Skip to content

HiveTableScans Execution Planning Strategy

HiveTableScans is an execution planning strategy (of Hive-specific SparkPlanner) that <>.

=== [[apply]] Planning Logical Plan for Execution -- apply Method

[source, scala]

apply( plan: LogicalPlan): Seq[SparkPlan]


apply converts (destructures) the input ../spark-sql-LogicalPlan.md[logical query plan] into projection expressions, predicate expressions, and a HiveTableRelation.md[HiveTableRelation].

apply requests the HiveTableRelation for the HiveTableRelation.md#partitionCols[partition columns] and filters the predicates to find so-called pruning predicates (that are expressions with no references and among the partition columns).

In the end, apply creates a "partial" HiveTableScanExec.md[HiveTableScanExec] physical operator (with the HiveTableRelation and the pruning predicates only) and pruneFilterProject.

apply is part of GenericStrategy abstraction.


Last update: 2020-08-20