liukun4515 commented on a change in pull request #1387: URL: https://github.com/apache/arrow-datafusion/pull/1387#discussion_r763728811
########## File path: datafusion/src/physical_plan/coercion_rule/aggregate_rule.rs ########## @@ -0,0 +1,194 @@ +// Licensed to the Apache Software Foundation (ASF) under one +// or more contributor license agreements. See the NOTICE file +// distributed with this work for additional information +// regarding copyright ownership. The ASF licenses this file +// to you under the Apache License, Version 2.0 (the +// "License"); you may not use this file except in compliance +// with the License. You may obtain a copy of the License at +// +// http://www.apache.org/licenses/LICENSE-2.0 +// +// Unless required by applicable law or agreed to in writing, +// software distributed under the License is distributed on an +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +// KIND, either express or implied. See the License for the +// specific language governing permissions and limitations +// under the License. + +//! Support the coercion rule for aggregate function. + +use crate::arrow::datatypes::Schema; +use crate::error::{DataFusionError, Result}; +use crate::physical_plan::aggregates::AggregateFunction; +use crate::physical_plan::expressions::{ + is_avg_support_arg_type, is_sum_support_arg_type, try_cast, +}; +use crate::physical_plan::functions::{Signature, TypeSignature}; +use crate::physical_plan::PhysicalExpr; +use arrow::datatypes::DataType; +use std::ops::Deref; +use std::sync::Arc; + +pub fn coerce_types( + agg_fun: &AggregateFunction, + input_types: &[DataType], + signature: &Signature, +) -> Result<Vec<DataType>> { + match signature.type_signature { + TypeSignature::Uniform(agg_count, _) | TypeSignature::Any(agg_count) => { + if input_types.len() != agg_count { + return Err(DataFusionError::Plan(format!("The function {:?} expect argument number is {:?}, but the input argument number is {:?}", + agg_fun, agg_count, input_types.len()))); + } + } + _ => { + return Err(DataFusionError::Plan(format!( + "The aggregate coercion rule don't support this {:?}", + signature + ))); + } + }; + match agg_fun { + AggregateFunction::Count | AggregateFunction::ApproxDistinct => { + Ok(input_types.to_vec()) + } + AggregateFunction::ArrayAgg => Ok(input_types.to_vec()), + AggregateFunction::Min | AggregateFunction::Max => { + // min and max support the dictionary data type + // unpack the dictionary to get the value + get_min_max_result_type(input_types) + } + AggregateFunction::Sum => { + // Refer to https://www.postgresql.org/docs/8.2/functions-aggregate.html doc + // smallint, int, bigint, real, double precision, decimal, or interval. + if !is_sum_support_arg_type(&input_types[0]) { + return Err(DataFusionError::Plan(format!( + "The function {:?} do not support the {:?}.", + agg_fun, input_types[0] + ))); + } + Ok(input_types.to_vec()) + } + AggregateFunction::Avg => { + // Refer to https://www.postgresql.org/docs/8.2/functions-aggregate.html doc + // smallint, int, bigint, real, double precision, decimal, or interval + if !is_avg_support_arg_type(&input_types[0]) { + return Err(DataFusionError::Plan(format!( + "The function {:?} do not support the {:?}.", + agg_fun, input_types[0] + ))); + } + Ok(input_types.to_vec()) + } + } +} + +fn get_min_max_result_type(input_types: &[DataType]) -> Result<Vec<DataType>> { + // min and max support the dictionary data type + // unpack the dictionary to get the value + match &input_types[0] { + DataType::Dictionary(_, dict_value_type) => { + // TODO add checker, if the value type is complex data type + Ok(vec![dict_value_type.deref().clone()]) + } + // TODO add checker for datatype which min and max supported + // For example, the `Struct` and `Map` type are not supported in the MIN and MAX function + _ => Ok(input_types.to_vec()), + } +} + +pub fn coerce_exprs( + agg_fun: &AggregateFunction, + input_exprs: &[Arc<dyn PhysicalExpr>], + schema: &Schema, + signature: &Signature, +) -> Result<Vec<Arc<dyn PhysicalExpr>>> { + if input_exprs.is_empty() { + return Ok(vec![]); + } + let input_types = input_exprs + .iter() + .map(|e| e.data_type(schema)) + .collect::<Result<Vec<_>>>()?; + + // get the coerced data types + let coerced_types = coerce_types(agg_fun, &input_types, signature)?; + + // try cast if need + input_exprs + .iter() + .enumerate() + .map(|(i, expr)| try_cast(expr.clone(), schema, coerced_types[i].clone())) + .collect::<Result<Vec<_>>>() Review comment: great suggestions -- 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
