FilterExec Unary Physical Operator¶

FilterExec is a unary physical operator that represents Filter and TypedFilter unary logical operators at execution time.

Creating Instance¶

FilterExec takes the following to be created:

• Condition (Expression)
• Child physical operator

FilterExec is created when:

• BasicOperators execution planning strategy is executed (and plans Filter and TypedFilter unary logical operators
• others

Performance Metrics¶

Executing Operator¶

doExecute(): RDD[InternalRow]

doExecute creates a FilterEvaluatorFactory (with the condition, the output of the child operator, and the numOutputRows metric).

With spark.sql.execution.usePartitionEvaluator enabled, doExecute requests the child operator to execute (that gives a RDD[InternalRow]). In the end, doExecute uses RDD.mapPartitionsWithEvaluator operator (Spark Core) with the FilterEvaluatorFactory.

Otherwise, doExecute...FIXME