|
| 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 | + |
| 18 | +//! [`StringAgg`] and [`StringAggAccumulator`] accumulator for the `string_agg` function |
| 19 | +
|
| 20 | +use arrow::array::ArrayRef; |
| 21 | +use arrow_schema::DataType; |
| 22 | +use datafusion_common::cast::as_generic_string_array; |
| 23 | +use datafusion_common::Result; |
| 24 | +use datafusion_common::{not_impl_err, ScalarValue}; |
| 25 | +use datafusion_expr::function::AccumulatorArgs; |
| 26 | +use datafusion_expr::{ |
| 27 | + Accumulator, AggregateUDFImpl, Expr, Signature, TypeSignature, Volatility, |
| 28 | +}; |
| 29 | +use std::any::Any; |
| 30 | + |
| 31 | +make_udaf_expr_and_func!( |
| 32 | + StringAgg, |
| 33 | + string_agg, |
| 34 | + expr delimiter, |
| 35 | + "Concatenates the values of string expressions and places separator values between them", |
| 36 | + string_agg_udaf |
| 37 | +); |
| 38 | + |
| 39 | +/// STRING_AGG aggregate expression |
| 40 | +#[derive(Debug)] |
| 41 | +pub struct StringAgg { |
| 42 | + signature: Signature, |
| 43 | +} |
| 44 | + |
| 45 | +impl StringAgg { |
| 46 | + /// Create a new StringAgg aggregate function |
| 47 | + pub fn new() -> Self { |
| 48 | + Self { |
| 49 | + signature: Signature::one_of( |
| 50 | + vec![ |
| 51 | + TypeSignature::Exact(vec![DataType::LargeUtf8, DataType::Utf8]), |
| 52 | + TypeSignature::Exact(vec![DataType::LargeUtf8, DataType::LargeUtf8]), |
| 53 | + TypeSignature::Exact(vec![DataType::LargeUtf8, DataType::Null]), |
| 54 | + ], |
| 55 | + Volatility::Immutable, |
| 56 | + ), |
| 57 | + } |
| 58 | + } |
| 59 | +} |
| 60 | + |
| 61 | +impl Default for StringAgg { |
| 62 | + fn default() -> Self { |
| 63 | + Self::new() |
| 64 | + } |
| 65 | +} |
| 66 | + |
| 67 | +impl AggregateUDFImpl for StringAgg { |
| 68 | + fn as_any(&self) -> &dyn Any { |
| 69 | + self |
| 70 | + } |
| 71 | + |
| 72 | + fn name(&self) -> &str { |
| 73 | + "string_agg" |
| 74 | + } |
| 75 | + |
| 76 | + fn signature(&self) -> &Signature { |
| 77 | + &self.signature |
| 78 | + } |
| 79 | + |
| 80 | + fn return_type(&self, _arg_types: &[DataType]) -> Result<DataType> { |
| 81 | + Ok(DataType::LargeUtf8) |
| 82 | + } |
| 83 | + |
| 84 | + fn accumulator(&self, acc_args: AccumulatorArgs) -> Result<Box<dyn Accumulator>> { |
| 85 | + match &acc_args.input_exprs[1] { |
| 86 | + Expr::Literal(ScalarValue::Utf8(Some(delimiter))) |
| 87 | + | Expr::Literal(ScalarValue::LargeUtf8(Some(delimiter))) => { |
| 88 | + Ok(Box::new(StringAggAccumulator::new(delimiter))) |
| 89 | + } |
| 90 | + Expr::Literal(ScalarValue::Utf8(None)) |
| 91 | + | Expr::Literal(ScalarValue::LargeUtf8(None)) |
| 92 | + | Expr::Literal(ScalarValue::Null) => { |
| 93 | + Ok(Box::new(StringAggAccumulator::new(""))) |
| 94 | + } |
| 95 | + _ => not_impl_err!( |
| 96 | + "StringAgg not supported for delimiter {}", |
| 97 | + &acc_args.input_exprs[1] |
| 98 | + ), |
| 99 | + } |
| 100 | + } |
| 101 | +} |
| 102 | + |
| 103 | +#[derive(Debug)] |
| 104 | +pub(crate) struct StringAggAccumulator { |
| 105 | + values: Option<String>, |
| 106 | + delimiter: String, |
| 107 | +} |
| 108 | + |
| 109 | +impl StringAggAccumulator { |
| 110 | + pub fn new(delimiter: &str) -> Self { |
| 111 | + Self { |
| 112 | + values: None, |
| 113 | + delimiter: delimiter.to_string(), |
| 114 | + } |
| 115 | + } |
| 116 | +} |
| 117 | + |
| 118 | +impl Accumulator for StringAggAccumulator { |
| 119 | + fn update_batch(&mut self, values: &[ArrayRef]) -> Result<()> { |
| 120 | + let string_array: Vec<_> = as_generic_string_array::<i64>(&values[0])? |
| 121 | + .iter() |
| 122 | + .filter_map(|v| v.as_ref().map(ToString::to_string)) |
| 123 | + .collect(); |
| 124 | + if !string_array.is_empty() { |
| 125 | + let s = string_array.join(self.delimiter.as_str()); |
| 126 | + let v = self.values.get_or_insert("".to_string()); |
| 127 | + if !v.is_empty() { |
| 128 | + v.push_str(self.delimiter.as_str()); |
| 129 | + } |
| 130 | + v.push_str(s.as_str()); |
| 131 | + } |
| 132 | + Ok(()) |
| 133 | + } |
| 134 | + |
| 135 | + fn merge_batch(&mut self, values: &[ArrayRef]) -> Result<()> { |
| 136 | + self.update_batch(values)?; |
| 137 | + Ok(()) |
| 138 | + } |
| 139 | + |
| 140 | + fn state(&mut self) -> Result<Vec<ScalarValue>> { |
| 141 | + Ok(vec![self.evaluate()?]) |
| 142 | + } |
| 143 | + |
| 144 | + fn evaluate(&mut self) -> Result<ScalarValue> { |
| 145 | + Ok(ScalarValue::LargeUtf8(self.values.clone())) |
| 146 | + } |
| 147 | + |
| 148 | + fn size(&self) -> usize { |
| 149 | + std::mem::size_of_val(self) |
| 150 | + + self.values.as_ref().map(|v| v.capacity()).unwrap_or(0) |
| 151 | + + self.delimiter.capacity() |
| 152 | + } |
| 153 | +} |
0 commit comments