yashaggarwal85d commented on code in PR #26637: URL: https://github.com/apache/datafusion/pull/26637#discussion_r4379401480
########## datafusion/physical-plan/src/aggregates/clustered_test_util.rs: ########## @@ -0,0 +1,210 @@ +// 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. + +//! Test helpers for the clustered aggregate streams under memory pressure. + +use std::sync::Arc; + +use arrow::array::Int64Array; +use arrow::datatypes::{DataType, Field, Schema}; +use arrow::record_batch::RecordBatch; +use datafusion_common::test_util::batches_to_string; +use datafusion_common::{DataFusionError, Result}; +use datafusion_execution::TaskContext; +use datafusion_execution::config::SessionConfig; +use datafusion_execution::disk_manager::{DiskManagerBuilder, DiskManagerMode}; +use datafusion_execution::memory_pool::{ + GreedyMemoryPool, MemoryPool, UnboundedMemoryPool, +}; +use datafusion_execution::runtime_env::RuntimeEnvBuilder; +use datafusion_functions_aggregate::sum::sum_udaf; +use datafusion_physical_expr::PhysicalSortExpr; +use datafusion_physical_expr::aggregate::AggregateExprBuilder; +use datafusion_physical_expr::expressions::col; +use datafusion_physical_expr_common::sort_expr::LexOrdering; +use futures::StreamExt; + +use super::{AggregateExec, AggregateMode, GroupClusteringMode, PhysicalGroupBy}; +use crate::ExecutionPlan; +use crate::test::TestMemoryExec; + +/// Key columns of two batches sorted on `k` that share the boundary key 256, so +/// groups complete within each batch under full group clustering. +pub(super) fn full_batches() -> Vec<Vec<Vec<i64>>> { + vec![vec![(0..=256).collect()], vec![(256..=512).collect()]] +} + +/// Key columns `(a, b)` of batches sorted on `a` only, with 256 groups for each +/// `a` spread over two batches. The groups of one `a` complete together when the +/// next `a` starts. +pub(super) fn partial_batches() -> Vec<Vec<Vec<i64>>> { + [0, 0, 1, 1, 2, 2] + .into_iter() + .map(|a| vec![vec![a; 256], (0..256).collect()]) + .collect() +} + +/// Key column of two batches that hold the same single key, so no group +/// completes before the input ends. +pub(super) fn single_key_batches() -> Vec<Vec<Vec<i64>>> { + vec![vec![vec![0; 512]]; 2] +} + +/// Builds `SUM(v) GROUP BY <keys>` in `mode` over `batches` of `(keys..., v)` +/// columns, where `v` equals the last key. The input is declared sorted on the +/// first key only, so one key gives full group clustering and more keys give +/// partial group clustering. +/// +/// For `Final` mode, the batches are the partial states, since the state of +/// `SUM(v)` is a single `Int64` column. +pub(super) fn sum_over_sorted_input( Review Comment: Good spot. It now builds the key arrays once and pushes an `Arc::clone` of the last one for `v` (c3ea1ce). -- 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]
