datafusion_functions/core/
least.rs

1// Licensed to the Apache Software Foundation (ASF) under one
2// or more contributor license agreements.  See the NOTICE file
3// distributed with this work for additional information
4// regarding copyright ownership.  The ASF licenses this file
5// to you under the Apache License, Version 2.0 (the
6// "License"); you may not use this file except in compliance
7// with the License.  You may obtain a copy of the License at
8//
9//   http://www.apache.org/licenses/LICENSE-2.0
10//
11// Unless required by applicable law or agreed to in writing,
12// software distributed under the License is distributed on an
13// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
14// KIND, either express or implied.  See the License for the
15// specific language governing permissions and limitations
16// under the License.
17
18use crate::core::greatest_least_utils::GreatestLeastOperator;
19use arrow::array::{make_comparator, Array, BooleanArray};
20use arrow::buffer::BooleanBuffer;
21use arrow::compute::kernels::cmp;
22use arrow::compute::SortOptions;
23use arrow::datatypes::DataType;
24use datafusion_common::{internal_err, Result, ScalarValue};
25use datafusion_doc::Documentation;
26use datafusion_expr::{ColumnarValue, ScalarFunctionArgs};
27use datafusion_expr::{ScalarUDFImpl, Signature, Volatility};
28use datafusion_macros::user_doc;
29use std::any::Any;
30
31const SORT_OPTIONS: SortOptions = SortOptions {
32    // Having the smallest result first
33    descending: false,
34
35    // NULL will be greater than any other value
36    nulls_first: false,
37};
38
39#[user_doc(
40    doc_section(label = "Conditional Functions"),
41    description = "Returns the smallest value in a list of expressions. Returns _null_ if all expressions are _null_.",
42    syntax_example = "least(expression1[, ..., expression_n])",
43    sql_example = r#"```sql
44> select least(4, 7, 5);
45+---------------------------+
46| least(4,7,5)              |
47+---------------------------+
48| 4                         |
49+---------------------------+
50```"#,
51    argument(
52        name = "expression1, expression_n",
53        description = "Expressions to compare and return the smallest value. Can be a constant, column, or function, and any combination of arithmetic operators. Pass as many expression arguments as necessary."
54    )
55)]
56#[derive(Debug)]
57pub struct LeastFunc {
58    signature: Signature,
59}
60
61impl Default for LeastFunc {
62    fn default() -> Self {
63        LeastFunc::new()
64    }
65}
66
67impl LeastFunc {
68    pub fn new() -> Self {
69        Self {
70            signature: Signature::user_defined(Volatility::Immutable),
71        }
72    }
73}
74
75impl GreatestLeastOperator for LeastFunc {
76    const NAME: &'static str = "least";
77
78    fn keep_scalar<'a>(
79        lhs: &'a ScalarValue,
80        rhs: &'a ScalarValue,
81    ) -> Result<&'a ScalarValue> {
82        // Manual checking for nulls as:
83        // 1. If we're going to use <=, in Rust None is smaller than Some(T), which we don't want
84        // 2. And we can't use make_comparator as it has no natural order (Arrow error)
85        if lhs.is_null() {
86            return Ok(rhs);
87        }
88
89        if rhs.is_null() {
90            return Ok(lhs);
91        }
92
93        if !lhs.data_type().is_nested() {
94            return if lhs <= rhs { Ok(lhs) } else { Ok(rhs) };
95        }
96
97        // Not using <= as in Rust None is smaller than Some(T)
98
99        // If complex type we can't compare directly as we want null values to be larger
100        let cmp = make_comparator(
101            lhs.to_array()?.as_ref(),
102            rhs.to_array()?.as_ref(),
103            SORT_OPTIONS,
104        )?;
105
106        if cmp(0, 0).is_le() {
107            Ok(lhs)
108        } else {
109            Ok(rhs)
110        }
111    }
112
113    /// Return boolean array where `arr[i] = lhs[i] <= rhs[i]` for all i, where `arr` is the result array
114    /// Nulls are always considered larger than any other value
115    fn get_indexes_to_keep(lhs: &dyn Array, rhs: &dyn Array) -> Result<BooleanArray> {
116        // Fast path:
117        // If both arrays are not nested, have the same length and no nulls, we can use the faster vectorized kernel
118        // - If both arrays are not nested: Nested types, such as lists, are not supported as the null semantics are not well-defined.
119        // - both array does not have any nulls: cmp::lt_eq will return null if any of the input is null while we want to return false in that case
120        if !lhs.data_type().is_nested()
121            && lhs.logical_null_count() == 0
122            && rhs.logical_null_count() == 0
123        {
124            return cmp::lt_eq(&lhs, &rhs).map_err(|e| e.into());
125        }
126
127        let cmp = make_comparator(lhs, rhs, SORT_OPTIONS)?;
128
129        if lhs.len() != rhs.len() {
130            return internal_err!(
131                "All arrays should have the same length for least comparison"
132            );
133        }
134
135        let values = BooleanBuffer::collect_bool(lhs.len(), |i| cmp(i, i).is_le());
136
137        // No nulls as we only want to keep the values that are smaller, its either true or false
138        Ok(BooleanArray::new(values, None))
139    }
140}
141
142impl ScalarUDFImpl for LeastFunc {
143    fn as_any(&self) -> &dyn Any {
144        self
145    }
146
147    fn name(&self) -> &str {
148        "least"
149    }
150
151    fn signature(&self) -> &Signature {
152        &self.signature
153    }
154
155    fn return_type(&self, arg_types: &[DataType]) -> Result<DataType> {
156        Ok(arg_types[0].clone())
157    }
158
159    fn invoke_with_args(&self, args: ScalarFunctionArgs) -> Result<ColumnarValue> {
160        super::greatest_least_utils::execute_conditional::<Self>(&args.args)
161    }
162
163    fn coerce_types(&self, arg_types: &[DataType]) -> Result<Vec<DataType>> {
164        let coerced_type =
165            super::greatest_least_utils::find_coerced_type::<Self>(arg_types)?;
166
167        Ok(vec![coerced_type; arg_types.len()])
168    }
169
170    fn documentation(&self) -> Option<&Documentation> {
171        self.doc()
172    }
173}
174
175#[cfg(test)]
176mod test {
177    use crate::core::least::LeastFunc;
178    use arrow::datatypes::DataType;
179    use datafusion_expr::ScalarUDFImpl;
180
181    #[test]
182    fn test_least_return_types_without_common_supertype_in_arg_type() {
183        let least = LeastFunc::new();
184        let return_type = least
185            .coerce_types(&[DataType::Decimal128(10, 3), DataType::Decimal128(10, 4)])
186            .unwrap();
187        assert_eq!(
188            return_type,
189            vec![DataType::Decimal128(11, 4), DataType::Decimal128(11, 4)]
190        );
191    }
192}
datafusion_functions/core/least.rs

datafusion_functions/core/
least.rs
