All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.apache.hudi.examples.spark.HoodieDataSourceExample.scala Maven / Gradle / Ivy

The newest version!
/*
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
 *
 *      http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package org.apache.hudi.examples.spark

import org.apache.hudi.DataSourceReadOptions.{END_COMMIT, QUERY_TYPE, QUERY_TYPE_INCREMENTAL_OPT_VAL, START_COMMIT}
import org.apache.hudi.DataSourceWriteOptions.{DELETE_OPERATION_OPT_VAL, DELETE_PARTITION_OPERATION_OPT_VAL, OPERATION, PARTITIONPATH_FIELD, PARTITIONS_TO_DELETE, PRECOMBINE_FIELD, RECORDKEY_FIELD}
import org.apache.hudi.QuickstartUtils.getQuickstartWriteConfigs
import org.apache.hudi.common.model.HoodieAvroPayload
import org.apache.hudi.config.HoodieWriteConfig.TBL_NAME
import org.apache.hudi.examples.common.{HoodieExampleDataGenerator, HoodieExampleSparkUtils}

import org.apache.spark.sql.SaveMode.{Append, Overwrite}
import org.apache.spark.sql.SparkSession

import scala.collection.JavaConverters._

/**
 * Simple examples of [[org.apache.hudi.DefaultSource]]
 *
 * To run this example, you should
 *   1. For running in IDE, set VM options `-Dspark.master=local[2]`
 *      2. For running in shell, using `spark-submit`
 *
 * Usage: HoodieWriteClientExample  .
 *  and  describe root path of hudi and table name
 * for example, `HoodieDataSourceExample file:///tmp/hoodie/hudi_cow_table hudi_cow_table`
 */
object HoodieDataSourceExample {

  def main(args: Array[String]): Unit = {

    if (args.length < 2) {
      System.err.println("Usage: HoodieDataSourceExample  ")
      System.exit(1)
    }
    val tablePath = args(0)
    val tableName = args(1)

    val spark = HoodieExampleSparkUtils.defaultSparkSession("Hudi Spark basic example")

    val dataGen = new HoodieExampleDataGenerator[HoodieAvroPayload]
    insertData(spark, tablePath, tableName, dataGen)
    updateData(spark, tablePath, tableName, dataGen)
    queryData(spark, tablePath, tableName, dataGen)

    incrementalQuery(spark, tablePath, tableName)
    pointInTimeQuery(spark, tablePath, tableName)

    delete(spark, tablePath, tableName)
    deleteByPartition(spark, tablePath, tableName)

    spark.stop()
  }

  /**
    * Generate some new trips, load them into a DataFrame and write the DataFrame into the Hudi dataset as below.
    */
  def insertData(spark: SparkSession, tablePath: String, tableName: String, dataGen: HoodieExampleDataGenerator[HoodieAvroPayload]): Unit = {

    val commitTime: String = System.currentTimeMillis().toString
    val inserts = dataGen.convertToStringList(dataGen.generateInserts(commitTime, 20)).asScala.toSeq
    val df = spark.read.json(spark.sparkContext.parallelize(inserts, 1))
    df.write.format("hudi").
      options(getQuickstartWriteConfigs).
      option(PRECOMBINE_FIELD.key, "ts").
      option(RECORDKEY_FIELD.key, "uuid").
      option(PARTITIONPATH_FIELD.key, "partitionpath").
      option(TBL_NAME.key, tableName).
      mode(Overwrite).
      save(tablePath)
  }

  /**
    * Load the data files into a DataFrame.
    */
  def queryData(spark: SparkSession, tablePath: String, tableName: String, dataGen: HoodieExampleDataGenerator[HoodieAvroPayload]): Unit = {
    val roViewDF = spark.read.format("hudi").load(tablePath)

    roViewDF.createOrReplaceTempView("hudi_ro_table")

    spark.sql("select fare, begin_lon, begin_lat, ts from  hudi_ro_table where fare > 20.0").show()
    //  +-----------------+-------------------+-------------------+---+
    //  |             fare|          begin_lon|          begin_lat| ts|
    //  +-----------------+-------------------+-------------------+---+
    //  |98.88075495133515|0.39556048623031603|0.17851135255091155|0.0|
    //  ...

    spark.sql("select _hoodie_commit_time, _hoodie_record_key, _hoodie_partition_path, rider, driver, fare from  hudi_ro_table").show()
    //  +-------------------+--------------------+----------------------+-------------------+--------------------+------------------+
    //  |_hoodie_commit_time|  _hoodie_record_key|_hoodie_partition_path|              rider|              driver|              fare|
    //  +-------------------+--------------------+----------------------+-------------------+--------------------+------------------+
    //  |     20191231181501|31cafb9f-0196-4b1...|            2020/01/02|rider-1577787297889|driver-1577787297889| 98.88075495133515|
    //  ...
  }

  /**
    * This is similar to inserting new data. Generate updates to existing trips using the data generator,
    * load into a DataFrame and write DataFrame into the hudi dataset.
    */
  def updateData(spark: SparkSession, tablePath: String, tableName: String, dataGen: HoodieExampleDataGenerator[HoodieAvroPayload]): Unit = {

    val commitTime: String = System.currentTimeMillis().toString
    val updates = dataGen.convertToStringList(dataGen.generateUpdates(commitTime, 10)).asScala.toSeq
    val df = spark.read.json(spark.sparkContext.parallelize(updates, 1))
    df.write.format("hudi").
      options(getQuickstartWriteConfigs).
      option(PRECOMBINE_FIELD.key, "ts").
      option(RECORDKEY_FIELD.key, "uuid").
      option(PARTITIONPATH_FIELD.key, "partitionpath").
      option(TBL_NAME.key, tableName).
      mode(Append).
      save(tablePath)
  }

  /**
   * Delete data based in data information.
   */
  def delete(spark: SparkSession, tablePath: String, tableName: String): Unit = {

    val roViewDF = spark.read.format("hudi").load(tablePath)
    roViewDF.createOrReplaceTempView("hudi_ro_table")
    val df = spark.sql("select uuid, partitionpath, ts from  hudi_ro_table limit 2")

    df.write.format("hudi").
      options(getQuickstartWriteConfigs).
      option(PRECOMBINE_FIELD.key, "ts").
      option(RECORDKEY_FIELD.key, "uuid").
      option(PARTITIONPATH_FIELD.key, "partitionpath").
      option(TBL_NAME.key, tableName).
      option(OPERATION.key, DELETE_OPERATION_OPT_VAL).
      mode(Append).
      save(tablePath)
  }

  /**
   *  Delete the data of a single or multiple partitions.
   */
  def deleteByPartition(spark: SparkSession, tablePath: String, tableName: String): Unit = {
    val df = spark.emptyDataFrame
    df.write.format("hudi").
      options(getQuickstartWriteConfigs).
      option(PRECOMBINE_FIELD.key, "ts").
      option(RECORDKEY_FIELD.key, "uuid").
      option(PARTITIONPATH_FIELD.key, "partitionpath").
      option(TBL_NAME.key, tableName).
      option(OPERATION.key, DELETE_PARTITION_OPERATION_OPT_VAL).
      option(PARTITIONS_TO_DELETE.key(), HoodieExampleDataGenerator.DEFAULT_PARTITION_PATHS.mkString(",")).
      mode(Append).
      save(tablePath)
  }

  /**
    * Hudi also provides capability to obtain a stream of records that changed since given commit timestamp.
   * This can be achieved using Hudi’s incremental view and providing a start time from which changes need to be streamed.
    * We do not need to specify endTime, if we want all changes after the given commit (as is the common case).
    */
  def incrementalQuery(spark: SparkSession, tablePath: String, tableName: String): Unit = {
    import spark.implicits._
    val commits = spark.sql("select distinct(_hoodie_commit_time) as commitTime from hudi_ro_table order by commitTime").map(k => k.getString(0)).take(50)
    val startTime = commits(commits.length - 1) // commit time we are interested in

    // incrementally query data
    val incViewDF = spark.
      read.
      format("hudi").
      option(QUERY_TYPE.key, QUERY_TYPE_INCREMENTAL_OPT_VAL).
      option(START_COMMIT.key, startTime).
        load(tablePath)
    incViewDF.createOrReplaceTempView("hudi_incr_table")
    spark.sql("select `_hoodie_commit_time`, fare, begin_lon, begin_lat, ts from hudi_incr_table where fare > 20.0").show()
  }

  /**
    * Lets look at how to query data as of a specific time.
    * The specific time can be represented by pointing endTime to a specific commit time
    * and beginTime to “000” (denoting earliest possible commit time).
    */
  def pointInTimeQuery(spark: SparkSession, tablePath: String, tableName: String): Unit = {
    import spark.implicits._
    val commits = spark.sql("select distinct(_hoodie_commit_time) as commitTime from  hudi_ro_table order by commitTime").map(k => k.getString(0)).take(50)
    val startTime = "000" // Represents all commits > this time.
    val endTime = commits(commits.length - 2) // commit time we are interested in

    //incrementally query data
    val incViewDF = spark.read.format("hudi").
      option(QUERY_TYPE.key, QUERY_TYPE_INCREMENTAL_OPT_VAL).
      option(START_COMMIT.key, startTime).
      option(END_COMMIT.key, endTime).
      load(tablePath)
    incViewDF.createOrReplaceTempView("hudi_incr_table")
    spark.sql("select `_hoodie_commit_time`, fare, begin_lon, begin_lat, ts from  hudi_incr_table where fare > 20.0").show()
  }
}




© 2015 - 2025 Weber Informatics LLC | Privacy Policy