Flink的DataSet基本算子总结_flink dateset 算子详解

最新推荐文章于 2024-12-23 09:14:48 发布

2401_87034401

最新推荐文章于 2024-12-23 09:14:48 发布

阅读量965

点赞数 29

文章标签： flink c# 大数据

本文链接：https://blog.csdn.net/2401_87034401/article/details/142134384

版权

/* new MapPartitionFunction<String, String>
第一个String：表示分区中的数据元素类型
第二个String：表示处理后的数据元素类型*/
DataSet mapPartitionData = text.mapPartition(new MapPartitionFunction<String, String>() {

public void mapPartition(Iterable<String> values, Collector<String> out) throws Exception {
	//针对分区进行操作的好处是：比如要进行数据库的操作，一个分区只需要创建一个Connection
	//values中保存了一个分区的数据
	 Iterator<String> it = values.iterator();
	while (it.hasNext()) {
		String next = it.next();
		String[] split = next.split(" ");
		for (String word : split) {
			out.collect(word);
		}
	}
	//关闭链接
}

});
mapPartitionData.print();


## 2、Filter与Distinct

//获取运行环境
ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();

ArrayList data = new ArrayList();
data.add(“I love Beijing”);
data.add(“I love China”);
data.add(“Beijing is the capital of China”);
DataSource text = env.fromCollection(data);

DataSet flatMapData = text.flatMap(new FlatMapFunction<String, String>() {

public void flatMap(String data, Collector<String> collection) throws Exception {
	String[] words = data.split(" ");
	for(String w:words){
		collection.collect(w);
	}
}

});

//去掉重复的单词
flatMapData.distinct().print();
System.out.println(“*********************”);

//选出长度大于3的单词
flatMapData.filter(new FilterFunction() {

public boolean filter(String word) throws Exception {
	int length = word.length();
	return length>3?true:false;
}

}).print();


## 3、Join操作

//获取运行的环境
ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();

//创建第一张表：用户ID 姓名
ArrayList<Tuple2<Integer, String>> data1 = new ArrayList<Tuple2<Integer,String>>();
data1.add(new Tuple2(1,“Tom”));
data1.add(new Tuple2(2,“Mike”));
data1.add(new Tuple2(3,“Mary”));
data1.add(new Tuple2(4,“Jone”));
//创建第二张表：用户ID 所在的城市
ArrayList<Tuple2<Integer, String>> data2 = new ArrayList<Tuple2<Integer,String>>();
data2.add(new Tuple2(1,“北京”));
data2.add(new Tuple2(2,“上海”));
data2.add(new Tuple2(3,“广州”));
data2.add(new Tuple2(4,“重庆”));

//实现join的多表查询：用户ID 姓名所在的程序
DataSet<Tuple2<Integer, String>> table1 = env.fromCollection(data1);
DataSet<Tuple2<Integer, String>> table2 = env.fromCollection(data2);

table1.join(table2).where(0).equalTo(0)
/*第一个Tuple2<Integer,String>：表示第一张表

第二个Tuple2<Integer,String>：表示第二张表
Tuple3<Integer,String, String>：多表join连接查询后的返回结果 */
.with(new JoinFunction<Tuple2<Integer,String>, Tuple2<Integer,String>, Tuple3<Integer,String, String>>() {
public Tuple3<Integer, String, String> join(Tuple2<Integer, String> table1,
Tuple2<Integer, String> table2) throws Exception {
return new Tuple3<Integer, String, String>(table1.f0,table1.f1,table2.f1);
} }).print();


## 4、笛卡尔积

ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();

//创建第二张表：用户ID 所在的城市
ArrayList<Tuple2<Integer, String>> data2 = new ArrayList<Tuple2<Integer,String>>();
data2.add(new Tuple2(1,“北京”));
data2.add(new Tuple2(2,“上海”));
data2.add(new Tuple2(3,“广州”));
data2.add(new Tuple2(4,“重庆”));

//实现join的多表查询：用户ID 姓名所在的程序
DataSet<Tuple2<Integer, String>> table1 = env.fromCollection(data1);
DataSet<Tuple2<Integer, String>> table2 = env.fromCollection(data2);

//生成笛卡尔积
table1.cross(table2).print();


## 5、First-N

ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();

//这里的数据是：员工姓名、薪水、部门号
DataSet<Tuple3<String, Integer,Integer>> grade =
env.fromElements(new Tuple3<String, Integer,Integer>(“Tom”,1000,10),
new Tuple3<String, Integer,Integer>(“Mary”,1500,20),
new Tuple3<String, Integer,Integer>(“Mike”,1200,30),
new Tuple3<String, Integer,Integer>(“Jerry”,2000,10));

//按照插入顺序取前三条记录
grade.first(3).print();
System.out.println(“**********************”);

//先按照部门号排序，在按照薪水排序
grade.sortPartition(2, Order.ASCENDING).sortPartition(1, Order.ASCENDING).print();
System.out.println(“**********************”);

//按照部门号分组，求每组的第一条记录
grade.groupBy(2).first(1).print();


## 6、外链接操作

ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();

//创建第一张表：用户ID 姓名
ArrayList<Tuple2<Integer, String>> data1 = new ArrayList<Tuple2<Integer,String>>();
data1.add(new Tuple2(1,“Tom”));
data1.add(new Tuple2(3,“Mary”));
data1.add(new Tuple2(4,“Jone”));

//实现join的多表查询：用户ID 姓名所在的程序
DataSet<Tuple2<Integer, String>> table1 = env.fromCollection(data1);
DataSet<Tuple2<Integer, String>> table2 = env.fromCollection(data2);

//左外连接
table1.leftOuterJoin(table2).where(0).equalTo(0)
.with(new JoinFunction<Tuple2<Integer,String>, Tuple2<Integer,String>, Tuple3<Integer,String,String>>() {

	public Tuple3<Integer, String, String> join(Tuple2<Integer, String> table1,
			Tuple2<Integer, String> table2) throws Exception {
		// 左外连接表示等号左边的信息会被包含
		if(table2 == null){
			return new Tuple3<Integer, String, String>(table1.f0,table1.f1,null);
		}else{
			return new Tuple3<Integer, String, String>(table1.f0,table1.f1,table2.f1);
		}
	}
}).print();

System.out.println(“***********************************”);
//右外连接