be/src/exprs/function/cast/cast_to_map.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 | | #include <array> |
19 | | |
20 | | #include "core/column/column_map.h" |
21 | | #include "core/column/column_nullable.h" |
22 | | #include "core/data_type/data_type_map.h" |
23 | | #include "exprs/function/cast/cast_base.h" |
24 | | |
25 | | namespace doris::CastWrapper { |
26 | | |
27 | 3 | inline Status deduplicate_map_keys_in_result(Block& block, uint32_t result) { |
28 | 3 | auto result_column_name = block.get_by_position(result).column->get_name(); |
29 | 3 | auto mutable_result_column = IColumn::mutate(std::move(block.get_by_position(result).column)); |
30 | | |
31 | 3 | if (auto* nullable_column = check_and_get_column<ColumnNullable>(*mutable_result_column)) { |
32 | 3 | auto nested_column = IColumn::mutate(nullable_column->get_nested_column_ptr()); |
33 | 3 | auto* map_column = check_and_get_column<ColumnMap>(*nested_column); |
34 | 3 | if (!map_column) { |
35 | 0 | return Status::RuntimeError("Illegal column {} for function CAST AS MAP", |
36 | 0 | result_column_name); |
37 | 0 | } |
38 | | |
39 | 3 | RETURN_IF_ERROR(map_column->deduplicate_keys(true)); |
40 | 3 | ColumnPtr nested_column_ptr = std::move(nested_column); |
41 | 3 | nullable_column->change_nested_column(nested_column_ptr); |
42 | 3 | } else { |
43 | 0 | auto* map_column = check_and_get_column<ColumnMap>(*mutable_result_column); |
44 | 0 | if (!map_column) { |
45 | 0 | return Status::RuntimeError("Illegal column {} for function CAST AS MAP", |
46 | 0 | result_column_name); |
47 | 0 | } |
48 | | |
49 | 0 | RETURN_IF_ERROR(map_column->deduplicate_keys(true)); |
50 | 0 | } |
51 | | |
52 | 3 | block.get_by_position(result).column = std::move(mutable_result_column); |
53 | 3 | return Status::OK(); |
54 | 3 | } |
55 | | |
56 | 3 | inline WrapperType wrap_string_to_map_wrapper(WrapperType wrapper) { |
57 | 3 | return [wrapper = std::move(wrapper)](FunctionContext* context, Block& block, |
58 | 3 | const ColumnNumbers& arguments, uint32_t result, |
59 | 3 | size_t input_rows_count, |
60 | 3 | const NullMap::value_type* null_map = nullptr) { |
61 | 3 | RETURN_IF_ERROR(wrapper(context, block, arguments, result, input_rows_count, null_map)); |
62 | 3 | return deduplicate_map_keys_in_result(block, result); |
63 | 3 | }; |
64 | 3 | } |
65 | | |
66 | | //TODO(Amory) . Need support more cast for key , value for map |
67 | | WrapperType create_map_wrapper(FunctionContext* context, const DataTypePtr& from_type, |
68 | 19 | const DataTypeMap& to_type) { |
69 | 19 | if (is_string_type(from_type->get_primitive_type())) { |
70 | 3 | auto wrapper = context->enable_strict_mode() ? cast_from_string_to_complex_type_strict_mode |
71 | 3 | : cast_from_string_to_complex_type; |
72 | 3 | return wrap_string_to_map_wrapper(wrapper); |
73 | 3 | } |
74 | 16 | const auto* from = check_and_get_data_type<DataTypeMap>(from_type.get()); |
75 | 16 | if (!from) { |
76 | 0 | return CastWrapper::create_unsupport_wrapper( |
77 | 0 | fmt::format("CAST AS Map can only be performed between Map types or from " |
78 | 0 | "String. from type: {}, to type: {}", |
79 | 0 | from_type->get_name(), to_type.get_name())); |
80 | 0 | } |
81 | 16 | DataTypes from_kv_types; |
82 | 16 | DataTypes to_kv_types; |
83 | 16 | from_kv_types.reserve(2); |
84 | 16 | to_kv_types.reserve(2); |
85 | 16 | from_kv_types.push_back(from->get_key_type()); |
86 | 16 | from_kv_types.push_back(from->get_value_type()); |
87 | 16 | to_kv_types.push_back(to_type.get_key_type()); |
88 | 16 | to_kv_types.push_back(to_type.get_value_type()); |
89 | | |
90 | 16 | auto kv_wrappers = get_element_wrappers(context, from_kv_types, to_kv_types); |
91 | | /// A key or value whose type does not change is passed through, so it needs neither a child mask |
92 | | /// nor a call into its (identity) wrapper. |
93 | 16 | std::array<bool, 2> unchanged_kv = {from_kv_types[0]->equals(*to_kv_types[0]), |
94 | 16 | from_kv_types[1]->equals(*to_kv_types[1])}; |
95 | 16 | return [kv_wrappers, from_kv_types, to_kv_types, unchanged_kv]( |
96 | 16 | FunctionContext* context, Block& block, const ColumnNumbers& arguments, |
97 | 16 | uint32_t result, size_t /*input_rows_count*/, |
98 | 16 | const NullMap::value_type* null_map = nullptr) -> Status { |
99 | 16 | auto& from_column = block.get_by_position(arguments.front()).column; |
100 | 16 | const auto* from_col_map = check_and_get_column<ColumnMap>(from_column.get()); |
101 | 16 | if (!from_col_map) { |
102 | 0 | return Status::RuntimeError("Illegal column {} for function CAST AS MAP", |
103 | 0 | from_column->get_name()); |
104 | 0 | } |
105 | | |
106 | 16 | Columns converted_columns(2); |
107 | 16 | ColumnsWithTypeAndName columnsWithTypeAndName(2); |
108 | 16 | columnsWithTypeAndName[0] = {from_col_map->get_keys_ptr(), from_kv_types[0], ""}; |
109 | 16 | columnsWithTypeAndName[1] = {from_col_map->get_values_ptr(), from_kv_types[1], ""}; |
110 | | |
111 | 16 | const auto& offsets = from_col_map->get_offsets(); |
112 | | // Keys and values share the rows of their parent, so the parent mask is scanned once. |
113 | 16 | const NullMap::value_type* inherited_null_map = |
114 | 16 | (null_map != nullptr && has_masked_row(null_map, offsets.size())) ? null_map |
115 | 16 | : nullptr; |
116 | 42 | for (size_t i = 0; i < 2; ++i) { |
117 | 31 | if (unchanged_kv[i]) { |
118 | 14 | converted_columns[i] = columnsWithTypeAndName[i].column; |
119 | 14 | continue; |
120 | 14 | } |
121 | | /// Keys and values are flattened, so the NULL of a map row has to be inherited by the |
122 | | /// entries that belong to it before they are cast. |
123 | 17 | auto child_mask = build_child_null_mask(inherited_null_map, &offsets, |
124 | 17 | columnsWithTypeAndName[i].column); |
125 | 17 | ColumnWithTypeAndName child_column {child_mask.column, from_kv_types[i], ""}; |
126 | 17 | ColumnNumbers element_arguments {block.columns()}; |
127 | 17 | block.insert(child_column); |
128 | 17 | auto element_result = block.columns(); |
129 | 17 | block.insert({to_kv_types[i], ""}); |
130 | 17 | RETURN_IF_ERROR(kv_wrappers[i](context, block, element_arguments, element_result, |
131 | 17 | child_mask.column->size(), child_mask.null_map)); |
132 | 12 | converted_columns[i] = block.get_by_position(element_result).column; |
133 | 12 | } |
134 | | |
135 | 11 | auto map_column = ColumnMap::create(converted_columns[0], converted_columns[1], |
136 | 11 | from_col_map->get_offsets_ptr()); |
137 | 11 | static_cast<void>(map_column->deduplicate_keys()); |
138 | 11 | block.get_by_position(result).column = std::move(map_column); |
139 | 11 | return Status::OK(); |
140 | 16 | }; |
141 | 16 | } |
142 | | } // namespace doris::CastWrapper |