query/optimizer/
remove_duplicate.rs

1// Copyright 2023 Greptime Team
2//
3// Licensed under the Apache License, Version 2.0 (the "License");
4// you may not use this file except in compliance with the License.
5// You may obtain a copy of the License at
6//
7//     http://www.apache.org/licenses/LICENSE-2.0
8//
9// Unless required by applicable law or agreed to in writing, software
10// distributed under the License is distributed on an "AS IS" BASIS,
11// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12// See the License for the specific language governing permissions and
13// limitations under the License.
14
15use std::sync::Arc;
16
17use datafusion::config::ConfigOptions;
18use datafusion::physical_optimizer::PhysicalOptimizerRule;
19use datafusion::physical_plan::coalesce_batches::CoalesceBatchesExec;
20use datafusion::physical_plan::repartition::RepartitionExec;
21use datafusion::physical_plan::ExecutionPlan;
22use datafusion_common::tree_node::{Transformed, TreeNode};
23use datafusion_common::Result as DfResult;
24
25/// This is [PhysicalOptimizerRule] to remove duplicate physical plans such as two
26/// adjoining [CoalesceBatchesExec] or [RepartitionExec]. They won't have any effect
27/// if one runs right after another.
28///
29/// This rule is expected to be run in the final stage of the optimization process.
30#[derive(Debug)]
31pub struct RemoveDuplicate;
32
33impl PhysicalOptimizerRule for RemoveDuplicate {
34    fn optimize(
35        &self,
36        plan: Arc<dyn ExecutionPlan>,
37        _config: &ConfigOptions,
38    ) -> DfResult<Arc<dyn ExecutionPlan>> {
39        Self::do_optimize(plan)
40    }
41
42    fn name(&self) -> &str {
43        "RemoveDuplicateRule"
44    }
45
46    fn schema_check(&self) -> bool {
47        false
48    }
49}
50
51impl RemoveDuplicate {
52    fn do_optimize(plan: Arc<dyn ExecutionPlan>) -> DfResult<Arc<dyn ExecutionPlan>> {
53        let result = plan
54            .transform_down(|plan| {
55                if plan.as_any().is::<CoalesceBatchesExec>()
56                    || plan.as_any().is::<RepartitionExec>()
57                {
58                    // check child
59                    let child = plan.children()[0].clone();
60                    if child.as_any().type_id() == plan.as_any().type_id() {
61                        // remove child
62                        let grand_child = child.children()[0].clone();
63                        let new_plan = plan.with_new_children(vec![grand_child])?;
64                        return Ok(Transformed::yes(new_plan));
65                    }
66                }
67
68                Ok(Transformed::no(plan))
69            })?
70            .data;
71
72        Ok(result)
73    }
74}
75
76#[cfg(test)]
77mod test {
78    use std::sync::Arc;
79
80    use arrow_schema::Schema;
81    use datafusion::physical_plan::displayable;
82    use datafusion::physical_plan::empty::EmptyExec;
83    use datafusion_physical_expr::Partitioning;
84
85    use super::*;
86
87    #[test]
88    fn remove_coalesce_batches() {
89        let empty = Arc::new(EmptyExec::new(Arc::new(Schema::empty())));
90        let coalesce_batches = Arc::new(CoalesceBatchesExec::new(empty, 1024));
91        let another_coalesce_batches = Arc::new(CoalesceBatchesExec::new(coalesce_batches, 8192));
92
93        let optimized = RemoveDuplicate::do_optimize(another_coalesce_batches).unwrap();
94        let formatted = displayable(optimized.as_ref()).indent(true).to_string();
95        let expected = "CoalesceBatchesExec: target_batch_size=8192\
96        \n  EmptyExec\n";
97
98        assert_eq!(expected, formatted);
99    }
100
101    #[test]
102    fn non_continuous_coalesce_batches() {
103        let empty = Arc::new(EmptyExec::new(Arc::new(Schema::empty())));
104        let coalesce_batches = Arc::new(CoalesceBatchesExec::new(empty, 1024));
105        let repartition = Arc::new(
106            RepartitionExec::try_new(coalesce_batches, Partitioning::UnknownPartitioning(1))
107                .unwrap(),
108        );
109        let another_coalesce_batches = Arc::new(CoalesceBatchesExec::new(repartition, 8192));
110
111        let optimized = RemoveDuplicate::do_optimize(another_coalesce_batches).unwrap();
112        let formatted = displayable(optimized.as_ref()).indent(true).to_string();
113        let expected = "CoalesceBatchesExec: target_batch_size=8192\
114        \n  RepartitionExec: partitioning=UnknownPartitioning(1), input_partitions=1\
115        \n    CoalesceBatchesExec: target_batch_size=1024\
116        \n      EmptyExec\n";
117
118        assert_eq!(expected, formatted);
119    }
120}
query/optimizer/remove_duplicate.rs

query/optimizer/
remove_duplicate.rs
