alamb commented on a change in pull request #8034:
URL: https://github.com/apache/arrow/pull/8034#discussion_r475537669
##########
File path: rust/datafusion/src/execution/physical_plan/mod.rs
##########
@@ -74,6 +85,15 @@ impl Partitioning {
}
}
+/// Distribution schemes
+#[derive(Debug, Clone)]
+pub enum Distribution {
+ /// Unspecified distribution
Review comment:
I think "unspecified" is more accurate -- specifically the data is
distributed around all the partitions, but the optimizer/planner doesn't know
*how* it is distributed
##########
File path: rust/datafusion/src/execution/physical_plan/parquet.rs
##########
@@ -91,11 +91,26 @@ impl ExecutionPlan for ParquetExec {
self.schema.clone()
}
+ fn children(&self) -> Vec<Arc<dyn ExecutionPlan>> {
+ // this is a leaf node and has no children
+ vec![]
+ }
+
/// Get the output partitioning of this plan
fn output_partitioning(&self) -> Partitioning {
Partitioning::UnknownPartitioning(self.filenames.len())
}
+ fn with_new_children(
+ &self,
+ _: Vec<Arc<dyn ExecutionPlan>>,
+ ) -> Result<Arc<dyn ExecutionPlan>> {
+ Err(ExecutionError::General(format!(
Review comment:
This is fine for this PR, but I think this implementation could actually
check that children was empty and then return `self.clone()`.
##########
File path: rust/datafusion/src/execution/physical_plan/mod.rs
##########
@@ -50,6 +50,17 @@ pub trait ExecutionPlan: Debug + Send + Sync {
fn schema(&self) -> SchemaRef;
/// Specifies the output partitioning scheme of this plan
fn output_partitioning(&self) -> Partitioning;
+ /// Specifies the data distribution requirements of all the children for
this operator
+ fn required_child_distribution(&self) -> Distribution {
+ Distribution::UnspecifiedDistribution
+ }
+ /// Get the children of this plan
Review comment:
Or maybe
```
/// `children` of this plan. This corresponds to all plans that provide
input to this plan.
/// This function should return an empty vector for leaf nodes such as
`scans`.
```
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
