be/src/exprs/aggregate/aggregate_function_retention.h
Line | Count | Source |
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 | | // This file is copied from |
19 | | // https://github.com/ClickHouse/ClickHouse/blob/master/AggregateFunctionRetention.h |
20 | | // and modified by Doris |
21 | | |
22 | | #pragma once |
23 | | |
24 | | #include <stddef.h> |
25 | | #include <stdint.h> |
26 | | |
27 | | #include <boost/iterator/iterator_facade.hpp> |
28 | | #include <memory> |
29 | | |
30 | | #include "common/exception.h" |
31 | | #include "common/status.h" |
32 | | #include "core/assert_cast.h" |
33 | | #include "core/column/column.h" |
34 | | #include "core/column/column_array.h" |
35 | | #include "core/column/column_nullable.h" |
36 | | #include "core/column/column_vector.h" |
37 | | #include "core/data_type/data_type_array.h" |
38 | | #include "core/data_type/data_type_nullable.h" |
39 | | #include "core/data_type/data_type_number.h" |
40 | | #include "core/types.h" |
41 | | #include "exprs/aggregate/aggregate_function.h" |
42 | | #include "util/var_int.h" |
43 | | |
44 | | namespace doris { |
45 | | class Arena; |
46 | | class BufferReadable; |
47 | | class BufferWritable; |
48 | | } // namespace doris |
49 | | |
50 | | namespace doris { |
51 | | struct RetentionState { |
52 | | static constexpr size_t MAX_EVENTS = 32; |
53 | | uint8_t events[MAX_EVENTS] = {0}; |
54 | | |
55 | 353 | RetentionState() = default; |
56 | | |
57 | 40 | void reset() { |
58 | 1.32k | for (int64_t i = 0; i < MAX_EVENTS; i++) { |
59 | 1.28k | events[i] = 0; |
60 | 1.28k | } |
61 | 40 | } |
62 | | |
63 | 83 | void set(int event) { events[event] = 1; } |
64 | | |
65 | 118 | void merge(const RetentionState& other) { |
66 | 3.89k | for (int64_t i = 0; i < MAX_EVENTS; i++) { |
67 | 3.77k | events[i] |= other.events[i]; |
68 | 3.77k | } |
69 | 118 | } |
70 | | |
71 | 156 | void write(BufferWritable& out) const { |
72 | 156 | int64_t serialized_events = 0; |
73 | 5.14k | for (int64_t i = 0; i < MAX_EVENTS; i++) { |
74 | 4.99k | serialized_events |= events[i]; |
75 | 4.99k | serialized_events <<= 1; |
76 | 4.99k | } |
77 | 156 | write_var_int(serialized_events, out); |
78 | 156 | } |
79 | | |
80 | 117 | void read(BufferReadable& in) { |
81 | 117 | int64_t serialized_events = 0; |
82 | 117 | uint64_t u_serialized_events = 0; |
83 | 117 | read_var_int(serialized_events, in); |
84 | 117 | u_serialized_events = serialized_events; |
85 | | |
86 | 117 | u_serialized_events >>= 1; |
87 | 3.86k | for (int64_t i = MAX_EVENTS - 1; i >= 0; i--) { |
88 | 3.74k | events[i] = (uint8_t)(1 & u_serialized_events); |
89 | 3.74k | u_serialized_events >>= 1; |
90 | 3.74k | } |
91 | 117 | } |
92 | | |
93 | 70 | void insert_result_into(IColumn& to, size_t events_size, const uint8_t* arg_events) const { |
94 | 70 | auto& data_to = assert_cast<ColumnUInt8&>(to).get_data(); |
95 | | |
96 | 70 | ColumnArray::Offset64 current_offset = data_to.size(); |
97 | 70 | data_to.resize(current_offset + events_size); |
98 | | |
99 | 70 | bool first_flag = arg_events[0]; |
100 | 70 | data_to[current_offset] = first_flag; |
101 | 70 | ++current_offset; |
102 | | |
103 | 219 | for (size_t i = 1; i < events_size; ++i) { |
104 | 149 | data_to[current_offset] = (first_flag && arg_events[i]); |
105 | 149 | ++current_offset; |
106 | 149 | } |
107 | 70 | } |
108 | | }; |
109 | | |
110 | | class AggregateFunctionRetention final |
111 | | : public IAggregateFunctionDataHelper<RetentionState, AggregateFunctionRetention>, |
112 | | VarargsExpression, |
113 | | NullableAggregateFunction { |
114 | | public: |
115 | | AggregateFunctionRetention(const DataTypes& argument_types_) |
116 | 442 | : IAggregateFunctionDataHelper<RetentionState, AggregateFunctionRetention>( |
117 | 442 | argument_types_) { |
118 | | // RetentionState only has room for MAX_EVENTS(32) events (fixed-size events[] array, |
119 | | // plus an int64 serialized bitmap). More params would overflow events[] in add()/ |
120 | | // insert_result_into() and corrupt the heap, so reject it at construction time. |
121 | 442 | if (argument_types_.size() > RetentionState::MAX_EVENTS) { |
122 | 1 | throw Exception(ErrorCode::INVALID_ARGUMENT, |
123 | 1 | "retention function can accept at most {} params, but got {}", |
124 | 1 | RetentionState::MAX_EVENTS, argument_types_.size()); |
125 | 1 | } |
126 | 442 | } |
127 | | |
128 | 31 | String get_name() const override { return "retention"; } |
129 | | |
130 | 61 | DataTypePtr get_return_type() const override { |
131 | 61 | return std::make_shared<DataTypeArray>(make_nullable(std::make_shared<DataTypeUInt8>())); |
132 | 61 | } |
133 | | |
134 | 40 | void reset(AggregateDataPtr __restrict place) const override { this->data(place).reset(); } |
135 | | void add(AggregateDataPtr __restrict place, const IColumn** columns, const ssize_t row_num, |
136 | 231 | Arena&) const override { |
137 | 921 | for (int i = 0; i < get_argument_types().size(); i++) { |
138 | 690 | auto event = assert_cast<const ColumnUInt8*, TypeCheckOnRelease::DISABLE>(columns[i]) |
139 | 690 | ->get_data()[row_num]; |
140 | 690 | if (event) { |
141 | 83 | this->data(place).set(i); |
142 | 83 | } |
143 | 690 | } |
144 | 231 | } |
145 | | |
146 | | void merge(AggregateDataPtr __restrict place, ConstAggregateDataPtr rhs, |
147 | 118 | Arena&) const override { |
148 | 118 | this->data(place).merge(this->data(rhs)); |
149 | 118 | } |
150 | | |
151 | 156 | void serialize(ConstAggregateDataPtr __restrict place, BufferWritable& buf) const override { |
152 | 156 | this->data(place).write(buf); |
153 | 156 | } |
154 | | |
155 | | void deserialize(AggregateDataPtr __restrict place, BufferReadable& buf, |
156 | 117 | Arena&) const override { |
157 | 117 | this->data(place).read(buf); |
158 | 117 | } |
159 | | |
160 | 70 | void insert_result_into(ConstAggregateDataPtr __restrict place, IColumn& to) const override { |
161 | 70 | auto& to_arr = assert_cast<ColumnArray&>(to); |
162 | 70 | auto& to_nested_col = to_arr.get_data(); |
163 | 70 | if (is_column_nullable(to_nested_col)) { |
164 | 70 | auto col_null = reinterpret_cast<ColumnNullable*>(&to_nested_col); |
165 | 70 | this->data(place).insert_result_into(col_null->get_nested_column(), |
166 | 70 | get_argument_types().size(), |
167 | 70 | this->data(place).events); |
168 | 70 | col_null->get_null_map_data().resize_fill(col_null->get_nested_column().size(), 0); |
169 | 70 | } else { |
170 | 0 | this->data(place).insert_result_into(to_nested_col, get_argument_types().size(), |
171 | 0 | this->data(place).events); |
172 | 0 | } |
173 | 70 | to_arr.get_offsets().push_back(to_nested_col.size()); |
174 | 70 | } |
175 | | }; |
176 | | } // namespace doris |