spark-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From rxin <...@git.apache.org>
Subject [GitHub] spark pull request: [SPARK-9023] [SQL] Efficiency improvements for...
Date Mon, 20 Jul 2015 06:31:48 GMT
Github user rxin commented on a diff in the pull request:

    https://github.com/apache/spark/pull/7456#discussion_r34970391
  
    --- Diff: sql/core/src/main/scala/org/apache/spark/sql/execution/UnsafeRowSerializer.scala
---
    @@ -0,0 +1,142 @@
    +/*
    + * 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.spark.sql.execution
    +
    +import java.io.{DataInputStream, DataOutputStream, OutputStream, InputStream}
    +import java.nio.ByteBuffer
    +
    +import scala.reflect.ClassTag
    +
    +import com.google.common.io.ByteStreams
    +
    +import org.apache.spark.serializer.{SerializationStream, DeserializationStream, SerializerInstance,
Serializer}
    +import org.apache.spark.sql.catalyst.expressions.UnsafeRow
    +import org.apache.spark.unsafe.PlatformDependent
    +
    +/**
    + * Serializer for serializing [[UnsafeRow]]s during shuffle. Since UnsafeRows are already
stored as
    + * bytes, this serializer simply copies those bytes to the underlying output stream.
When
    + * deserializing a stream of rows, instances of this serializer mutate and return a single
UnsafeRow
    + * instance that is backed by an on-heap byte array.
    + *
    + * Note that this serializer implements only the [[Serializer]] methods that are used
during
    + * shuffle, so certain [[SerializerInstance]] methods will throw UnsupportedOperationException.
    + *
    + * This serializer does not support UnsafeRows that use
    + * [[org.apache.spark.sql.catalyst.util.ObjectPool]].
    + *
    + * @param numFields the number of fields in the row being serialized.
    + */
    +private[sql] class UnsafeRowSerializer(numFields: Int) extends Serializer with Serializable
{
    +  override def newInstance(): SerializerInstance = new UnsafeRowSerializerInstance(numFields)
    +  override private[spark] def supportsRelocationOfSerializedObjects: Boolean = true
    +}
    +
    +private class UnsafeRowSerializerInstance(numFields: Int) extends SerializerInstance
{
    +
    +  private[this] val EOF: Int = -1
    +
    +  override def serializeStream(out: OutputStream): SerializationStream = new SerializationStream
{
    +    private[this] var writeBuffer: Array[Byte] = new Array[Byte](4096)
    +    private[this] val dOut: DataOutputStream = new DataOutputStream(out)
    +
    +    override def writeValue[T: ClassTag](value: T): SerializationStream = {
    +      val row = value.asInstanceOf[UnsafeRow]
    +      assert(row.getPool == null, "UnsafeRowSerializer does not support ObjectPool")
    +      dOut.writeInt(row.getSizeInBytes)
    +      var dataRemaining: Int = row.getSizeInBytes
    +      val baseObject = row.getBaseObject
    +      var rowReadPosition: Long = row.getBaseOffset
    +      while (dataRemaining > 0) {
    +        val toTransfer: Int = Math.min(writeBuffer.length, dataRemaining)
    +        PlatformDependent.copyMemory(
    +          baseObject,
    +          rowReadPosition,
    +          writeBuffer,
    +          PlatformDependent.BYTE_ARRAY_OFFSET,
    +          toTransfer)
    +        out.write(writeBuffer, 0, toTransfer)
    +        rowReadPosition += toTransfer
    +        dataRemaining -= toTransfer
    +      }
    +      this
    +    }
    +    override def writeKey[T: ClassTag](key: T): SerializationStream = {
    +      assert(key.isInstanceOf[Int])
    --- End diff --
    
    you need to add some comment explaining why we are not doing anything when writing keys.



---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


Mime
View raw message