chaokunyang commented on code in PR #2597: URL: https://github.com/apache/uniffle/pull/2597#discussion_r2307144902
########## client-spark/common/src/main/scala/org/apache/spark/serializer/ForySerializer.scala: ########## @@ -0,0 +1,181 @@ +/* + * 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.serializer + +import org.apache.fory.Fory +import org.apache.fory.config.{CompatibleMode, Language} +import org.apache.spark.internal.Logging + +import java.io.{InputStream, OutputStream, Serializable} +import java.nio.ByteBuffer +import scala.reflect.ClassTag + +class ForySerializer extends org.apache.spark.serializer.Serializer + with Logging + with Serializable { + + override def newInstance(): SerializerInstance = new ForySerializerInstance() + + override def supportsRelocationOfSerializedObjects: Boolean = true + +} + +class ForySerializerInstance extends org.apache.spark.serializer.SerializerInstance { + + // Thread-local Fury instance for thread safety + private val fury: ThreadLocal[Fory] = ThreadLocal.withInitial(() => { + val f = Fory.builder() + .withLanguage(Language.JAVA) + .withRefTracking(true) + .withCompatibleMode(CompatibleMode.COMPATIBLE) + .requireClassRegistration(false) + .build() + f + }) + + override def serialize[T: ClassTag](t: T): ByteBuffer = { + val bytes = fury.get().serialize(t.asInstanceOf[AnyRef]) + ByteBuffer.wrap(bytes) + } + + override def deserialize[T: ClassTag](bytes: ByteBuffer): T = { + val array = if (bytes.hasArray) { + val offset = bytes.arrayOffset() + bytes.position() + val length = bytes.remaining() + java.util.Arrays.copyOfRange(bytes.array(), offset, offset + length) + } else { + val array = new Array[Byte](bytes.remaining()) + bytes.get(array) + array + } + fury.get().deserialize(array).asInstanceOf[T] + } + + override def deserialize[T: ClassTag](bytes: ByteBuffer, loader: ClassLoader): T = { + // Fury handles class loading internally, so we can use the standard deserialize method + deserialize[T](bytes) + } + + override def serializeStream(s: OutputStream): SerializationStream = { + new ForySerializationStream(fury.get(), s) + } + + override def deserializeStream(s: InputStream): DeserializationStream = { + new ForyDeserializationStream(fury.get(), s) + } +} + +class ForySerializationStream(fury: Fory, outputStream: OutputStream) + extends org.apache.spark.serializer.SerializationStream { + + private val out = outputStream + private var closed = false + + override def writeObject[T: ClassTag](t: T): SerializationStream = { + if (closed) { + throw new IllegalStateException("Stream is closed") + } + + val bytes = fury.serialize(t.asInstanceOf[AnyRef]) + // Write length first, then data + writeInt(bytes.length) + out.write(bytes) + this + } + + private def writeInt(value: Int): Unit = { + out.write((value >>> 24) & 0xFF) Review Comment: Just use: ``` public void writeInt64(MemoryBuffer buffer, long value) { LongSerializer.writeInt64(buffer, value, longEncoding); } public long readInt64(MemoryBuffer buffer) { return LongSerializer.readInt64(buffer, longEncoding); } ``` This will be faster and simple, it also compress data -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
