Github user kiszk commented on a diff in the pull request:
https://github.com/apache/spark/pull/21208#discussion_r187692007
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/collectionOperations.scala
---
@@ -1229,3 +1229,140 @@ case class Flatten(child: Expression) extends
UnaryExpression {
override def prettyName: String = "flatten"
}
+
+/**
+ * Returns the array containing the given input value (left) count (right)
times.
+ */
+@ExpressionDescription(
+ usage = "_FUNC_(element, count) - Returns the array containing element
count times.",
+ examples = """
+ Examples:
+ > SELECT _FUNC_('123', 2);
+ ['123', '123']
+ """)
+case class ArrayRepeat(left: Expression, right: Expression)
+ extends BinaryExpression with ExpectsInputTypes {
+
+ override def dataType: ArrayType = ArrayType(left.dataType,
left.nullable)
+
+ override def inputTypes: Seq[AbstractDataType] = Seq(AnyDataType,
IntegerType)
+
+ override def nullable: Boolean = right.nullable
+
+ override def eval(input: InternalRow): Any = {
+ val count = right.eval(input)
+ if (count == null) {
+ null
+ } else {
+ new
GenericArrayData(List.fill(count.asInstanceOf[Int])(left.eval(input)))
+ }
+ }
+
+ override def prettyName: String = "array_repeat"
+
+ override def nullSafeCodeGen(ctx: CodegenContext,
+ ev: ExprCode,
+ f: (String, String) => String): ExprCode = {
+ val leftGen = left.genCode(ctx)
+ val rightGen = right.genCode(ctx)
+ val resultCode = f(leftGen.value, rightGen.value)
+
+ if (nullable) {
+ val nullSafeEval =
+ leftGen.code +
+ rightGen.code + ctx.nullSafeExec(right.nullable,
rightGen.isNull) {
+ s"""
+ ${ev.isNull} = false;
+ $resultCode
+ """
+ }
+
+ ev.copy(code =
+ s"""
+ | boolean ${ev.isNull} = true;
+ | ${CodeGenerator.javaType(dataType)} ${ev.value} =
+ | ${CodeGenerator.defaultValue(dataType)};
+ | $nullSafeEval
+ """.stripMargin
+ )
+ } else {
+ ev.copy(code =
+ s"""
+ | boolean ${ev.isNull} = false;
+ | ${leftGen.code}
+ | ${rightGen.code}
+ | ${CodeGenerator.javaType(dataType)} ${ev.value} =
+ | ${CodeGenerator.defaultValue(dataType)};
+ | $resultCode
+ """.stripMargin
+ , isNull = FalseLiteral)
+ }
+
+ }
+
+ override def doGenCode(ctx: CodegenContext, ev: ExprCode): ExprCode = {
+
+ nullSafeCodeGen(ctx, ev, (l, r) => {
+ val et = dataType.elementType
+ val isPrimitive = CodeGenerator.isPrimitiveType(et)
+
+ val arrayDataName = ctx.freshName("arrayData")
+ val arrayName = ctx.freshName("arrayObject")
+ val numElements = ctx.freshName("numElements")
+
+ val genNumElements =
+ s"""
+ | int $numElements = 0;
+ | if ($r > 0) {
+ | $numElements = $r;
+ | }
+ """.stripMargin
+
+ val initialization = if (isPrimitive) {
+ val arrayName = ctx.freshName("array")
+ val baseOffset = Platform.BYTE_ARRAY_OFFSET
+ s"""
+ | int numBytes = ${et.defaultSize} * $numElements;
+ | int unsafeArraySizeInBytes =
+ | UnsafeArrayData.calculateHeaderPortionInBytes($numElements)
+ | + org.apache.spark.unsafe.array.ByteArrayMethods
+ | .roundNumberOfBytesToNearestWord(numBytes);
+ | byte[] $arrayName = new byte[unsafeArraySizeInBytes];
+ | UnsafeArrayData $arrayDataName = new UnsafeArrayData();
+ | Platform.putLong($arrayName, $baseOffset, $numElements);
+ | $arrayDataName.pointTo($arrayName, $baseOffset,
unsafeArraySizeInBytes);
+ | ${ev.value} = $arrayDataName;
+ """.stripMargin
+ } else {
+ s"${ev.value} = new ${classOf[GenericArrayData].getName()}(new
Object[$numElements]);"
+ }
+
+ val primitiveValueTypeName = CodeGenerator.primitiveTypeName(et)
+ val assignments = {
+ val updateArray = if (isPrimitive) {
+ val isNull = left.genCode(ctx).isNull
+ s"""
+ | if ($isNull) {
+ | ${ev.value}.setNullAt(k);
+ | } else {
+ | ${ev.value}.set$primitiveValueTypeName(k, $l);
+ | }
+ """.stripMargin
+ } else {
+ s"${ev.value}.update(k, $l);"
--- End diff --
I realized that other places use `update()` without null check. It looks
good as-is.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
