Phoenix500526 commented on code in PR #23327:
URL: https://github.com/apache/datafusion/pull/23327#discussion_r3549751496
##########
datafusion/core/tests/sql/unparser.rs:
##########
@@ -399,6 +402,118 @@ async fn
optimized_duckdb_unparse_qualifies_nested_passthrough_column() -> Resul
Ok(())
}
+// https://github.com/apache/datafusion/issues/23317
+//
+// `SingleDistinctToGroupBy` can produce two stacked Aggregates where the inner
+// Aggregate defines intermediate fields such as `group_alias_0`, `alias1`, and
+// `alias2`. The unparser must preserve that inner Aggregate as a derived table
+// before the outer Aggregate references those fields.
+//
+// Without `SingleDistinctToGroupBy`, the Aggregate still sits over an unnamed
+// derived Projection. In that SQL scope, base table aliases `cs` and `c` are
no
+// longer visible, so aggregate expressions must refer to the derived table's
+// output columns unqualified.
+const ISSUE_23317_QUERY: &str = r#"
+WITH cohort AS (
+ SELECT
+ signup_year,
+ sum(customers) AS customers,
+ sum(revenue) AS revenue
+ FROM
+ (
+ SELECT
+ date_part('year', c.signup_date) AS signup_year,
+ count(DISTINCT cs.customer_id) AS customers,
+ round(sum(cs.total_revenue), 2) AS revenue
+ FROM
+ "warehouse"."main"."sales" cs
+ JOIN "warehouse"."main"."customers" c USING (customer_id)
+ GROUP BY
Review Comment:
Good catch. The sql has already parsed in
`assert_issue_23317_unparsed_sql_plans` now.
##########
datafusion/sql/src/unparser/plan.rs:
##########
Review Comment:
Done in commit b56a339
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
