|
| 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 | +use criterion::*; |
| 19 | + |
| 20 | +use arrow::datatypes::*; |
| 21 | +use arrow::util::bench_util::{ |
| 22 | + create_primitive_array, create_string_array, create_string_array_with_len, |
| 23 | + create_string_dict_array, |
| 24 | +}; |
| 25 | +use arrow::util::test_util::seedable_rng; |
| 26 | +use arrow_array::{Array, ListArray, RecordBatch, StructArray}; |
| 27 | +use arrow_buffer::{BooleanBuffer, NullBuffer, OffsetBuffer}; |
| 28 | +use arrow_json::LineDelimitedWriter; |
| 29 | +use rand::Rng; |
| 30 | +use std::sync::Arc; |
| 31 | + |
| 32 | +const NUM_ROWS: usize = 65536; |
| 33 | + |
| 34 | +fn do_bench(c: &mut Criterion, name: &str, batch: &RecordBatch) { |
| 35 | + c.bench_function(name, |b| { |
| 36 | + b.iter(|| { |
| 37 | + let mut out = Vec::with_capacity(1024); |
| 38 | + LineDelimitedWriter::new(&mut out).write(&batch).unwrap(); |
| 39 | + out |
| 40 | + }) |
| 41 | + }); |
| 42 | +} |
| 43 | + |
| 44 | +fn create_mixed(len: usize) -> RecordBatch { |
| 45 | + let c1 = Arc::new(create_string_array::<i32>(len, 0.)); |
| 46 | + let c2 = Arc::new(create_primitive_array::<Int32Type>(len, 0.)); |
| 47 | + let c3 = Arc::new(create_primitive_array::<UInt32Type>(len, 0.)); |
| 48 | + let c4 = Arc::new(create_string_array_with_len::<i32>(len, 0.2, 10)); |
| 49 | + let c5 = Arc::new(create_string_array_with_len::<i32>(len, 0.2, 20)); |
| 50 | + let c6 = Arc::new(create_primitive_array::<Float32Type>(len, 0.2)); |
| 51 | + RecordBatch::try_from_iter([ |
| 52 | + ("c1", c1 as _), |
| 53 | + ("c2", c2 as _), |
| 54 | + ("c3", c3 as _), |
| 55 | + ("c4", c4 as _), |
| 56 | + ("c5", c5 as _), |
| 57 | + ("c6", c6 as _), |
| 58 | + ]) |
| 59 | + .unwrap() |
| 60 | +} |
| 61 | + |
| 62 | +fn create_nulls(len: usize) -> NullBuffer { |
| 63 | + let mut rng = seedable_rng(); |
| 64 | + BooleanBuffer::from_iter((0..len).map(|_| rng.gen_bool(0.2))).into() |
| 65 | +} |
| 66 | + |
| 67 | +fn create_offsets(len: usize) -> (usize, OffsetBuffer<i32>) { |
| 68 | + let mut rng = seedable_rng(); |
| 69 | + let mut last_offset = 0; |
| 70 | + let mut offsets = Vec::with_capacity(len + 1); |
| 71 | + offsets.push(0); |
| 72 | + for _ in 0..len { |
| 73 | + let len = rng.gen_range(0..10); |
| 74 | + offsets.push(last_offset + len); |
| 75 | + last_offset += len; |
| 76 | + } |
| 77 | + ( |
| 78 | + *offsets.last().unwrap() as _, |
| 79 | + OffsetBuffer::new(offsets.into()), |
| 80 | + ) |
| 81 | +} |
| 82 | + |
| 83 | +fn create_nullable_struct(len: usize) -> StructArray { |
| 84 | + let c2 = StructArray::from(create_mixed(len)); |
| 85 | + StructArray::new( |
| 86 | + c2.fields().clone(), |
| 87 | + c2.columns().to_vec(), |
| 88 | + Some(create_nulls(c2.len())), |
| 89 | + ) |
| 90 | +} |
| 91 | + |
| 92 | +fn bench_primitive(c: &mut Criterion) { |
| 93 | + let c1 = Arc::new(create_string_array::<i32>(NUM_ROWS, 0.)); |
| 94 | + let c2 = Arc::new(create_primitive_array::<Int32Type>(NUM_ROWS, 0.)); |
| 95 | + let c3 = Arc::new(create_primitive_array::<UInt32Type>(NUM_ROWS, 0.)); |
| 96 | + |
| 97 | + let batch = |
| 98 | + RecordBatch::try_from_iter([("c1", c1 as _), ("c2", c2 as _), ("c3", c3 as _)]).unwrap(); |
| 99 | + |
| 100 | + do_bench(c, "bench_primitive", &batch) |
| 101 | +} |
| 102 | + |
| 103 | +fn bench_mixed(c: &mut Criterion) { |
| 104 | + let batch = create_mixed(NUM_ROWS); |
| 105 | + do_bench(c, "bench_mixed", &batch) |
| 106 | +} |
| 107 | + |
| 108 | +fn bench_dict_array(c: &mut Criterion) { |
| 109 | + let c1 = Arc::new(create_string_array::<i32>(NUM_ROWS, 0.)); |
| 110 | + let c2 = Arc::new(create_string_dict_array::<Int32Type>(NUM_ROWS, 0., 20)); |
| 111 | + let c3 = Arc::new(create_string_dict_array::<Int32Type>(NUM_ROWS, 0.1, 20)); |
| 112 | + |
| 113 | + let batch = |
| 114 | + RecordBatch::try_from_iter([("c1", c1 as _), ("c2", c2 as _), ("c3", c3 as _)]).unwrap(); |
| 115 | + |
| 116 | + do_bench(c, "bench_dict_array", &batch) |
| 117 | +} |
| 118 | + |
| 119 | +fn bench_string_array(c: &mut Criterion) { |
| 120 | + let c1 = Arc::new(create_string_array::<i32>(NUM_ROWS, 0.)); |
| 121 | + let c2 = Arc::new(create_string_dict_array::<Int32Type>(NUM_ROWS, 0., 20)); |
| 122 | + let c3 = Arc::new(create_string_dict_array::<Int32Type>(NUM_ROWS, 0.1, 20)); |
| 123 | + |
| 124 | + let batch = |
| 125 | + RecordBatch::try_from_iter([("c1", c1 as _), ("c2", c2 as _), ("c3", c3 as _)]).unwrap(); |
| 126 | + |
| 127 | + do_bench(c, "bench_dict_array", &batch) |
| 128 | +} |
| 129 | + |
| 130 | +fn bench_struct(c: &mut Criterion) { |
| 131 | + let c1 = Arc::new(create_string_array::<i32>(NUM_ROWS, 0.)); |
| 132 | + let c2 = Arc::new(StructArray::from(create_mixed(NUM_ROWS))); |
| 133 | + let batch = RecordBatch::try_from_iter([("c1", c1 as _), ("c2", c2 as _)]).unwrap(); |
| 134 | + |
| 135 | + do_bench(c, "bench_struct", &batch) |
| 136 | +} |
| 137 | + |
| 138 | +fn bench_nullable_struct(c: &mut Criterion) { |
| 139 | + let c1 = Arc::new(create_string_array::<i32>(NUM_ROWS, 0.)); |
| 140 | + let c2 = Arc::new(create_nullable_struct(NUM_ROWS)); |
| 141 | + let batch = RecordBatch::try_from_iter([("c1", c1 as _), ("c2", c2 as _)]).unwrap(); |
| 142 | + |
| 143 | + do_bench(c, "bench_nullable_struct", &batch) |
| 144 | +} |
| 145 | + |
| 146 | +fn bench_list(c: &mut Criterion) { |
| 147 | + let (values_len, offsets) = create_offsets(NUM_ROWS); |
| 148 | + let c1_values = Arc::new(create_string_array::<i32>(values_len, 0.)); |
| 149 | + let c1_field = Arc::new(Field::new_list_field(c1_values.data_type().clone(), false)); |
| 150 | + let c1 = Arc::new(ListArray::new(c1_field, offsets, c1_values, None)); |
| 151 | + let batch = RecordBatch::try_from_iter([("c1", c1 as _)]).unwrap(); |
| 152 | + do_bench(c, "bench_list", &batch) |
| 153 | +} |
| 154 | + |
| 155 | +fn bench_nullable_list(c: &mut Criterion) { |
| 156 | + let (values_len, offsets) = create_offsets(NUM_ROWS); |
| 157 | + let c1_values = Arc::new(create_string_array::<i32>(values_len, 0.1)); |
| 158 | + let c1_field = Arc::new(Field::new_list_field(c1_values.data_type().clone(), true)); |
| 159 | + let c1_nulls = create_nulls(NUM_ROWS); |
| 160 | + let c1 = Arc::new(ListArray::new(c1_field, offsets, c1_values, Some(c1_nulls))); |
| 161 | + let batch = RecordBatch::try_from_iter([("c1", c1 as _)]).unwrap(); |
| 162 | + do_bench(c, "bench_nullable_list", &batch) |
| 163 | +} |
| 164 | + |
| 165 | +fn bench_struct_list(c: &mut Criterion) { |
| 166 | + let (values_len, offsets) = create_offsets(NUM_ROWS); |
| 167 | + let c1_values = Arc::new(create_nullable_struct(values_len)); |
| 168 | + let c1_field = Arc::new(Field::new_list_field(c1_values.data_type().clone(), true)); |
| 169 | + let c1_nulls = create_nulls(NUM_ROWS); |
| 170 | + let c1 = Arc::new(ListArray::new(c1_field, offsets, c1_values, Some(c1_nulls))); |
| 171 | + let batch = RecordBatch::try_from_iter([("c1", c1 as _)]).unwrap(); |
| 172 | + do_bench(c, "bench_struct_list", &batch) |
| 173 | +} |
| 174 | + |
| 175 | +fn criterion_benchmark(c: &mut Criterion) { |
| 176 | + bench_primitive(c); |
| 177 | + bench_string_array(c); |
| 178 | + bench_mixed(c); |
| 179 | + bench_dict_array(c); |
| 180 | + bench_struct(c); |
| 181 | + bench_nullable_struct(c); |
| 182 | + bench_list(c); |
| 183 | + bench_nullable_list(c); |
| 184 | + bench_struct_list(c); |
| 185 | +} |
| 186 | + |
| 187 | +criterion_group!(benches, criterion_benchmark); |
| 188 | +criterion_main!(benches); |
0 commit comments