datafusion_physical_expr/expressions/
like.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::PhysicalExpr;
19use arrow::datatypes::{DataType, Schema};
20use arrow::record_batch::RecordBatch;
21use datafusion_common::{internal_err, Result};
22use datafusion_expr::ColumnarValue;
23use datafusion_physical_expr_common::datum::apply_cmp;
24use std::hash::Hash;
25use std::{any::Any, sync::Arc};
26
27// Like expression
28#[derive(Debug, Eq)]
29pub struct LikeExpr {
30    negated: bool,
31    case_insensitive: bool,
32    expr: Arc<dyn PhysicalExpr>,
33    pattern: Arc<dyn PhysicalExpr>,
34}
35
36// Manually derive PartialEq and Hash to work around https://github.com/rust-lang/rust/issues/78808
37impl PartialEq for LikeExpr {
38    fn eq(&self, other: &Self) -> bool {
39        self.negated == other.negated
40            && self.case_insensitive == other.case_insensitive
41            && self.expr.eq(&other.expr)
42            && self.pattern.eq(&other.pattern)
43    }
44}
45
46impl Hash for LikeExpr {
47    fn hash<H: std::hash::Hasher>(&self, state: &mut H) {
48        self.negated.hash(state);
49        self.case_insensitive.hash(state);
50        self.expr.hash(state);
51        self.pattern.hash(state);
52    }
53}
54
55impl LikeExpr {
56    pub fn new(
57        negated: bool,
58        case_insensitive: bool,
59        expr: Arc<dyn PhysicalExpr>,
60        pattern: Arc<dyn PhysicalExpr>,
61    ) -> Self {
62        Self {
63            negated,
64            case_insensitive,
65            expr,
66            pattern,
67        }
68    }
69
70    /// Is negated
71    pub fn negated(&self) -> bool {
72        self.negated
73    }
74
75    /// Is case insensitive
76    pub fn case_insensitive(&self) -> bool {
77        self.case_insensitive
78    }
79
80    /// Input expression
81    pub fn expr(&self) -> &Arc<dyn PhysicalExpr> {
82        &self.expr
83    }
84
85    /// Pattern expression
86    pub fn pattern(&self) -> &Arc<dyn PhysicalExpr> {
87        &self.pattern
88    }
89
90    /// Operator name
91    fn op_name(&self) -> &str {
92        match (self.negated, self.case_insensitive) {
93            (false, false) => "LIKE",
94            (true, false) => "NOT LIKE",
95            (false, true) => "ILIKE",
96            (true, true) => "NOT ILIKE",
97        }
98    }
99}
100
101impl std::fmt::Display for LikeExpr {
102    fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
103        write!(f, "{} {} {}", self.expr, self.op_name(), self.pattern)
104    }
105}
106
107impl PhysicalExpr for LikeExpr {
108    fn as_any(&self) -> &dyn Any {
109        self
110    }
111
112    fn data_type(&self, _input_schema: &Schema) -> Result<DataType> {
113        Ok(DataType::Boolean)
114    }
115
116    fn nullable(&self, input_schema: &Schema) -> Result<bool> {
117        Ok(self.expr.nullable(input_schema)? || self.pattern.nullable(input_schema)?)
118    }
119
120    fn evaluate(&self, batch: &RecordBatch) -> Result<ColumnarValue> {
121        use arrow::compute::*;
122        let lhs = self.expr.evaluate(batch)?;
123        let rhs = self.pattern.evaluate(batch)?;
124        match (self.negated, self.case_insensitive) {
125            (false, false) => apply_cmp(&lhs, &rhs, like),
126            (false, true) => apply_cmp(&lhs, &rhs, ilike),
127            (true, false) => apply_cmp(&lhs, &rhs, nlike),
128            (true, true) => apply_cmp(&lhs, &rhs, nilike),
129        }
130    }
131
132    fn children(&self) -> Vec<&Arc<dyn PhysicalExpr>> {
133        vec![&self.expr, &self.pattern]
134    }
135
136    fn with_new_children(
137        self: Arc<Self>,
138        children: Vec<Arc<dyn PhysicalExpr>>,
139    ) -> Result<Arc<dyn PhysicalExpr>> {
140        Ok(Arc::new(LikeExpr::new(
141            self.negated,
142            self.case_insensitive,
143            Arc::clone(&children[0]),
144            Arc::clone(&children[1]),
145        )))
146    }
147
148    fn fmt_sql(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
149        self.expr.fmt_sql(f)?;
150        write!(f, " {} ", self.op_name())?;
151        self.pattern.fmt_sql(f)
152    }
153}
154
155/// used for optimize Dictionary like
156fn can_like_type(from_type: &DataType) -> bool {
157    match from_type {
158        DataType::Dictionary(_, inner_type_from) => **inner_type_from == DataType::Utf8,
159        _ => false,
160    }
161}
162
163/// Create a like expression, erroring if the argument types are not compatible.
164pub fn like(
165    negated: bool,
166    case_insensitive: bool,
167    expr: Arc<dyn PhysicalExpr>,
168    pattern: Arc<dyn PhysicalExpr>,
169    input_schema: &Schema,
170) -> Result<Arc<dyn PhysicalExpr>> {
171    let expr_type = &expr.data_type(input_schema)?;
172    let pattern_type = &pattern.data_type(input_schema)?;
173    if !expr_type.eq(pattern_type) && !can_like_type(expr_type) {
174        return internal_err!(
175            "The type of {expr_type} AND {pattern_type} of like physical should be same"
176        );
177    }
178    Ok(Arc::new(LikeExpr::new(
179        negated,
180        case_insensitive,
181        expr,
182        pattern,
183    )))
184}
185
186#[cfg(test)]
187mod test {
188    use super::*;
189    use crate::expressions::col;
190    use arrow::array::*;
191    use arrow::datatypes::Field;
192    use datafusion_common::cast::as_boolean_array;
193    use datafusion_physical_expr_common::physical_expr::fmt_sql;
194
195    macro_rules! test_like {
196        ($A_VEC:expr, $B_VEC:expr, $VEC:expr, $NULLABLE: expr, $NEGATED:expr, $CASE_INSENSITIVE:expr,) => {{
197            let schema = Schema::new(vec![
198                Field::new("a", DataType::Utf8, $NULLABLE),
199                Field::new("b", DataType::Utf8, $NULLABLE),
200            ]);
201            let a = StringArray::from($A_VEC);
202            let b = StringArray::from($B_VEC);
203
204            let expression = like(
205                $NEGATED,
206                $CASE_INSENSITIVE,
207                col("a", &schema)?,
208                col("b", &schema)?,
209                &schema,
210            )?;
211            let batch = RecordBatch::try_new(
212                Arc::new(schema.clone()),
213                vec![Arc::new(a), Arc::new(b)],
214            )?;
215
216            // compute
217            let result = expression
218                .evaluate(&batch)?
219                .into_array(batch.num_rows())
220                .expect("Failed to convert to array");
221            let result =
222                as_boolean_array(&result).expect("failed to downcast to BooleanArray");
223            let expected = &BooleanArray::from($VEC);
224            assert_eq!(expected, result);
225        }};
226    }
227
228    #[test]
229    fn like_op() -> Result<()> {
230        test_like!(
231            vec!["hello world", "world"],
232            vec!["%hello%", "%hello%"],
233            vec![true, false],
234            false,
235            false,
236            false,
237        ); // like
238        test_like!(
239            vec![Some("hello world"), None, Some("world")],
240            vec![Some("%hello%"), None, Some("%hello%")],
241            vec![Some(false), None, Some(true)],
242            true,
243            true,
244            false,
245        ); // not like
246        test_like!(
247            vec!["hello world", "world"],
248            vec!["%helLo%", "%helLo%"],
249            vec![true, false],
250            false,
251            false,
252            true,
253        ); // ilike
254        test_like!(
255            vec![Some("hello world"), None, Some("world")],
256            vec![Some("%helLo%"), None, Some("%helLo%")],
257            vec![Some(false), None, Some(true)],
258            true,
259            true,
260            true,
261        ); // not ilike
262
263        Ok(())
264    }
265
266    #[test]
267    fn test_fmt_sql() -> Result<()> {
268        let schema = Schema::new(vec![
269            Field::new("a", DataType::Utf8, false),
270            Field::new("b", DataType::Utf8, false),
271        ]);
272
273        let expr = like(
274            false,
275            false,
276            col("a", &schema)?,
277            col("b", &schema)?,
278            &schema,
279        )?;
280
281        // Display format
282        let display_string = expr.to_string();
283        assert_eq!(display_string, "a@0 LIKE b@1");
284
285        // fmt_sql format
286        let sql_string = fmt_sql(expr.as_ref()).to_string();
287        assert_eq!(sql_string, "a LIKE b");
288
289        Ok(())
290    }
291}
datafusion_physical_expr/expressions/like.rs

datafusion_physical_expr/expressions/
like.rs
