【java-spark】sparksql入门小例子

TestData.json

{"regNo": "1237855","sortTime": 1511402600000,"pname": "北京服装有限公司","brandRank": "服装-羽绒服"}
{"regNo": "2236811","sortTime": 1461404600000,"pname": "上海市器械有限公司","brandRank": "运动球类球"}
{"regNo": "3337815","sortTime": 1411401600000,"pname": "泉州奇鹭物联网科技有限公司","brandRank": "服装-婴儿全套衣"}
{"regNo": "3044199","sortTime": 1374163200000,"pname": "泉州锻炼器械科技有限公司","brandRank": "锻炼身体器械；健身球；运动球类球胆"}

TestData02.json

{"pname": "北京服装有限公司","body": "服装-羽绒服-body-hahaha"}
{"pname": "泉州奇鹭物联网科技有限公司","body": "服装-婴儿全套衣-body-hahaha"}
{"pname": "泉州锻炼器械科技有限公司","body": "锻炼身体器械；健身球；运动球类球胆-body-hahaha"}

import org.apache.spark.sql.AnalysisException;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Encoder;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.Row;
import org.apache.spark.sql.SparkSession;

public class SparkSQLTest {
	public static void main(String[] args) throws AnalysisException {
		SparkSession spark = SparkSession.builder().master("local").appName("Spark SQL Test")
				.getOrCreate();

		createView(spark);//创建视图

		createClassView(spark);//创建实体类视图
		
		createJoinTest(spark);//关联测试

	}

	private static void createJoinTest(SparkSession spark) {
		Encoder<regCom> regComEncoder = Encoders.bean(regCom.class);
		Dataset<regCom> regComDS = spark.read().json("C:\\Users\\A\\Desktop\\TestData.json").as(regComEncoder);
//		regComDS.show();

		regComDS.createOrReplaceTempView("regCom");

		Dataset<Row> regBodyDS = spark.read().json("C:\\Users\\A\\Desktop\\TestData02.json");
		regBodyDS.createOrReplaceTempView("regBody");// 临时会话
		
		Dataset<Row> sqlDS = spark.sql("select * from regCom rc left join regBody rb  on rc.pname=rb.pname");
		sqlDS.show();
		
	}

	private static void createClassView(SparkSession spark) {

		Encoder<regCom> regComEncoder = Encoders.bean(regCom.class);
		Dataset<regCom> regComDS = spark.read().json("C:\\Users\\A\\Desktop\\TestData.json").as(regComEncoder);
//		regComDS.show();

		regComDS.createOrReplaceTempView("regCom");

		Dataset<Row> sqlDS = spark.sql("select sum(sortTime) from regCom where pname like '%泉州%' ");
		sqlDS.show();

	}

	private static void createView(SparkSession spark) {
		Dataset<Row> df = spark.read().json("C:\\Users\\A\\Desktop\\TestData.json");
//		df.show();
		df.createOrReplaceTempView("testData");// 临时会话
//		df.createGlobalTempView("people");//全局视图
		Dataset<Row> sqlDF = spark.sql("SELECT * FROM testData");
		sqlDF.show();
	}

	public static class regCom {
		private String regNo;
		private Long sortTime;
		private String pname;
		private String brandRank;

		public String getRegNo() {
			return regNo;
		}

		public void setRegNo(String regNo) {
			this.regNo = regNo;
		}

		public Long getSortTime() {
			return sortTime;
		}

		public void setSortTime(Long sortTime) {
			this.sortTime = sortTime;
		}

		public String getPname() {
			return pname;
		}

		public void setPname(String pname) {
			this.pname = pname;
		}

		public String getBrandRank() {
			return brandRank;
		}

		public void setBrandRank(String brandRank) {
			this.brandRank = brandRank;
		}

	}

}