ceph/src/arrow/cpp/src/arrow/array/builder_union.h

   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 #pragma once
  19
  20 #include <cstdint>
  21 #include <memory>
  22 #include <string>
  23 #include <vector>
  24
  25 #include "arrow/array/array_nested.h"
  26 #include "arrow/array/builder_base.h"
  27 #include "arrow/array/data.h"
  28 #include "arrow/buffer_builder.h"
  29 #include "arrow/memory_pool.h"
  30 #include "arrow/status.h"
  31 #include "arrow/type.h"
  32 #include "arrow/util/visibility.h"
  33
  34 namespace arrow {
  35
  36 /// \brief Base class for union array builds.
  37 ///
  38 /// Note that while we subclass ArrayBuilder, as union types do not have a
  39 /// validity bitmap, the bitmap builder member of ArrayBuilder is not used.
  40 class ARROW_EXPORT BasicUnionBuilder : public ArrayBuilder {
  41  public:
  42   Status FinishInternal(std::shared_ptr<ArrayData>* out) override;
  43
  44   /// \cond FALSE
  45   using ArrayBuilder::Finish;
  46   /// \endcond
  47
  48   Status Finish(std::shared_ptr<UnionArray>* out) { return FinishTyped(out); }
  49
  50   /// \brief Make a new child builder available to the UnionArray
  51   ///
  52   /// \param[in] new_child the child builder
  53   /// \param[in] field_name the name of the field in the union array type
  54   /// if type inference is used
  55   /// \return child index, which is the "type" argument that needs
  56   /// to be passed to the "Append" method to add a new element to
  57   /// the union array.
  58   int8_t AppendChild(const std::shared_ptr<ArrayBuilder>& new_child,
  59                      const std::string& field_name = "");
  60
  61   std::shared_ptr<DataType> type() const override;
  62
  63   int64_t length() const override { return types_builder_.length(); }
  64
  65  protected:
  66   BasicUnionBuilder(MemoryPool* pool,
  67                     const std::vector<std::shared_ptr<ArrayBuilder>>& children,
  68                     const std::shared_ptr<DataType>& type);
  69
  70   int8_t NextTypeId();
  71
  72   std::vector<std::shared_ptr<Field>> child_fields_;
  73   std::vector<int8_t> type_codes_;
  74   UnionMode::type mode_;
  75
  76   std::vector<ArrayBuilder*> type_id_to_children_;
  77   std::vector<int> type_id_to_child_id_;
  78   // for all type_id < dense_type_id_, type_id_to_children_[type_id] != nullptr
  79   int8_t dense_type_id_ = 0;
  80   TypedBufferBuilder<int8_t> types_builder_;
  81 };
  82
  83 /// \class DenseUnionBuilder
  84 ///
  85 /// This API is EXPERIMENTAL.
  86 class ARROW_EXPORT DenseUnionBuilder : public BasicUnionBuilder {
  87  public:
  88   /// Use this constructor to initialize the UnionBuilder with no child builders,
  89   /// allowing type to be inferred. You will need to call AppendChild for each of the
  90   /// children builders you want to use.
  91   explicit DenseUnionBuilder(MemoryPool* pool)
  92       : BasicUnionBuilder(pool, {}, dense_union(FieldVector{})), offsets_builder_(pool) {}
  93
  94   /// Use this constructor to specify the type explicitly.
  95   /// You can still add child builders to the union after using this constructor
  96   DenseUnionBuilder(MemoryPool* pool,
  97                     const std::vector<std::shared_ptr<ArrayBuilder>>& children,
  98                     const std::shared_ptr<DataType>& type)
  99       : BasicUnionBuilder(pool, children, type), offsets_builder_(pool) {}
 100
 101   Status AppendNull() final {
 102     const int8_t first_child_code = type_codes_[0];
 103     ArrayBuilder* child_builder = type_id_to_children_[first_child_code];
 104     ARROW_RETURN_NOT_OK(types_builder_.Append(first_child_code));
 105     ARROW_RETURN_NOT_OK(
 106         offsets_builder_.Append(static_cast<int32_t>(child_builder->length())));
 107     // Append a null arbitrarily to the first child
 108     return child_builder->AppendNull();
 109   }
 110
 111   Status AppendNulls(int64_t length) final {
 112     const int8_t first_child_code = type_codes_[0];
 113     ArrayBuilder* child_builder = type_id_to_children_[first_child_code];
 114     ARROW_RETURN_NOT_OK(types_builder_.Append(length, first_child_code));
 115     ARROW_RETURN_NOT_OK(
 116         offsets_builder_.Append(length, static_cast<int32_t>(child_builder->length())));
 117     // Append just a single null to the first child
 118     return child_builder->AppendNull();
 119   }
 120
 121   Status AppendEmptyValue() final {
 122     const int8_t first_child_code = type_codes_[0];
 123     ArrayBuilder* child_builder = type_id_to_children_[first_child_code];
 124     ARROW_RETURN_NOT_OK(types_builder_.Append(first_child_code));
 125     ARROW_RETURN_NOT_OK(
 126         offsets_builder_.Append(static_cast<int32_t>(child_builder->length())));
 127     // Append an empty value arbitrarily to the first child
 128     return child_builder->AppendEmptyValue();
 129   }
 130
 131   Status AppendEmptyValues(int64_t length) final {
 132     const int8_t first_child_code = type_codes_[0];
 133     ArrayBuilder* child_builder = type_id_to_children_[first_child_code];
 134     ARROW_RETURN_NOT_OK(types_builder_.Append(length, first_child_code));
 135     ARROW_RETURN_NOT_OK(
 136         offsets_builder_.Append(length, static_cast<int32_t>(child_builder->length())));
 137     // Append just a single empty value to the first child
 138     return child_builder->AppendEmptyValue();
 139   }
 140
 141   /// \brief Append an element to the UnionArray. This must be followed
 142   ///        by an append to the appropriate child builder.
 143   ///
 144   /// \param[in] next_type type_id of the child to which the next value will be appended.
 145   ///
 146   /// The corresponding child builder must be appended to independently after this method
 147   /// is called.
 148   Status Append(int8_t next_type) {
 149     ARROW_RETURN_NOT_OK(types_builder_.Append(next_type));
 150     if (type_id_to_children_[next_type]->length() == kListMaximumElements) {
 151       return Status::CapacityError(
 152           "a dense UnionArray cannot contain more than 2^31 - 1 elements from a single "
 153           "child");
 154     }
 155     auto offset = static_cast<int32_t>(type_id_to_children_[next_type]->length());
 156     return offsets_builder_.Append(offset);
 157   }
 158
 159   Status AppendArraySlice(const ArrayData& array, int64_t offset,
 160                           int64_t length) override;
 161
 162   Status FinishInternal(std::shared_ptr<ArrayData>* out) override;
 163
 164  private:
 165   TypedBufferBuilder<int32_t> offsets_builder_;
 166 };
 167
 168 /// \class SparseUnionBuilder
 169 ///
 170 /// This API is EXPERIMENTAL.
 171 class ARROW_EXPORT SparseUnionBuilder : public BasicUnionBuilder {
 172  public:
 173   /// Use this constructor to initialize the UnionBuilder with no child builders,
 174   /// allowing type to be inferred. You will need to call AppendChild for each of the
 175   /// children builders you want to use.
 176   explicit SparseUnionBuilder(MemoryPool* pool)
 177       : BasicUnionBuilder(pool, {}, sparse_union(FieldVector{})) {}
 178
 179   /// Use this constructor to specify the type explicitly.
 180   /// You can still add child builders to the union after using this constructor
 181   SparseUnionBuilder(MemoryPool* pool,
 182                      const std::vector<std::shared_ptr<ArrayBuilder>>& children,
 183                      const std::shared_ptr<DataType>& type)
 184       : BasicUnionBuilder(pool, children, type) {}
 185
 186   /// \brief Append a null value.
 187   ///
 188   /// A null is appended to the first child, empty values to the other children.
 189   Status AppendNull() final {
 190     const auto first_child_code = type_codes_[0];
 191     ARROW_RETURN_NOT_OK(types_builder_.Append(first_child_code));
 192     ARROW_RETURN_NOT_OK(type_id_to_children_[first_child_code]->AppendNull());
 193     for (int i = 1; i < static_cast<int>(type_codes_.size()); ++i) {
 194       ARROW_RETURN_NOT_OK(type_id_to_children_[type_codes_[i]]->AppendEmptyValue());
 195     }
 196     return Status::OK();
 197   }
 198
 199   /// \brief Append multiple null values.
 200   ///
 201   /// Nulls are appended to the first child, empty values to the other children.
 202   Status AppendNulls(int64_t length) final {
 203     const auto first_child_code = type_codes_[0];
 204     ARROW_RETURN_NOT_OK(types_builder_.Append(length, first_child_code));
 205     ARROW_RETURN_NOT_OK(type_id_to_children_[first_child_code]->AppendNulls(length));
 206     for (int i = 1; i < static_cast<int>(type_codes_.size()); ++i) {
 207       ARROW_RETURN_NOT_OK(
 208           type_id_to_children_[type_codes_[i]]->AppendEmptyValues(length));
 209     }
 210     return Status::OK();
 211   }
 212
 213   Status AppendEmptyValue() final {
 214     ARROW_RETURN_NOT_OK(types_builder_.Append(type_codes_[0]));
 215     for (int8_t code : type_codes_) {
 216       ARROW_RETURN_NOT_OK(type_id_to_children_[code]->AppendEmptyValue());
 217     }
 218     return Status::OK();
 219   }
 220
 221   Status AppendEmptyValues(int64_t length) final {
 222     ARROW_RETURN_NOT_OK(types_builder_.Append(length, type_codes_[0]));
 223     for (int8_t code : type_codes_) {
 224       ARROW_RETURN_NOT_OK(type_id_to_children_[code]->AppendEmptyValues(length));
 225     }
 226     return Status::OK();
 227   }
 228
 229   /// \brief Append an element to the UnionArray. This must be followed
 230   ///        by an append to the appropriate child builder.
 231   ///
 232   /// \param[in] next_type type_id of the child to which the next value will be appended.
 233   ///
 234   /// The corresponding child builder must be appended to independently after this method
 235   /// is called, and all other child builders must have null or empty value appended.
 236   Status Append(int8_t next_type) { return types_builder_.Append(next_type); }
 237
 238   Status AppendArraySlice(const ArrayData& array, int64_t offset,
 239                           int64_t length) override;
 240 };
 241
 242 }  // namespace arrow