houqp commented on a change in pull request #605:
URL: https://github.com/apache/arrow-datafusion/pull/605#discussion_r667388265
##########
File path: datafusion/src/logical_plan/plan.rs
##########
@@ -354,6 +356,43 @@ impl LogicalPlan {
| LogicalPlan::CreateExternalTable { .. } => vec![],
}
}
+
+ /// returns all `Using` join columns in a logical plan
+ pub fn using_columns(&self) -> Result<Vec<HashSet<Column>>,
DataFusionError> {
+ struct UsingJoinColumnVisitor {
+ using_columns: Vec<HashSet<Column>>,
+ }
+
+ impl PlanVisitor for UsingJoinColumnVisitor {
+ type Error = DataFusionError;
+
+ fn pre_visit(&mut self, plan: &LogicalPlan) -> Result<bool,
Self::Error> {
+ if let LogicalPlan::Join {
+ join_constraint: JoinConstraint::Using,
+ on,
+ ..
+ } = plan
+ {
+ self.using_columns.push(
+ on.iter()
+ .map(|entry| {
+ std::iter::once(entry.0.clone())
+ .chain(std::iter::once(entry.1.clone()))
Review comment:
@Jimexist sent https://github.com/apache/arrow-datafusion/pull/704 to
address this. I found a comprise that's more readable and avoids the extra
memory allocation.
Here is my test to compare the code gen with different approaches:
https://godbolt.org/z/j3dcjbnvM.
--
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.
To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
