Trait datafusion::physical_plan::ExecutionPlan

source · [−]
pub trait ExecutionPlan: Debug + Send + Sync {

Show 14 methods    fn as_any(&self) -> &dyn Any;
    fn schema(&self) -> SchemaRef;
    fn output_partitioning(&self) -> Partitioning;
    fn output_ordering(&self) -> Option<&[PhysicalSortExpr]>;
    fn children(&self) -> Vec<Arc<dyn ExecutionPlan>>ⓘNotable traits for Vec<u8, A>impl<A> Write for Vec<u8, A> where
    A: Allocator, ;
    fn with_new_children(
        self: Arc<Self>,
        children: Vec<Arc<dyn ExecutionPlan>>
    ) -> Result<Arc<dyn ExecutionPlan>>;
    fn execute(
        &self,
        partition: usize,
        context: Arc<TaskContext>
    ) -> Result<SendableRecordBatchStream>;
    fn statistics(&self) -> Statistics;

    fn required_child_distribution(&self) -> Distribution { ... }
    fn relies_on_input_order(&self) -> bool { ... }
    fn maintains_input_order(&self) -> bool { ... }
    fn benefits_from_input_partitioning(&self) -> bool { ... }
    fn metrics(&self) -> Option<MetricsSet> { ... }
    fn fmt_as(&self, _t: DisplayFormatType, f: &mut Formatter<'_>) -> Result { ... }

}
Expand description

ExecutionPlan represent nodes in the DataFusion Physical Plan.

Each ExecutionPlan is Partition-aware and is responsible for creating the actual async SendableRecordBatchStreams of RecordBatch that incrementally compute the operator’s output from its input partition.

ExecutionPlan can be displayed in an simplified form using the return value from displayable in addition to the (normally quite verbose) Debug output.

Required Methods

Returns the execution plan as Any so that it can be downcast to a specific implementation.

Get the schema for this execution plan

Specifies the output partitioning scheme of this plan

If the output of this operator is sorted, returns Some(keys) with the description of how it was sorted.

For example, Sort, (obviously) produces sorted output as does SortPreservingMergeStream. Less obviously Projection produces sorted output if its input was sorted as it does not reorder the input rows,

It is safe to return None here if your operator does not have any particular output order here

Get a list of child execution plans that provide the input for this plan. The returned list will be empty for leaf nodes, will contain a single value for unary nodes, or two values for binary nodes (such as joins).

Returns a new plan where all children were replaced by new plans.

creates an iterator

Returns the global output statistics for this ExecutionPlan node.

Provided Methods

Specifies the data distribution requirements of all the children for this operator

Returns true if this operator relies on its inputs being produced in a certain order (for example that they are sorted a particular way) for correctness.

If true is returned, DataFusion will not apply certain optimizations which might reorder the inputs (such as repartitioning to increase concurrency).

The default implementation returns true

WARNING: if you override this default and return false, your operator can not rely on datafusion preserving the input order as it will likely not.

Returns false if this operator’s implementation may reorder rows within or between partitions.

For example, Projection, Filter, and Limit maintain the order of inputs – they may transform values (Projection) or not produce the same number of rows that went in (Filter and Limit), but the rows that are produced go in the same way.

DataFusion uses this metadata to apply certain optimizations such as automatically repartitioning correctly.

The default implementation returns false

WARNING: if you override this default, you MUST ensure that the operator’s maintains the ordering invariant or else DataFusion may produce incorrect results.

Returns true if this operator would benefit from partitioning its input (and thus from more parallelism). For operators that do very little work the overhead of extra parallelism may outweigh any benefits

The default implementation returns true unless this operator has signalled it requiers a single child input partition.

Return a snapshot of the set of Metrics for this ExecutionPlan.

While the values of the metrics in the returned MetricsSets may change as execution progresses, the specific metrics will not.

Once self.execute() has returned (technically the future is resolved) for all available partitions, the set of metrics should be complete. If this function is called prior to execute() new metrics may appear in subsequent calls.

Format this ExecutionPlan to f in the specified type.

Should not include a newline

Note this function prints a placeholder by default to preserve backwards compatibility.

Implementors