be/src/exprs/function/array/function_array_zip.cpp
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 | | // This file is copied from |
18 | | // https://github.com/ClickHouse/ClickHouse/blob/master/src/Functions/array/arrayZip.cpp |
19 | | // and modified by Doris |
20 | | |
21 | | #include <fmt/format.h> |
22 | | #include <glog/logging.h> |
23 | | #include <stddef.h> |
24 | | |
25 | | #include <algorithm> |
26 | | #include <memory> |
27 | | #include <ostream> |
28 | | #include <string> |
29 | | #include <utility> |
30 | | |
31 | | #include "common/status.h" |
32 | | #include "core/block/block.h" |
33 | | #include "core/block/column_numbers.h" |
34 | | #include "core/block/column_with_type_and_name.h" |
35 | | #include "core/column/column.h" |
36 | | #include "core/column/column_array.h" |
37 | | #include "core/column/column_nullable.h" |
38 | | #include "core/column/column_struct.h" |
39 | | #include "core/column/column_vector.h" |
40 | | #include "core/data_type/data_type.h" |
41 | | #include "core/data_type/data_type_array.h" |
42 | | #include "core/data_type/data_type_nullable.h" |
43 | | #include "core/data_type/data_type_struct.h" |
44 | | #include "core/types.h" |
45 | | #include "exprs/aggregate/aggregate_function.h" |
46 | | #include "exprs/function/function.h" |
47 | | #include "exprs/function/function_helpers.h" |
48 | | #include "exprs/function/simple_function_factory.h" |
49 | | |
50 | | namespace doris { |
51 | | class FunctionContext; |
52 | | } // namespace doris |
53 | | |
54 | | namespace doris { |
55 | | |
56 | | // Combines multiple arrays into a single array |
57 | | // array_zip(['d', 'o', 'r', 'i', 's'], [1, 2, 3, 4, 5]) -> [('d', 1), ('o', 2), ('r', 3), ('i', 4), ('s', 5)] |
58 | | class FunctionArrayZip : public IFunction { |
59 | | public: |
60 | | static constexpr auto name = "array_zip"; |
61 | 2 | static FunctionPtr create() { return std::make_shared<FunctionArrayZip>(); } |
62 | | |
63 | | /// Get function name. |
64 | 0 | String get_name() const override { return name; } |
65 | | |
66 | 1 | bool is_variadic() const override { return true; } |
67 | | |
68 | 0 | size_t get_number_of_arguments() const override { return 0; } |
69 | | |
70 | 0 | DataTypePtr get_return_type_impl(const DataTypes& arguments) const override { |
71 | 0 | DCHECK(arguments.size() > 0) |
72 | 0 | << "function: " << get_name() << ", arguments should not be empty"; |
73 | |
|
74 | 0 | DataTypes res_data_types; |
75 | 0 | size_t num_elements = arguments.size(); |
76 | 0 | for (size_t i = 0; i < num_elements; ++i) { |
77 | 0 | DCHECK(arguments[i]->get_primitive_type() == TYPE_ARRAY) |
78 | 0 | << i << "-th element is not array type"; |
79 | |
|
80 | 0 | const auto* array_type = check_and_get_data_type<DataTypeArray>(arguments[i].get()); |
81 | 0 | DCHECK(array_type) << "function: " << get_name() << " " << i + 1 |
82 | 0 | << "-th argument is not array"; |
83 | |
|
84 | 0 | res_data_types.emplace_back( |
85 | 0 | make_nullable(remove_nullable((array_type->get_nested_type())))); |
86 | 0 | } |
87 | |
|
88 | 0 | auto res = std::make_shared<DataTypeArray>( |
89 | 0 | make_nullable(std::make_shared<DataTypeStruct>(res_data_types))); |
90 | 0 | return res; |
91 | 0 | } |
92 | | |
93 | | Status execute_impl(FunctionContext* context, Block& block, const ColumnNumbers& arguments, |
94 | 0 | uint32_t result, size_t input_rows_count) const override { |
95 | 0 | size_t num_element = arguments.size(); |
96 | | |
97 | | // all the columns must have the same size as the first column |
98 | 0 | ColumnPtr first_array_column; |
99 | 0 | Columns tuple_columns(num_element); |
100 | |
|
101 | 0 | for (size_t i = 0; i < num_element; ++i) { |
102 | 0 | auto col = block.get_by_position(arguments[i]).column; |
103 | 0 | col = col->convert_to_full_column_if_const(); |
104 | |
|
105 | 0 | const auto* column_array = check_and_get_column<ColumnArray>(col.get()); |
106 | 0 | if (!column_array) { |
107 | 0 | return Status::RuntimeError(fmt::format( |
108 | 0 | "execute failed, function {}'s {}-th argument should be array bet get {}", |
109 | 0 | get_name(), i + 1, block.get_by_position(arguments[i]).type->get_name())); |
110 | 0 | } |
111 | | |
112 | 0 | if (i == 0) { |
113 | 0 | first_array_column = col; |
114 | 0 | } else if (!column_array->has_equal_offsets( |
115 | 0 | static_cast<const ColumnArray&>(*first_array_column))) { |
116 | 0 | return Status::RuntimeError( |
117 | 0 | fmt::format("execute failed, function {}'s {}-th argument should have same " |
118 | 0 | "offsets with first argument", |
119 | 0 | get_name(), i + 1)); |
120 | 0 | } |
121 | | |
122 | 0 | tuple_columns[i] = column_array->get_data_ptr(); |
123 | 0 | } |
124 | | |
125 | 0 | auto tuples = ColumnStruct::create(tuple_columns); |
126 | 0 | auto nullable_tuples = |
127 | 0 | ColumnNullable::create(std::move(tuples), ColumnUInt8::create(tuples->size(), 0)); |
128 | 0 | auto res_column = ColumnArray::create( |
129 | 0 | std::move(nullable_tuples), |
130 | 0 | static_cast<const ColumnArray&>(*first_array_column).get_offsets_ptr()); |
131 | 0 | block.replace_by_position(result, std::move(res_column)); |
132 | 0 | return Status::OK(); |
133 | 0 | } |
134 | | }; |
135 | | |
136 | 1 | void register_function_array_zip(SimpleFunctionFactory& factory) { |
137 | 1 | factory.register_function<FunctionArrayZip>(); |
138 | 1 | } |
139 | | |
140 | | } // namespace doris |