|
| 1 | +/* |
| 2 | + * Licensed to the Apache Software Foundation (ASF) under one |
| 3 | + * or more contributor license agreements. See the NOTICE file |
| 4 | + * distributed with this work for additional information |
| 5 | + * regarding copyright ownership. The ASF licenses this file |
| 6 | + * to you under the Apache License, Version 2.0 (the |
| 7 | + * "License"); you may not use this file except in compliance |
| 8 | + * with the License. You may obtain a copy of the License at |
| 9 | + * |
| 10 | + * http://www.apache.org/licenses/LICENSE-2.0 |
| 11 | + * |
| 12 | + * Unless required by applicable law or agreed to in writing, |
| 13 | + * software distributed under the License is distributed on an |
| 14 | + * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
| 15 | + * KIND, either express or implied. See the License for the |
| 16 | + * specific language governing permissions and limitations |
| 17 | + * under the License. |
| 18 | + */ |
| 19 | + |
| 20 | +#include "paimon/core/mergetree/compact/full_changelog_merge_function_wrapper.h" |
| 21 | + |
| 22 | +#include <memory> |
| 23 | +#include <utility> |
| 24 | + |
| 25 | +#include "gtest/gtest.h" |
| 26 | +#include "paimon/core/mergetree/compact/deduplicate_merge_function.h" |
| 27 | +#include "paimon/core/mergetree/compact/internal_row_equalizer.h" |
| 28 | +#include "paimon/memory/memory_pool.h" |
| 29 | +#include "paimon/testing/utils/binary_row_generator.h" |
| 30 | +#include "paimon/testing/utils/testharness.h" |
| 31 | + |
| 32 | +namespace paimon::test { |
| 33 | +namespace { |
| 34 | + |
| 35 | +constexpr int32_t MAX_LEVEL = 3; |
| 36 | + |
| 37 | +KeyValue MakeKeyValue(const RowKind* kind, int64_t sequence_number, int32_t level, int32_t key, |
| 38 | + int32_t value, const std::shared_ptr<MemoryPool>& pool) { |
| 39 | + return KeyValue(kind, sequence_number, level, |
| 40 | + BinaryRowGenerator::GenerateRowPtr({key}, pool.get()), |
| 41 | + BinaryRowGenerator::GenerateRowPtr({value}, pool.get())); |
| 42 | +} |
| 43 | + |
| 44 | +std::unique_ptr<RowCompactedSerializer> CreateValueSerializer( |
| 45 | + const std::shared_ptr<MemoryPool>& pool) { |
| 46 | + return RowCompactedSerializer::Create(arrow::schema({arrow::field("value", arrow::int32())}), |
| 47 | + pool) |
| 48 | + .value(); |
| 49 | +} |
| 50 | + |
| 51 | +std::unique_ptr<FullChangelogMergeFunctionWrapper> CreateWrapper( |
| 52 | + const std::shared_ptr<MemoryPool>& pool, |
| 53 | + FieldsComparator::FieldComparatorFunc value_equalizer = {}) { |
| 54 | + return std::make_unique<FullChangelogMergeFunctionWrapper>( |
| 55 | + std::make_unique<DeduplicateMergeFunction>(/*ignore_delete=*/false), MAX_LEVEL, |
| 56 | + CreateValueSerializer(pool), std::move(value_equalizer)); |
| 57 | +} |
| 58 | + |
| 59 | +void CheckKeyValue(const KeyValue& actual, const RowKind* kind, int64_t sequence_number, |
| 60 | + int32_t level, int32_t value) { |
| 61 | + ASSERT_EQ(kind, actual.value_kind); |
| 62 | + ASSERT_EQ(sequence_number, actual.sequence_number); |
| 63 | + ASSERT_EQ(level, actual.level); |
| 64 | + ASSERT_EQ(value, actual.value->GetInt(0)); |
| 65 | +} |
| 66 | + |
| 67 | +} // namespace |
| 68 | + |
| 69 | +TEST(FullChangelogMergeFunctionWrapperTest, TestSingleRecord) { |
| 70 | + auto pool = GetDefaultPool(); |
| 71 | + auto wrapper = CreateWrapper(pool); |
| 72 | + |
| 73 | + wrapper->Reset(); |
| 74 | + ASSERT_OK(wrapper->Add( |
| 75 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/1, /*level=*/0, 1, 10, pool))); |
| 76 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> insert_result, wrapper->GetResult()); |
| 77 | + ASSERT_TRUE(insert_result); |
| 78 | + ASSERT_TRUE(insert_result->result); |
| 79 | + ASSERT_EQ(1, insert_result->changelogs.size()); |
| 80 | + CheckKeyValue(insert_result->changelogs[0], RowKind::Insert(), 1, KeyValue::UNKNOWN_LEVEL, 10); |
| 81 | + CheckKeyValue(*insert_result->result, RowKind::Insert(), 1, 0, 10); |
| 82 | + |
| 83 | + wrapper->Reset(); |
| 84 | + ASSERT_OK(wrapper->Add( |
| 85 | + MakeKeyValue(RowKind::Delete(), /*sequence_number=*/2, /*level=*/0, 2, 20, pool))); |
| 86 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> delete_result, wrapper->GetResult()); |
| 87 | + ASSERT_TRUE(delete_result); |
| 88 | + ASSERT_FALSE(delete_result->result); |
| 89 | + ASSERT_TRUE(delete_result->changelogs.empty()); |
| 90 | + |
| 91 | + wrapper->Reset(); |
| 92 | + ASSERT_OK(wrapper->Add( |
| 93 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/3, MAX_LEVEL, 3, 30, pool))); |
| 94 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> top_level_result, wrapper->GetResult()); |
| 95 | + ASSERT_TRUE(top_level_result); |
| 96 | + ASSERT_TRUE(top_level_result->result); |
| 97 | + ASSERT_TRUE(top_level_result->changelogs.empty()); |
| 98 | + CheckKeyValue(*top_level_result->result, RowKind::Insert(), 3, MAX_LEVEL, 30); |
| 99 | +} |
| 100 | + |
| 101 | +TEST(FullChangelogMergeFunctionWrapperTest, TestInsertUpdateAndDelete) { |
| 102 | + auto pool = GetDefaultPool(); |
| 103 | + auto wrapper = CreateWrapper(pool); |
| 104 | + |
| 105 | + wrapper->Reset(); |
| 106 | + ASSERT_OK(wrapper->Add( |
| 107 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/1, MAX_LEVEL, 1, 10, pool))); |
| 108 | + ASSERT_OK(wrapper->Add( |
| 109 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/2, /*level=*/0, 1, 20, pool))); |
| 110 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> update_result, wrapper->GetResult()); |
| 111 | + ASSERT_TRUE(update_result); |
| 112 | + ASSERT_TRUE(update_result->result); |
| 113 | + ASSERT_EQ(2, update_result->changelogs.size()); |
| 114 | + CheckKeyValue(update_result->changelogs[0], RowKind::UpdateBefore(), 1, KeyValue::UNKNOWN_LEVEL, |
| 115 | + 10); |
| 116 | + CheckKeyValue(update_result->changelogs[1], RowKind::UpdateAfter(), 2, KeyValue::UNKNOWN_LEVEL, |
| 117 | + 20); |
| 118 | + CheckKeyValue(*update_result->result, RowKind::Insert(), 2, 0, 20); |
| 119 | + |
| 120 | + wrapper->Reset(); |
| 121 | + ASSERT_OK(wrapper->Add( |
| 122 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/3, MAX_LEVEL, 2, 30, pool))); |
| 123 | + ASSERT_OK(wrapper->Add( |
| 124 | + MakeKeyValue(RowKind::Delete(), /*sequence_number=*/4, /*level=*/0, 2, 30, pool))); |
| 125 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> delete_result, wrapper->GetResult()); |
| 126 | + ASSERT_TRUE(delete_result); |
| 127 | + ASSERT_FALSE(delete_result->result); |
| 128 | + ASSERT_EQ(1, delete_result->changelogs.size()); |
| 129 | + CheckKeyValue(delete_result->changelogs[0], RowKind::Delete(), 3, KeyValue::UNKNOWN_LEVEL, 30); |
| 130 | + |
| 131 | + wrapper->Reset(); |
| 132 | + ASSERT_OK(wrapper->Add( |
| 133 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/5, /*level=*/0, 3, 40, pool))); |
| 134 | + ASSERT_OK(wrapper->Add( |
| 135 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/6, /*level=*/0, 3, 50, pool))); |
| 136 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> insert_result, wrapper->GetResult()); |
| 137 | + ASSERT_TRUE(insert_result); |
| 138 | + ASSERT_TRUE(insert_result->result); |
| 139 | + ASSERT_EQ(1, insert_result->changelogs.size()); |
| 140 | + CheckKeyValue(insert_result->changelogs[0], RowKind::Insert(), 6, KeyValue::UNKNOWN_LEVEL, 50); |
| 141 | + CheckKeyValue(*insert_result->result, RowKind::Insert(), 6, 0, 50); |
| 142 | +} |
| 143 | + |
| 144 | +TEST(FullChangelogMergeFunctionWrapperTest, TestRowDeduplicate) { |
| 145 | + auto pool = GetDefaultPool(); |
| 146 | + |
| 147 | + auto wrapper_without_deduplicate = CreateWrapper(pool); |
| 148 | + wrapper_without_deduplicate->Reset(); |
| 149 | + ASSERT_OK(wrapper_without_deduplicate->Add( |
| 150 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/1, MAX_LEVEL, 1, 10, pool))); |
| 151 | + ASSERT_OK(wrapper_without_deduplicate->Add( |
| 152 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/2, /*level=*/0, 1, 10, pool))); |
| 153 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> result_without_deduplicate, |
| 154 | + wrapper_without_deduplicate->GetResult()); |
| 155 | + ASSERT_TRUE(result_without_deduplicate); |
| 156 | + ASSERT_EQ(2, result_without_deduplicate->changelogs.size()); |
| 157 | + |
| 158 | + auto value_schema = arrow::schema({arrow::field("value", arrow::int32())}); |
| 159 | + ASSERT_OK_AND_ASSIGN(FieldsComparator::FieldComparatorFunc value_equalizer, |
| 160 | + InternalRowEqualizer::Create(value_schema, /*ignore_fields=*/{})); |
| 161 | + auto wrapper = CreateWrapper(pool, std::move(value_equalizer)); |
| 162 | + |
| 163 | + wrapper->Reset(); |
| 164 | + ASSERT_OK(wrapper->Add( |
| 165 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/1, MAX_LEVEL, 1, 10, pool))); |
| 166 | + ASSERT_OK(wrapper->Add( |
| 167 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/2, /*level=*/0, 1, 10, pool))); |
| 168 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> result, wrapper->GetResult()); |
| 169 | + ASSERT_TRUE(result); |
| 170 | + ASSERT_TRUE(result->result); |
| 171 | + ASSERT_TRUE(result->changelogs.empty()); |
| 172 | + CheckKeyValue(*result->result, RowKind::Insert(), 2, 0, 10); |
| 173 | +} |
| 174 | + |
| 175 | +TEST(FullChangelogMergeFunctionWrapperTest, TestRowDeduplicateWithIgnoreFields) { |
| 176 | + auto pool = GetDefaultPool(); |
| 177 | + auto value_schema = arrow::schema( |
| 178 | + {arrow::field("value", arrow::int32()), arrow::field("ignored", arrow::int32())}); |
| 179 | + ASSERT_OK_AND_ASSIGN(FieldsComparator::FieldComparatorFunc value_equalizer, |
| 180 | + InternalRowEqualizer::Create(value_schema, {"ignored"})); |
| 181 | + ASSERT_OK_AND_ASSIGN(std::unique_ptr<RowCompactedSerializer> value_serializer, |
| 182 | + RowCompactedSerializer::Create(value_schema, pool)); |
| 183 | + FullChangelogMergeFunctionWrapper wrapper( |
| 184 | + std::make_unique<DeduplicateMergeFunction>(/*ignore_delete=*/false), MAX_LEVEL, |
| 185 | + std::move(value_serializer), std::move(value_equalizer)); |
| 186 | + |
| 187 | + wrapper.Reset(); |
| 188 | + ASSERT_OK(wrapper.Add(KeyValue(RowKind::Insert(), /*sequence_number=*/1, MAX_LEVEL, |
| 189 | + BinaryRowGenerator::GenerateRowPtr({1}, pool.get()), |
| 190 | + BinaryRowGenerator::GenerateRowPtr({10, 1}, pool.get())))); |
| 191 | + ASSERT_OK(wrapper.Add(KeyValue(RowKind::Insert(), /*sequence_number=*/2, /*level=*/0, |
| 192 | + BinaryRowGenerator::GenerateRowPtr({1}, pool.get()), |
| 193 | + BinaryRowGenerator::GenerateRowPtr({10, 2}, pool.get())))); |
| 194 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> ignored_field_result, wrapper.GetResult()); |
| 195 | + ASSERT_TRUE(ignored_field_result); |
| 196 | + ASSERT_TRUE(ignored_field_result->result); |
| 197 | + ASSERT_TRUE(ignored_field_result->changelogs.empty()); |
| 198 | + ASSERT_EQ(10, ignored_field_result->result->value->GetInt(0)); |
| 199 | + ASSERT_EQ(2, ignored_field_result->result->value->GetInt(1)); |
| 200 | + |
| 201 | + wrapper.Reset(); |
| 202 | + ASSERT_OK(wrapper.Add(KeyValue(RowKind::Insert(), /*sequence_number=*/3, MAX_LEVEL, |
| 203 | + BinaryRowGenerator::GenerateRowPtr({1}, pool.get()), |
| 204 | + BinaryRowGenerator::GenerateRowPtr({10, 1}, pool.get())))); |
| 205 | + ASSERT_OK(wrapper.Add(KeyValue(RowKind::Insert(), /*sequence_number=*/4, /*level=*/0, |
| 206 | + BinaryRowGenerator::GenerateRowPtr({1}, pool.get()), |
| 207 | + BinaryRowGenerator::GenerateRowPtr({11, 2}, pool.get())))); |
| 208 | + ASSERT_OK_AND_ASSIGN(std::optional<ChangelogResult> value_field_result, wrapper.GetResult()); |
| 209 | + ASSERT_TRUE(value_field_result); |
| 210 | + ASSERT_TRUE(value_field_result->result); |
| 211 | + ASSERT_EQ(2, value_field_result->changelogs.size()); |
| 212 | + ASSERT_EQ(RowKind::UpdateBefore(), value_field_result->changelogs[0].value_kind); |
| 213 | + ASSERT_EQ(RowKind::UpdateAfter(), value_field_result->changelogs[1].value_kind); |
| 214 | +} |
| 215 | + |
| 216 | +TEST(FullChangelogMergeFunctionWrapperTest, TestRejectMultipleTopLevelRecords) { |
| 217 | + auto pool = GetDefaultPool(); |
| 218 | + auto wrapper = CreateWrapper(pool); |
| 219 | + |
| 220 | + wrapper->Reset(); |
| 221 | + ASSERT_OK(wrapper->Add( |
| 222 | + MakeKeyValue(RowKind::Insert(), /*sequence_number=*/1, MAX_LEVEL, 1, 10, pool))); |
| 223 | + ASSERT_NOK_WITH_MSG(wrapper->Add(MakeKeyValue(RowKind::Insert(), /*sequence_number=*/2, |
| 224 | + MAX_LEVEL, 1, 20, pool)), |
| 225 | + "Top level key-value already exists"); |
| 226 | +} |
| 227 | + |
| 228 | +} // namespace paimon::test |
0 commit comments