption.getAsInt();
                out.println(maxHeight);
            }
            //Оба способа записи параметров mapToInt одинаковы. Я предпочитаю следующий, но idea выдаст предупреждение
            OptionalInt minWeightOption = list1.stream().mapToInt(o -> o.getHeight()).min();
            int minWeight = list1.stream().mapToInt(o -> o.getHeight()).reduce(Integer.MAX_VALUE, Integer::min);
1
2
3
4
            int maxHeight = list1.Select(o => o.Height).Max();
            //То же, что list1.Max(o => o.Height);
            int minWeight = list1.Min(o => o.Weight);            
  • Пропуск (skip/Skip), извлечение (limit/Take)
1
2
3
            //Оба параметра skip и limit имеют тип long, обратите на это внимание
            list1.stream().skip(1L).limit(2L);

Сортировка

  • Удаление дубликатов (Distinct)
1
list1.stream().map(Person::getIdentifier).distinct();
1
list1.Select(o=>o.Identifier).Distinct();
1
 list1.Skip(1).Take(2);
  • По возрастанию (sort/OrderBy)
1
2
3
4
5
6
            out.println("------------------------------------|По возрастанию|------------------------------------");
            list1 = list1.stream().sorted(Comparator.comparing(Person::getBirthday)).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));
            list1 = list1.stream().sorted((left, right) -> left.getBirthday().compareTo(right.getBirthday())).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));
1
2
            //По возрастанию
            list1 = list1.OrderBy(o => o.Birthday).ToList();
  • По убыванию (sort/OrderByDescending)
1
2
3
4
5
6
7
            out.println("------------------------------------|По убыванию|------------------------------------");
            list1 = list1.stream().sorted(Comparator.comparing(Person::getBirthday).reversed()).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));
            list1 = list1.stream().sorted((left, right) -> right.getBirthday().compareTo(left.getBirthday())).collect(Collectors.toList());
            out.println(new Gson().toJson(list1));

1
2
3
            //По убыванию
            list1 = list1.OrderByDescending(o => o.Birthday).ToList();

Несколько коллекций

  • Пересечение list1 ∩ list2
1
2
3
            out.println("------------------------------------|Пересечение list1 ∩ list2|------------------------------------");
            list1.stream().filter(o -> list2.contains(o)).collect(Collectors.toList());
1
2
3
4
5
//Соединение, ниже означает извлечение элементов с одинаковым номером ID из list1 и list2, создание новой коллекции            
            //На самом деле, join имеет другое использование, подобное многотабличным соединениям в sqlserver, объединяя разные источники данных вместе для создания новых структур данных
            var intersect = list1.Join(list2, o => o.Identifier, o => o.Identifier, (a, b) => a).ToList();
            //Пересечение list1 ∩ list2                        
            intersect = list1.Intersect(list2).ToList();
  • Объединение list1 ∪ list2
1
2
3
            out.println("------------------------------------|Объединение list1 ∪ list2 |------------------------------------");
            list1.addAll(list2);
1
2
3
            //Объединение list1 ∪ list2 
            var union = list1.Union(list2).ToList();
  • Разность list1 - list2
1
2
3
            out.println("------------------------------------|Разность list1 - list2|------------------------------------");
            list1.stream().filter(item1 -> !list2.contains(item1)).collect(Collectors.toList());
1
2
            //Разность list1 - list2
            var except = list1.Except(list2).ToList();

Преобразование структуры данных

1
2
3
4
5
6
7
            out.println("------------------------------------|Преобразование структуры данных|------------------------------------");
            List<Person> list3 = list1.stream().filter(o -> true).collect(Collectors.toList());
            ArrayList<Person> list4 = list1.stream().filter(o -> true).collect(Collectors.toCollection(ArrayList::new));
            Set<Person> list5 = list1.stream().filter(o -> true).collect(Collectors.toSet());
            Object[] list6 = list1.stream().toArray();
            Person[] list7 = list1.stream().toArray(Person[]::new);
1
2
3
4
5
            //Преобразование структуры данных
            list1.ToArray();
            //Обратите внимание, если ключ дублируется, ToDictionary вызовет ошибку
            list1.ToDictionary(o => o.Identifier, o => o);
            list1.ToHashSet();