Github user hvanhovell commented on a diff in the pull request:
https://github.com/apache/spark/pull/20350#discussion_r182820983
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/decimalExpressions.scala
---
@@ -80,30 +80,44 @@ case class PromotePrecision(child: Expression) extends
UnaryExpression {
/**
* Rounds the decimal to given scale and check whether the decimal can fit
in provided precision
- * or not, returns null if not.
+ * or not. If not, if `nullOnOverflow` is `true`, it returns `null`;
otherwise an
+ * `ArithmeticException` is thrown.
*/
-case class CheckOverflow(child: Expression, dataType: DecimalType) extends
UnaryExpression {
+case class CheckOverflow(
+ child: Expression,
+ dataType: DecimalType,
+ nullOnOverflow: Boolean) extends UnaryExpression {
override def nullable: Boolean = true
override def nullSafeEval(input: Any): Any =
- input.asInstanceOf[Decimal].toPrecision(dataType.precision,
dataType.scale)
+ input.asInstanceOf[Decimal].toPrecision(
+ dataType.precision,
+ dataType.scale,
+ Decimal.ROUND_HALF_UP,
+ nullOnOverflow)
override protected def doGenCode(ctx: CodegenContext, ev: ExprCode):
ExprCode = {
nullSafeCodeGen(ctx, ev, eval => {
val tmp = ctx.freshName("tmp")
+ val onOverflow = if (nullOnOverflow) {
--- End diff --
Why are you not just calling `Decimal.toPrecision` here? There seems to be
very little value in code generating this (no specialization).
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
