flink-libraries/flink-table/src/main/scala/org/apache/flink/table/functions/aggfunctions/AvgAggFunction.scala

/*
 * 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.flink.table.functions.aggfunctions

import java.math.{BigDecimal, BigInteger, MathContext}
import java.lang.{Iterable => JIterable}

import org.apache.flink.api.common.typeinfo.{BasicTypeInfo, TypeInformation}
import org.apache.flink.api.java.tuple.{Tuple2 => JTuple2}
import org.apache.flink.api.java.typeutils.TupleTypeInfo
import org.apache.flink.table.functions.AggregateFunction

/** The initial accumulator for Integral Avg aggregate function */
class IntegralAvgAccumulator extends JTuple2[Long, Long] {
  f0 = 0L //sum
  f1 = 0L //count
}

/**
  * Base class for built-in Integral Avg aggregate function
  *
  * @tparam T the type for the aggregation result
  */
abstract class IntegralAvgAggFunction[T] extends AggregateFunction[T, IntegralAvgAccumulator] {

  override def createAccumulator(): IntegralAvgAccumulator = {
    new IntegralAvgAccumulator
  }

  def accumulate(acc: IntegralAvgAccumulator, value: Any): Unit = {
    if (value != null) {
      val v = value.asInstanceOf[Number].longValue()
      acc.f0 += v
      acc.f1 += 1L
    }
  }

  def retract(acc: IntegralAvgAccumulator, value: Any): Unit = {
    if (value != null) {
      val v = value.asInstanceOf[Number].longValue()
      acc.f0 -= v
      acc.f1 -= 1L
    }
  }

  override def getValue(acc: IntegralAvgAccumulator): T = {
    if (acc.f1 == 0) {
      null.asInstanceOf[T]
    } else {
      resultTypeConvert(acc.f0 / acc.f1)
    }
  }

  def merge(acc: IntegralAvgAccumulator, its: JIterable[IntegralAvgAccumulator]): Unit = {
    val iter = its.iterator()
    while (iter.hasNext) {
      val a = iter.next()
      acc.f1 += a.f1
      acc.f0 += a.f0
    }
  }

  def resetAccumulator(acc: IntegralAvgAccumulator): Unit = {
    acc.f0 = 0L
    acc.f1 = 0L
  }

  override def getAccumulatorType: TypeInformation[IntegralAvgAccumulator] = {
    new TupleTypeInfo(
      classOf[IntegralAvgAccumulator],
      BasicTypeInfo.LONG_TYPE_INFO,
      BasicTypeInfo.LONG_TYPE_INFO)
  }

  /**
    * Convert the intermediate result to the expected aggregation result type
    *
    * @param value the intermediate result. We use a Long container to save
    *              the intermediate result to avoid the overflow by sum operation.
    * @return the result value with the expected aggregation result type
    */
  def resultTypeConvert(value: Long): T
}

/**
  * Built-in Byte Avg aggregate function
  */
class ByteAvgAggFunction extends IntegralAvgAggFunction[Byte] {
  override def resultTypeConvert(value: Long): Byte = value.toByte
}

/**
  * Built-in Short Avg aggregate function
  */
class ShortAvgAggFunction extends IntegralAvgAggFunction[Short] {
  override def resultTypeConvert(value: Long): Short = value.toShort
}

/**
  * Built-in Int Avg aggregate function
  */
class IntAvgAggFunction extends IntegralAvgAggFunction[Int] {
  override def resultTypeConvert(value: Long): Int = value.toInt
}

/** The initial accumulator for Big Integral Avg aggregate function */
class BigIntegralAvgAccumulator
  extends JTuple2[BigInteger, Long] {
  f0 = BigInteger.ZERO //sum
  f1 = 0L //count
}

/**
  * Base Class for Built-in Big Integral Avg aggregate function
  *
  * @tparam T the type for the aggregation result
  */
abstract class BigIntegralAvgAggFunction[T]
  extends AggregateFunction[T, BigIntegralAvgAccumulator] {

  override def createAccumulator(): BigIntegralAvgAccumulator = {
    new BigIntegralAvgAccumulator
  }

  def accumulate(acc: BigIntegralAvgAccumulator, value: Any): Unit = {
    if (value != null) {
      val v = value.asInstanceOf[Long]
      acc.f0 = acc.f0.add(BigInteger.valueOf(v))
      acc.f1 += 1L
    }
  }

  def retract(acc: BigIntegralAvgAccumulator, value: Any): Unit = {
    if (value != null) {
      val v = value.asInstanceOf[Long]
      acc.f0 = acc.f0.subtract(BigInteger.valueOf(v))
      acc.f1 -= 1L
    }
  }

  override def getValue(acc: BigIntegralAvgAccumulator): T = {
    if (acc.f1 == 0) {
      null.asInstanceOf[T]
    } else {
      resultTypeConvert(acc.f0.divide(BigInteger.valueOf(acc.f1)))
    }
  }

  def merge(acc: BigIntegralAvgAccumulator, its: JIterable[BigIntegralAvgAccumulator]): Unit = {
    val iter = its.iterator()
    while (iter.hasNext) {
      val a = iter.next()
      acc.f1 += a.f1
      acc.f0 = acc.f0.add(a.f0)
    }
  }

  def resetAccumulator(acc: BigIntegralAvgAccumulator): Unit = {
    acc.f0 = BigInteger.ZERO
    acc.f1 = 0
  }

  override def getAccumulatorType: TypeInformation[BigIntegralAvgAccumulator] = {
    new TupleTypeInfo(
      classOf[BigIntegralAvgAccumulator],
      BasicTypeInfo.BIG_INT_TYPE_INFO,
      BasicTypeInfo.LONG_TYPE_INFO)
  }

  /**
    * Convert the intermediate result to the expected aggregation result type
    *
    * @param value the intermediate result. We use a BigInteger container to
    *              save the intermediate result to avoid the overflow by sum
    *              operation.
    * @return the result value with the expected aggregation result type
    */
  def resultTypeConvert(value: BigInteger): T
}

/**
  * Built-in Long Avg aggregate function
  */
class LongAvgAggFunction extends BigIntegralAvgAggFunction[Long] {
  override def resultTypeConvert(value: BigInteger): Long = value.longValue()
}

/** The initial accumulator for Floating Avg aggregate function */
class FloatingAvgAccumulator extends JTuple2[Double, Long] {
  f0 = 0 //sum
  f1 = 0L //count
}

/**
  * Base class for built-in Floating Avg aggregate function
  *
  * @tparam T the type for the aggregation result
  */
abstract class FloatingAvgAggFunction[T] extends AggregateFunction[T, FloatingAvgAccumulator] {

  override def createAccumulator(): FloatingAvgAccumulator = {
    new FloatingAvgAccumulator
  }

  def accumulate(acc: FloatingAvgAccumulator, value: Any): Unit = {
    if (value != null) {
      val v = value.asInstanceOf[Number].doubleValue()
      acc.f0 += v
      acc.f1 += 1L
    }
  }

  def retract(acc: FloatingAvgAccumulator, value: Any): Unit = {
    if (value != null) {
      val v = value.asInstanceOf[Number].doubleValue()
      acc.f0 -= v
      acc.f1 -= 1L
    }
  }

  override def getValue(acc: FloatingAvgAccumulator): T = {
    if (acc.f1 == 0) {
      null.asInstanceOf[T]
    } else {
      resultTypeConvert(acc.f0 / acc.f1)
    }
  }

  def merge(acc: FloatingAvgAccumulator, its: JIterable[FloatingAvgAccumulator]): Unit = {
    val iter = its.iterator()
    while (iter.hasNext) {
      val a = iter.next()
      acc.f1 += a.f1
      acc.f0 += a.f0
    }
  }

  def resetAccumulator(acc: FloatingAvgAccumulator): Unit = {
    acc.f0 = 0
    acc.f1 = 0L
  }

  override def getAccumulatorType: TypeInformation[FloatingAvgAccumulator] = {
    new TupleTypeInfo(
      classOf[FloatingAvgAccumulator],
      BasicTypeInfo.DOUBLE_TYPE_INFO,
      BasicTypeInfo.LONG_TYPE_INFO)
  }

  /**
    * Convert the intermediate result to the expected aggregation result type
    *
    * @param value the intermediate result. We use a Double container to save
    *              the intermediate result to avoid the overflow by sum operation.
    * @return the result value with the expected aggregation result type
    */
  def resultTypeConvert(value: Double): T
}

/**
  * Built-in Float Avg aggregate function
  */
class FloatAvgAggFunction extends FloatingAvgAggFunction[Float] {
  override def resultTypeConvert(value: Double): Float = value.toFloat
}

/**
  * Built-in Int Double aggregate function
  */
class DoubleAvgAggFunction extends FloatingAvgAggFunction[Double] {
  override def resultTypeConvert(value: Double): Double = value
}

/** The initial accumulator for Big Decimal Avg aggregate function */
class DecimalAvgAccumulator extends JTuple2[BigDecimal, Long] {
  f0 = BigDecimal.ZERO //sum
  f1 = 0L //count
}

/**
  * Base class for built-in Big Decimal Avg aggregate function
  */
class DecimalAvgAggFunction(context: MathContext)
  extends AggregateFunction[BigDecimal, DecimalAvgAccumulator] {

  override def createAccumulator(): DecimalAvgAccumulator = {
    new DecimalAvgAccumulator
  }

  def accumulate(acc: DecimalAvgAccumulator, value: Any): Unit = {
    if (value != null) {
      val v = value.asInstanceOf[BigDecimal]
      acc.f0 = acc.f0.add(v)
      acc.f1 += 1L
    }
  }

  def retract(acc: DecimalAvgAccumulator, value: Any): Unit = {
    if (value != null) {
      val v = value.asInstanceOf[BigDecimal]
      acc.f0 = acc.f0.subtract(v)
      acc.f1 -= 1L
    }
  }

  override def getValue(acc: DecimalAvgAccumulator): BigDecimal = {
    if (acc.f1 == 0) {
      null.asInstanceOf[BigDecimal]
    } else {
      acc.f0.divide(BigDecimal.valueOf(acc.f1), context)
    }
  }

  def merge(acc: DecimalAvgAccumulator, its: JIterable[DecimalAvgAccumulator]): Unit = {
    val iter = its.iterator()
    while (iter.hasNext) {
      val a = iter.next()
      acc.f0 = acc.f0.add(a.f0)
      acc.f1 += a.f1
    }
  }

  def resetAccumulator(acc: DecimalAvgAccumulator): Unit = {
    acc.f0 = BigDecimal.ZERO
    acc.f1 = 0L
  }

  override def getAccumulatorType: TypeInformation[DecimalAvgAccumulator] = {
    new TupleTypeInfo(
      classOf[DecimalAvgAccumulator],
      BasicTypeInfo.BIG_DEC_TYPE_INFO,
      BasicTypeInfo.LONG_TYPE_INFO)
  }
}