java-Flink(二)

流处理分词统计

1.创建流处理的环境

 //创建流处理执行环境
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();

2.读取数据

 //从文件中读取数据
        String inputPath = "/Users/lvqing/Desktop/wm/javacode/flink/src/main/resources/hello.txt";
        DataStream<String> inputDataStream = env.readTextFile(inputPath);

3.基于流处理的转换计算

//基于数据流进行转换计算
        SingleOutputStreamOperator<Tuple2<String, Integer>> resultStrem = inputDataStream.flatMap(new WordCount.MyflatMapper())
                .keyBy(0)
                .sum(1);
        resultStrem.print();

4.执行任务

    env.execute();

5.执行结果

4> (flink,1)
3> (word,1)
2> (hello,1)
2> (hello,2)
2> (are,1)
3> (how,1)
2> (thank,1)
3> (,1)
3> (you,1)
3> (fine,1)
3> (you,2)
3> (you,3)
4> (and,1)
1> (spark,1)
2> (hello,3)
1> (scala,1)
2> (hello,4)

结果中的1>代表当前执行的线程的位数。