datafusion_spark/function/hash/
crc32.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 std::any::Any;
19use std::sync::Arc;
20
21use arrow::array::{ArrayRef, Int64Array};
22use arrow::datatypes::{DataType, Field, FieldRef};
23use crc32fast::Hasher;
24use datafusion_common::cast::{
25    as_binary_array, as_binary_view_array, as_fixed_size_binary_array,
26    as_large_binary_array,
27};
28use datafusion_common::types::{NativeType, logical_string};
29use datafusion_common::utils::take_function_args;
30use datafusion_common::{Result, internal_err};
31use datafusion_expr::{
32    Coercion, ColumnarValue, ReturnFieldArgs, ScalarFunctionArgs, ScalarUDFImpl,
33    Signature, TypeSignatureClass, Volatility,
34};
35use datafusion_functions::utils::make_scalar_function;
36
37/// <https://spark.apache.org/docs/latest/api/sql/index.html#crc32>
38#[derive(Debug, PartialEq, Eq, Hash)]
39pub struct SparkCrc32 {
40    signature: Signature,
41}
42
43impl Default for SparkCrc32 {
44    fn default() -> Self {
45        Self::new()
46    }
47}
48
49impl SparkCrc32 {
50    pub fn new() -> Self {
51        Self {
52            signature: Signature::coercible(
53                vec![Coercion::new_implicit(
54                    TypeSignatureClass::Binary,
55                    vec![TypeSignatureClass::Native(logical_string())],
56                    NativeType::Binary,
57                )],
58                Volatility::Immutable,
59            ),
60        }
61    }
62}
63
64impl ScalarUDFImpl for SparkCrc32 {
65    fn as_any(&self) -> &dyn Any {
66        self
67    }
68
69    fn name(&self) -> &str {
70        "crc32"
71    }
72
73    fn signature(&self) -> &Signature {
74        &self.signature
75    }
76
77    fn return_type(&self, _arg_types: &[DataType]) -> Result<DataType> {
78        internal_err!("return_field_from_args should be used instead")
79    }
80
81    fn return_field_from_args(&self, args: ReturnFieldArgs) -> Result<FieldRef> {
82        let nullable = args.arg_fields.iter().any(|f| f.is_nullable());
83        Ok(Arc::new(Field::new(self.name(), DataType::Int64, nullable)))
84    }
85
86    fn invoke_with_args(&self, args: ScalarFunctionArgs) -> Result<ColumnarValue> {
87        make_scalar_function(spark_crc32, vec![])(&args.args)
88    }
89}
90
91fn spark_crc32_digest(value: &[u8]) -> i64 {
92    let mut hasher = Hasher::new();
93    hasher.update(value);
94    hasher.finalize() as i64
95}
96
97fn spark_crc32_impl<'a>(input: impl Iterator<Item = Option<&'a [u8]>>) -> ArrayRef {
98    let result = input
99        .map(|value| value.map(spark_crc32_digest))
100        .collect::<Int64Array>();
101    Arc::new(result)
102}
103
104fn spark_crc32(args: &[ArrayRef]) -> Result<ArrayRef> {
105    let [input] = take_function_args("crc32", args)?;
106
107    match input.data_type() {
108        DataType::Null => Ok(Arc::new(Int64Array::new_null(input.len()))),
109        DataType::Binary => {
110            let input = as_binary_array(input)?;
111            Ok(spark_crc32_impl(input.iter()))
112        }
113        DataType::LargeBinary => {
114            let input = as_large_binary_array(input)?;
115            Ok(spark_crc32_impl(input.iter()))
116        }
117        DataType::BinaryView => {
118            let input = as_binary_view_array(input)?;
119            Ok(spark_crc32_impl(input.iter()))
120        }
121        DataType::FixedSizeBinary(_) => {
122            let input = as_fixed_size_binary_array(input)?;
123            Ok(spark_crc32_impl(input.iter()))
124        }
125        dt => {
126            internal_err!("Unsupported data type for crc32: {dt}")
127        }
128    }
129}
130
131#[cfg(test)]
132mod tests {
133    use super::*;
134
135    #[test]
136    fn test_crc32_nullability() -> Result<()> {
137        let crc32_func = SparkCrc32::new();
138
139        // non-nullable field should produce non-nullable output
140        let field_not_null = Arc::new(Field::new("data", DataType::Binary, false));
141        let result = crc32_func.return_field_from_args(ReturnFieldArgs {
142            arg_fields: std::slice::from_ref(&field_not_null),
143            scalar_arguments: &[None],
144        })?;
145        assert!(!result.is_nullable());
146        assert_eq!(result.data_type(), &DataType::Int64);
147
148        // nullable field should produce nullable output
149        let field_nullable = Arc::new(Field::new("data", DataType::Binary, true));
150        let result = crc32_func.return_field_from_args(ReturnFieldArgs {
151            arg_fields: &[field_nullable],
152            scalar_arguments: &[None],
153        })?;
154        assert!(result.is_nullable());
155        assert_eq!(result.data_type(), &DataType::Int64);
156
157        Ok(())
158    }
159}
datafusion_spark/function/hash/crc32.rs

datafusion_spark/function/hash/
crc32.rs
