Apache DataFusion SQL Query Engine
Fix within group aggregates with unparser (#22195)
Peter Ld61052a968ab
diff --git a/datafusion/sql/src/unparser/expr.rs b/datafusion/sql/src/unparser/expr.rsindex c659d8694e..89560b2379 100644--- a/datafusion/sql/src/unparser/expr.rs+++ b/datafusion/sql/src/unparser/expr.rs@@ -402,20 +402,25 @@ impl Unparser<'_> {..} = &agg.params;- let args = self.function_args_to_sql(args)?;+ let args_to_use;+ let within_group;++ // if this is a WITHIN GROUP aggregate, skip the prepended arg+ if agg.func.supports_within_group_clause() && !order_by.is_empty() {+ args_to_use = self.function_args_to_sql(&args[1..])?;+ within_group = order_by+ .iter()+ .map(|sort_expr| self.sort_to_sql(sort_expr))+ .collect::<Result<Vec<ast::OrderByExpr>>>()?;+ } else {+ args_to_use = self.function_args_to_sql(args)?;+ within_group = Vec::new();+ }+let filter = match filter {Some(filter) => Some(Box::new(self.expr_to_sql_inner(filter)?)),None => None,};- let within_group: Vec<ast::OrderByExpr> =- if agg.func.supports_within_group_clause() {- order_by- .iter()- .map(|sort_expr| self.sort_to_sql(sort_expr))- .collect::<Result<Vec<ast::OrderByExpr>>>()?- } else {- Vec::new()- };Ok(ast::Expr::Function(Function {name: ObjectName::from(vec![Ident {value: func_name.to_string(),@@ -425,7 +430,7 @@ impl Unparser<'_> {args: ast::FunctionArguments::List(ast::FunctionArgumentList {duplicate_treatment: distinct.then_some(DuplicateTreatment::Distinct),- args,+ args: args_to_use,clauses: vec![],}),filter,diff --git a/datafusion/sql/tests/cases/plan_to_sql.rs b/datafusion/sql/tests/cases/plan_to_sql.rsindex 2194085e65..d6c31570bf 100644--- a/datafusion/sql/tests/cases/plan_to_sql.rs+++ b/datafusion/sql/tests/cases/plan_to_sql.rs@@ -313,6 +313,12 @@ macro_rules! roundtrip_statement_with_dialect_helper {let state = MockSessionState::default().with_aggregate_function(max_udaf()).with_aggregate_function(min_udaf())+ .with_aggregate_function(+ datafusion_functions_aggregate::approx_percentile_cont::approx_percentile_cont_udaf(),+ )+ .with_aggregate_function(+ datafusion_functions_aggregate::percentile_cont::percentile_cont_udaf(),+ ).with_expr_planner(Arc::new(CoreFunctionPlanner::default())).with_expr_planner(Arc::new(NestedFunctionPlanner)).with_expr_planner(Arc::new(FieldAccessPlanner));@@ -4345,6 +4351,40 @@ fn snowflake_flatten_cross_join_unnest_table_column() -> Result<(), DataFusionErOk(())}+#[test]+fn roundtrip_approx_percentile_cont_within_group() -> Result<(), DataFusionError> {+ roundtrip_statement_with_dialect_helper!(+ sql: "SELECT approx_percentile_cont(0.5) WITHIN GROUP (ORDER BY salary) FROM person",+ parser_dialect: GenericDialect {},+ unparser_dialect: UnparserDefaultDialect {},+ expected: @"SELECT approx_percentile_cont(0.5) WITHIN GROUP (ORDER BY person.salary ASC NULLS LAST) FROM person",+ );+ Ok(())+}++#[test]+fn roundtrip_percentile_cont_within_group() -> Result<(), DataFusionError> {+ roundtrip_statement_with_dialect_helper!(+ sql: "SELECT percentile_cont(0.5) WITHIN GROUP (ORDER BY salary) FROM person",+ parser_dialect: GenericDialect {},+ unparser_dialect: UnparserDefaultDialect {},+ expected: @"SELECT percentile_cont(0.5) WITHIN GROUP (ORDER BY person.salary ASC NULLS LAST) FROM person",+ );+ Ok(())+}++#[test]+fn roundtrip_approx_percentile_cont_within_group_with_centroids()+-> Result<(), DataFusionError> {+ roundtrip_statement_with_dialect_helper!(+ sql: "SELECT approx_percentile_cont(0.9, 200) WITHIN GROUP (ORDER BY salary * 2 DESC) FROM person",+ parser_dialect: GenericDialect {},+ unparser_dialect: UnparserDefaultDialect {},+ expected: @"SELECT approx_percentile_cont(0.9, 200) WITHIN GROUP (ORDER BY (person.salary * 2) DESC NULLS FIRST) FROM person",+ );+ Ok(())+}+#[test]fn snowflake_flatten_multiple_unnest_cross_join() -> Result<(), DataFusionError> {// Realistic Snowflake pattern: