risingwave_frontend/optimizer/plan_node/
logical_dedup.rs

1// Copyright 2025 RisingWave Labs
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 fixedbitset::FixedBitSet;
16use itertools::Itertools;
17use risingwave_common::util::column_index_mapping::ColIndexMapping;
18
19use super::generic::{GenericPlanRef, TopNLimit};
20use super::utils::impl_distill_by_unit;
21use super::{
22    BatchGroupTopN, BatchPlanRef, ColPrunable, ColumnPruningContext, ExprRewritable, Logical,
23    LogicalPlanRef as PlanRef, LogicalProject, PlanBase, PlanTreeNodeUnary, PredicatePushdown,
24    PredicatePushdownContext, RewriteStreamContext, StreamDedup, StreamGroupTopN, ToBatch,
25    ToStream, ToStreamContext, gen_filter_and_pushdown, generic,
26};
27use crate::error::Result;
28use crate::optimizer::plan_node::expr_visitable::ExprVisitable;
29use crate::optimizer::property::{Order, RequiredDist};
30use crate::utils::Condition;
31
32/// [`LogicalDedup`] deduplicates data on specific columns. It is now used in `DISTINCT ON` without
33/// an `ORDER BY`.
34#[derive(Debug, Clone, PartialEq, Eq, Hash)]
35pub struct LogicalDedup {
36    pub base: PlanBase<Logical>,
37    core: generic::Dedup<PlanRef>,
38}
39
40impl LogicalDedup {
41    pub fn new(input: PlanRef, dedup_cols: Vec<usize>) -> Self {
42        let core = generic::Dedup::new(input, dedup_cols);
43        let base = PlanBase::new_logical_with_core(&core);
44        LogicalDedup { base, core }
45    }
46
47    pub fn dedup_cols(&self) -> &[usize] {
48        &self.core.dedup_cols
49    }
50}
51
52impl PlanTreeNodeUnary<Logical> for LogicalDedup {
53    fn input(&self) -> PlanRef {
54        self.core.input.clone()
55    }
56
57    fn clone_with_input(&self, input: PlanRef) -> Self {
58        Self::new(input, self.dedup_cols().to_vec())
59    }
60
61    fn rewrite_with_input(
62        &self,
63        input: PlanRef,
64        input_col_change: ColIndexMapping,
65    ) -> (Self, ColIndexMapping) {
66        (
67            Self::new(
68                input,
69                self.dedup_cols()
70                    .iter()
71                    .map(|idx| input_col_change.map(*idx))
72                    .collect_vec(),
73            ),
74            input_col_change,
75        )
76    }
77}
78
79impl_plan_tree_node_for_unary! { Logical, LogicalDedup}
80
81impl PredicatePushdown for LogicalDedup {
82    fn predicate_pushdown(
83        &self,
84        predicate: Condition,
85        ctx: &mut PredicatePushdownContext,
86    ) -> PlanRef {
87        gen_filter_and_pushdown(self, predicate, Condition::true_cond(), ctx)
88    }
89}
90
91impl ToStream for LogicalDedup {
92    fn logical_rewrite_for_stream(
93        &self,
94        ctx: &mut RewriteStreamContext,
95    ) -> Result<(PlanRef, ColIndexMapping)> {
96        let (input, input_col_change) = self.input().logical_rewrite_for_stream(ctx)?;
97        let (logical, out_col_change) = self.rewrite_with_input(input, input_col_change);
98        Ok((logical.into(), out_col_change))
99    }
100
101    fn to_stream(
102        &self,
103        ctx: &mut ToStreamContext,
104    ) -> Result<crate::optimizer::plan_node::StreamPlanRef> {
105        use super::stream::prelude::*;
106
107        let logical_input = self
108            .input()
109            .try_better_locality(self.dedup_cols())
110            .unwrap_or_else(|| self.input());
111        let input = logical_input.to_stream(ctx)?;
112        let input = RequiredDist::hash_shard(self.dedup_cols())
113            .streaming_enforce_if_not_satisfies(input)?;
114        if input.append_only() {
115            // `LogicalDedup` is transformed to `StreamDedup` only when the input is append-only.
116            let core = self.core.clone_with_input(input);
117            Ok(StreamDedup::new(core).into())
118        } else {
119            // If the input is not append-only, we use a `StreamGroupTopN` with the limit being 1.
120            let logical_top_n = generic::TopN::with_group(
121                input,
122                TopNLimit::new(1, false),
123                0,
124                Order::default(),
125                self.dedup_cols().to_vec(),
126            );
127            Ok(StreamGroupTopN::new(logical_top_n, None)?.into())
128        }
129    }
130}
131
132impl ToBatch for LogicalDedup {
133    fn to_batch(&self) -> Result<BatchPlanRef> {
134        let input = self.input().to_batch()?;
135        let logical_top_n = generic::TopN::with_group(
136            input,
137            TopNLimit::new(1, false),
138            0,
139            Order::default(),
140            self.dedup_cols().to_vec(),
141        );
142        Ok(BatchGroupTopN::new(logical_top_n).into())
143    }
144}
145
146impl ExprRewritable<Logical> for LogicalDedup {}
147
148impl ExprVisitable for LogicalDedup {}
149
150impl ColPrunable for LogicalDedup {
151    fn prune_col(&self, required_cols: &[usize], ctx: &mut ColumnPruningContext) -> PlanRef {
152        let input_required_bitset = FixedBitSet::from_iter(required_cols.iter().copied());
153        let dedup_required_bitset = {
154            let mut dedup_required_bitset = FixedBitSet::with_capacity(self.input().schema().len());
155            self.dedup_cols()
156                .iter()
157                .for_each(|idx| dedup_required_bitset.insert(*idx));
158            dedup_required_bitset
159        };
160        let input_required_cols = {
161            let mut tmp = input_required_bitset;
162            tmp.union_with(&dedup_required_bitset);
163            tmp.ones().collect_vec()
164        };
165        let mapping = ColIndexMapping::with_remaining_columns(
166            &input_required_cols,
167            self.input().schema().len(),
168        );
169
170        let new_input = self.input().prune_col(&input_required_cols, ctx);
171        let new_dedup_cols = self
172            .dedup_cols()
173            .iter()
174            .map(|&idx| mapping.map(idx))
175            .collect_vec();
176        let logical_dedup = Self::new(new_input, new_dedup_cols).into();
177
178        if input_required_cols == required_cols {
179            logical_dedup
180        } else {
181            let output_required_cols = required_cols
182                .iter()
183                .map(|&idx| mapping.map(idx))
184                .collect_vec();
185            let src_size = logical_dedup.schema().len();
186            LogicalProject::with_mapping(
187                logical_dedup,
188                ColIndexMapping::with_remaining_columns(&output_required_cols, src_size),
189            )
190            .into()
191        }
192    }
193}
194
195impl_distill_by_unit!(LogicalDedup, core, "LogicalDedup");
risingwave_frontend/optimizer/plan_node/logical_dedup.rs

risingwave_frontend/optimizer/plan_node/
logical_dedup.rs
