Apache spark 如何在spark结构化流媒体中将流数据集转换为JavaRDD

我有一个结构化的流媒体应用程序，可以从hdfs路径读取数据

structStream = spark.readStream().format("text").load(parameters.get("input"));

JavaRDD<String> transformedstructStreamRDD = structStream.as(Encoders.STRING()).toJavaRDD();

Dataset<Row> df = spark.createDataFrame(transformedstructStreamRDD, String.class);
//Dataset<Row> df = structStream.as("dummy");

StreamingQuery streamingQuery = df.writeStream().format("csv").option("checkpointLocation","/user/hadoop/chkpointpath/").option("path", "/user/hadoop/output/").start();
try {
        streamingQuery.awaitTermination();
    } catch (StreamingQueryException e) {
        e.printStackTrace();
    }

org.apache.spark.sql.AnalysisException: Queries with streaming sources must be executed with writeStream.start();;