Github user kiszk commented on a diff in the pull request:
https://github.com/apache/spark/pull/13909#discussion_r92341142
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/complexTypeCreator.scala
---
@@ -56,33 +58,93 @@ case class CreateArray(children: Seq[Expression])
extends Expression {
}
override def doGenCode(ctx: CodegenContext, ev: ExprCode): ExprCode = {
- val arrayClass = classOf[GenericArrayData].getName
- val values = ctx.freshName("values")
- ctx.addMutableState("Object[]", values, s"this.$values = null;")
+ val array = ctx.freshName("array")
- ev.copy(code = s"""
- this.$values = new Object[${children.size}];""" +
+ val et = dataType.elementType
+ val evals = children.map(e => e.genCode(ctx))
+ val isPrimitiveArray = ctx.isPrimitiveType(et)
+ val primitiveTypeName = if (isPrimitiveArray)
ctx.primitiveTypeName(et) else ""
+ val (preprocess, arrayData, arrayWriter) =
+ GenArrayData.getCodeArrayData(ctx, et, children.size,
isPrimitiveArray, array)
+
+ ev.copy(code =
+ preprocess +
ctx.splitExpressions(
ctx.INPUT_ROW,
- children.zipWithIndex.map { case (e, i) =>
- val eval = e.genCode(ctx)
- eval.code + s"""
- if (${eval.isNull}) {
- $values[$i] = null;
+ evals.zipWithIndex.map { case (eval, i) =>
+ eval.code +
+ (if (isPrimitiveArray) {
+ (if (!children(i).nullable) {
+ s"\n$arrayWriter.write($i, ${eval.value});"
+ } else {
+ s"""
+ if (${eval.isNull}) {
--- End diff --
@viiya and @cloud-fan , i see. This is not performance optimization. My
motivation is to reduce code size for avoiding 64KB problem as possible.
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
