[SPARK-9949] TakeOrderedAndProject returns wrong output attributes when project is pushed in to it

• Type: Sub-task
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Fix Version/s: 1.4.2, 1.5.0
• Component/s: SQL
• Target Version/s: 1.5.0
• Sprint: Spark 1.5 doc/QA sprint
• Assignee: Yin Huai
• Reporter: Yin Huai
• Created: 13/Aug/15 21:16
• Updated: 18/Aug/15 04:23
• Resolved: 15/Aug/15 00:35

Description

In TakeOrderedAndProject, we have

case class TakeOrderedAndProject(
    limit: Int,
    sortOrder: Seq[SortOrder],
    projectList: Option[Seq[NamedExpression]],
    child: SparkPlan) extends UnaryNode {

  override def output: Seq[Attribute] = child.output

When projectList is set, we should use its attributes as the output.

Issue Links

• links to: [Github] Pull Request #8179 (yhuai)
• links to: [Github] Pull Request #8252 (yhuai)