spark GenerateUnsafeProjection 源码

spark GenerateUnsafeProjection 代码

文件路径：/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/codegen/GenerateUnsafeProjection.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.spark.sql.catalyst.expressions.codegen

import org.apache.spark.sql.catalyst.expressions._
import org.apache.spark.sql.catalyst.expressions.BindReferences.bindReferences
import org.apache.spark.sql.catalyst.expressions.codegen.Block._
import org.apache.spark.sql.types._

/**
 * Generates a [[Projection]] that returns an [[UnsafeRow]].
 *
 * It generates the code for all the expressions, computes the total length for all the columns
 * (can be accessed via variables), and then copies the data into a scratch buffer space in the
 * form of UnsafeRow (the scratch buffer will grow as needed).
 *
 * @note The returned UnsafeRow will be pointed to a scratch buffer inside the projection.
 */
object GenerateUnsafeProjection extends CodeGenerator[Seq[Expression], UnsafeProjection] {

  case class Schema(dataType: DataType, nullable: Boolean)

  /** Returns true iff we support this data type. */
  def canSupport(dataType: DataType): Boolean = UserDefinedType.sqlType(dataType) match {
    case NullType => true
    case _: AtomicType => true
    case _: CalendarIntervalType => true
    case t: StructType => t.forall(field => canSupport(field.dataType))
    case t: ArrayType if canSupport(t.elementType) => true
    case MapType(kt, vt, _) if canSupport(kt) && canSupport(vt) => true
    case _ => false
  }

  private def writeStructToBuffer(
      ctx: CodegenContext,
      input: String,
      index: String,
      schemas: Seq[Schema],
      rowWriter: String): String = {
    // Puts `input` in a local variable to avoid to re-evaluate it if it's a statement.
    val tmpInput = ctx.freshName("tmpInput")
    val fieldEvals = schemas.zipWithIndex.map { case (Schema(dt, nullable), i) =>
      val isNull = if (nullable) {
        JavaCode.isNullExpression(s"$tmpInput.isNullAt($i)")
      } else {
        FalseLiteral
      }
      ExprCode(isNull, JavaCode.expression(CodeGenerator.getValue(tmpInput, dt, i.toString), dt))
    }

    val rowWriterClass = classOf[UnsafeRowWriter].getName
    val structRowWriter = ctx.addMutableState(rowWriterClass, "rowWriter",
      v => s"$v = new $rowWriterClass($rowWriter, ${fieldEvals.length});")
    val previousCursor = ctx.freshName("previousCursor")
    s"""
       |final InternalRow $tmpInput = $input;
       |if ($tmpInput instanceof UnsafeRow) {
       |  $rowWriter.write($index, (UnsafeRow) $tmpInput);
       |} else {
       |  // Remember the current cursor so that we can calculate how many bytes are
       |  // written later.
       |  final int $previousCursor = $rowWriter.cursor();
       |  ${writeExpressionsToBuffer(ctx, tmpInput, fieldEvals, schemas, structRowWriter)}
       |  $rowWriter.setOffsetAndSizeFromPreviousCursor($index, $previousCursor);
       |}
     """.stripMargin
  }

  private def writeExpressionsToBuffer(
      ctx: CodegenContext,
      row: String,
      inputs: Seq[ExprCode],
      schemas: Seq[Schema],
      rowWriter: String,
      isTopLevel: Boolean = false): String = {
    val resetWriter = if (isTopLevel) {
      // For top level row writer, it always writes to the beginning of the global buffer holder,
      // which means its fixed-size region always in the same position, so we don't need to call
      // `reset` to set up its fixed-size region every time.
      if (inputs.map(_.isNull).forall(_ == FalseLiteral)) {
        // If all fields are not nullable, which means the null bits never changes, then we don't
        // need to clear it out every time.
        ""
      } else {
        s"$rowWriter.zeroOutNullBytes();"
      }
    } else {
      s"$rowWriter.resetRowWriter();"
    }

    val writeFields = inputs.zip(schemas).zipWithIndex.map {
      case ((input, Schema(dataType, nullable)), index) =>
        val dt = UserDefinedType.sqlType(dataType)

        val setNull = dt match {
          case t: DecimalType if t.precision > Decimal.MAX_LONG_DIGITS =>
            // Can't call setNullAt() for DecimalType with precision larger than 18.
            s"$rowWriter.write($index, (Decimal) null, ${t.precision}, ${t.scale});"
          case CalendarIntervalType => s"$rowWriter.write($index, (CalendarInterval) null);"
          case _ => s"$rowWriter.setNullAt($index);"
        }

        val writeField = writeElement(ctx, input.value, index.toString, dt, rowWriter)
        if (!nullable) {
          s"""
             |${input.code}
             |${writeField.trim}
           """.stripMargin
        } else {
          s"""
             |${input.code}
             |if (${input.isNull}) {
             |  ${setNull.trim}
             |} else {
             |  ${writeField.trim}
             |}
           """.stripMargin
        }
    }

    val writeFieldsCode = if (isTopLevel && (row == null || ctx.currentVars != null)) {
      // TODO: support whole stage codegen
      writeFields.mkString("\n")
    } else {
      assert(row != null, "the input row name cannot be null when generating code to write it.")
      ctx.splitExpressions(
        expressions = writeFields,
        funcName = "writeFields",
        arguments = Seq("InternalRow" -> row))
    }
    s"""
       |$resetWriter
       |$writeFieldsCode
     """.stripMargin
  }

  private def writeArrayToBuffer(
      ctx: CodegenContext,
      input: String,
      elementType: DataType,
      containsNull: Boolean,
      rowWriter: String): String = {
    // Puts `input` in a local variable to avoid to re-evaluate it if it's a statement.
    val tmpInput = ctx.freshName("tmpInput")
    val numElements = ctx.freshName("numElements")
    val index = ctx.freshName("index")

    val et = UserDefinedType.sqlType(elementType)

    val jt = CodeGenerator.javaType(et)

    val elementOrOffsetSize = et match {
      case t: DecimalType if t.precision <= Decimal.MAX_LONG_DIGITS => 8
      case _ if CodeGenerator.isPrimitiveType(jt) => et.defaultSize
      case _ => 8  // we need 8 bytes to store offset and length
    }

    val arrayWriterClass = classOf[UnsafeArrayWriter].getName
    val arrayWriter = ctx.addMutableState(arrayWriterClass, "arrayWriter",
      v => s"$v = new $arrayWriterClass($rowWriter, $elementOrOffsetSize);")

    val element = CodeGenerator.getValue(tmpInput, et, index)

    val elementAssignment = if (containsNull) {
      s"""
         |if ($tmpInput.isNullAt($index)) {
         |  $arrayWriter.setNull${elementOrOffsetSize}Bytes($index);
         |} else {
         |  ${writeElement(ctx, element, index, et, arrayWriter)}
         |}
       """.stripMargin
    } else {
      writeElement(ctx, element, index, et, arrayWriter)
    }

    s"""
       |final ArrayData $tmpInput = $input;
       |if ($tmpInput instanceof UnsafeArrayData) {
       |  $rowWriter.write((UnsafeArrayData) $tmpInput);
       |} else {
       |  final int $numElements = $tmpInput.numElements();
       |  $arrayWriter.initialize($numElements);
       |
       |  for (int $index = 0; $index < $numElements; $index++) {
       |    $elementAssignment
       |  }
       |}
     """.stripMargin
  }

  private def writeMapToBuffer(
      ctx: CodegenContext,
      input: String,
      index: String,
      keyType: DataType,
      valueType: DataType,
      valueContainsNull: Boolean,
      rowWriter: String): String = {
    // Puts `input` in a local variable to avoid to re-evaluate it if it's a statement.
    val tmpInput = ctx.freshName("tmpInput")
    val tmpCursor = ctx.freshName("tmpCursor")
    val previousCursor = ctx.freshName("previousCursor")

    // Writes out unsafe map according to the format described in `UnsafeMapData`.
    val keyArray = writeArrayToBuffer(
      ctx, s"$tmpInput.keyArray()", keyType, false, rowWriter)
    val valueArray = writeArrayToBuffer(
      ctx, s"$tmpInput.valueArray()", valueType, valueContainsNull, rowWriter)

    s"""
       |final MapData $tmpInput = $input;
       |if ($tmpInput instanceof UnsafeMapData) {
       |  $rowWriter.write($index, (UnsafeMapData) $tmpInput);
       |} else {
       |  // Remember the current cursor so that we can calculate how many bytes are
       |  // written later.
       |  final int $previousCursor = $rowWriter.cursor();
       |
       |  // preserve 8 bytes to write the key array numBytes later.
       |  $rowWriter.grow(8);
       |  $rowWriter.increaseCursor(8);
       |
       |  // Remember the current cursor so that we can write numBytes of key array later.
       |  final int $tmpCursor = $rowWriter.cursor();
       |
       |  $keyArray
       |
       |  // Write the numBytes of key array into the first 8 bytes.
       |  Platform.putLong(
       |    $rowWriter.getBuffer(),
       |    $tmpCursor - 8,
       |    $rowWriter.cursor() - $tmpCursor);
       |
       |  $valueArray
       |  $rowWriter.setOffsetAndSizeFromPreviousCursor($index, $previousCursor);
       |}
     """.stripMargin
  }

  private def writeElement(
      ctx: CodegenContext,
      input: String,
      index: String,
      dt: DataType,
      writer: String): String = dt match {
    case t: StructType =>
      writeStructToBuffer(
        ctx, input, index, t.map(e => Schema(e.dataType, e.nullable)), writer)

    case ArrayType(et, en) =>
      val previousCursor = ctx.freshName("previousCursor")
      s"""
         |// Remember the current cursor so that we can calculate how many bytes are
         |// written later.
         |final int $previousCursor = $writer.cursor();
         |${writeArrayToBuffer(ctx, input, et, en, writer)}
         |$writer.setOffsetAndSizeFromPreviousCursor($index, $previousCursor);
       """.stripMargin

    case MapType(kt, vt, vn) =>
      writeMapToBuffer(ctx, input, index, kt, vt, vn, writer)

    case DecimalType.Fixed(precision, scale) =>
      s"$writer.write($index, $input, $precision, $scale);"

    case NullType => ""

    case _ => s"$writer.write($index, $input);"
  }

  def createCode(
      ctx: CodegenContext,
      expressions: Seq[Expression],
      useSubexprElimination: Boolean = false): ExprCode = {
    val exprEvals = ctx.generateExpressions(expressions, useSubexprElimination)
    val exprSchemas = expressions.map(e => Schema(e.dataType, e.nullable))

    val numVarLenFields = exprSchemas.count {
      case Schema(dt, _) => !UnsafeRow.isFixedLength(dt)
      // TODO: consider large decimal and interval type
    }

    val rowWriterClass = classOf[UnsafeRowWriter].getName
    val rowWriter = ctx.addMutableState(rowWriterClass, "rowWriter",
      v => s"$v = new $rowWriterClass(${expressions.length}, ${numVarLenFields * 32});")

    // Evaluate all the subexpression.
    val evalSubexpr = ctx.subexprFunctionsCode

    val writeExpressions = writeExpressionsToBuffer(
      ctx, ctx.INPUT_ROW, exprEvals, exprSchemas, rowWriter, isTopLevel = true)

    val code =
      code"""
         |$rowWriter.reset();
         |$evalSubexpr
         |$writeExpressions
       """.stripMargin
    // `rowWriter` is declared as a class field, so we can access it directly in methods.
    ExprCode(code, FalseLiteral, JavaCode.expression(s"$rowWriter.getRow()", classOf[UnsafeRow]))
  }

  protected def canonicalize(in: Seq[Expression]): Seq[Expression] =
    in.map(ExpressionCanonicalizer.execute)

  protected def bind(in: Seq[Expression], inputSchema: Seq[Attribute]): Seq[Expression] =
    bindReferences(in, inputSchema)

  def generate(
      expressions: Seq[Expression],
      subexpressionEliminationEnabled: Boolean): UnsafeProjection = {
    create(canonicalize(expressions), subexpressionEliminationEnabled)
  }

  protected def create(references: Seq[Expression]): UnsafeProjection = {
    create(references, subexpressionEliminationEnabled = false)
  }

  private def create(
      expressions: Seq[Expression],
      subexpressionEliminationEnabled: Boolean): UnsafeProjection = {
    val ctx = newCodeGenContext()
    val eval = createCode(ctx, expressions, subexpressionEliminationEnabled)

    val codeBody =
      s"""
         |public java.lang.Object generate(Object[] references) {
         |  return new SpecificUnsafeProjection(references);
         |}
         |
         |class SpecificUnsafeProjection extends ${classOf[UnsafeProjection].getName} {
         |
         |  private Object[] references;
         |  ${ctx.declareMutableStates()}
         |
         |  public SpecificUnsafeProjection(Object[] references) {
         |    this.references = references;
         |    ${ctx.initMutableStates()}
         |  }
         |
         |  public void initialize(int partitionIndex) {
         |    ${ctx.initPartition()}
         |  }
         |
         |  // Scala.Function1 need this
         |  public java.lang.Object apply(java.lang.Object row) {
         |    return apply((InternalRow) row);
         |  }
         |
         |  public UnsafeRow apply(InternalRow ${ctx.INPUT_ROW}) {
         |    ${eval.code}
         |    return ${eval.value};
         |  }
         |
         |  ${ctx.declareAddedFunctions()}
         |}
       """.stripMargin

    val code = CodeFormatter.stripOverlappingComments(
      new CodeAndComment(codeBody, ctx.getPlaceHolderToComments()))
    logDebug(s"code for ${expressions.mkString(",")}:\n${CodeFormatter.format(code)}")

    val (clazz, _) = CodeGenerator.compile(code)
    clazz.generate(ctx.references.toArray).asInstanceOf[UnsafeProjection]
  }
}

相关信息

spark 源码目录

相关文章

spark CodeFormatter 源码

spark CodeGenerator 源码

spark CodegenFallback 源码

spark GenerateMutableProjection 源码

spark GenerateOrdering 源码

spark GeneratePredicate 源码

spark GenerateSafeProjection 源码

spark GenerateUnsafeRowJoiner 源码

spark javaCode 源码

spark package 源码