| 1 | #pragma once |
| 2 | |
| 3 | #include <Columns/IColumn.h> |
| 4 | #include <Columns/IColumnImpl.h> |
| 5 | #include <Columns/ColumnVector.h> |
| 6 | #include <Core/Defines.h> |
| 7 | #include <Common/typeid_cast.h> |
| 8 | #include <Common/assert_cast.h> |
| 9 | |
| 10 | |
| 11 | namespace DB |
| 12 | { |
| 13 | |
| 14 | /** A column of array values. |
| 15 | * In memory, it is represented as one column of a nested type, whose size is equal to the sum of the sizes of all arrays, |
| 16 | * and as an array of offsets in it, which allows you to get each element. |
| 17 | */ |
| 18 | class ColumnArray final : public COWHelper<IColumn, ColumnArray> |
| 19 | { |
| 20 | private: |
| 21 | friend class COWHelper<IColumn, ColumnArray>; |
| 22 | |
| 23 | /** Create an array column with specified values and offsets. */ |
| 24 | ColumnArray(MutableColumnPtr && nested_column, MutableColumnPtr && offsets_column); |
| 25 | |
| 26 | /** Create an empty column of arrays with the type of values as in the column `nested_column` */ |
| 27 | explicit ColumnArray(MutableColumnPtr && nested_column); |
| 28 | |
| 29 | ColumnArray(const ColumnArray &) = default; |
| 30 | |
| 31 | public: |
| 32 | /** Create immutable column using immutable arguments. This arguments may be shared with other columns. |
| 33 | * Use IColumn::mutate in order to make mutable column and mutate shared nested columns. |
| 34 | */ |
| 35 | using Base = COWHelper<IColumn, ColumnArray>; |
| 36 | |
| 37 | static Ptr create(const ColumnPtr & nested_column, const ColumnPtr & offsets_column) |
| 38 | { |
| 39 | return ColumnArray::create(nested_column->assumeMutable(), offsets_column->assumeMutable()); |
| 40 | } |
| 41 | |
| 42 | static Ptr create(const ColumnPtr & nested_column) |
| 43 | { |
| 44 | return ColumnArray::create(nested_column->assumeMutable()); |
| 45 | } |
| 46 | |
| 47 | template <typename ... Args, typename = typename std::enable_if<IsMutableColumns<Args ...>::value>::type> |
| 48 | static MutablePtr create(Args &&... args) { return Base::create(std::forward<Args>(args)...); } |
| 49 | |
| 50 | /** On the index i there is an offset to the beginning of the i + 1 -th element. */ |
| 51 | using ColumnOffsets = ColumnVector<Offset>; |
| 52 | |
| 53 | std::string getName() const override; |
| 54 | const char * getFamilyName() const override { return "Array" ; } |
| 55 | MutableColumnPtr cloneResized(size_t size) const override; |
| 56 | size_t size() const override; |
| 57 | Field operator[](size_t n) const override; |
| 58 | void get(size_t n, Field & res) const override; |
| 59 | StringRef getDataAt(size_t n) const override; |
| 60 | void insertData(const char * pos, size_t length) override; |
| 61 | StringRef serializeValueIntoArena(size_t n, Arena & arena, char const *& begin) const override; |
| 62 | const char * deserializeAndInsertFromArena(const char * pos) override; |
| 63 | void updateHashWithValue(size_t n, SipHash & hash) const override; |
| 64 | void insertRangeFrom(const IColumn & src, size_t start, size_t length) override; |
| 65 | void insert(const Field & x) override; |
| 66 | void insertFrom(const IColumn & src_, size_t n) override; |
| 67 | void insertDefault() override; |
| 68 | void popBack(size_t n) override; |
| 69 | ColumnPtr filter(const Filter & filt, ssize_t result_size_hint) const override; |
| 70 | ColumnPtr permute(const Permutation & perm, size_t limit) const override; |
| 71 | ColumnPtr index(const IColumn & indexes, size_t limit) const override; |
| 72 | template <typename Type> ColumnPtr indexImpl(const PaddedPODArray<Type> & indexes, size_t limit) const; |
| 73 | int compareAt(size_t n, size_t m, const IColumn & rhs_, int nan_direction_hint) const override; |
| 74 | void getPermutation(bool reverse, size_t limit, int nan_direction_hint, Permutation & res) const override; |
| 75 | void reserve(size_t n) override; |
| 76 | size_t byteSize() const override; |
| 77 | size_t allocatedBytes() const override; |
| 78 | void protect() override; |
| 79 | ColumnPtr replicate(const Offsets & replicate_offsets) const override; |
| 80 | ColumnPtr convertToFullColumnIfConst() const override; |
| 81 | void getExtremes(Field & min, Field & max) const override; |
| 82 | |
| 83 | bool hasEqualOffsets(const ColumnArray & other) const; |
| 84 | |
| 85 | /** More efficient methods of manipulation */ |
| 86 | IColumn & getData() { return *data; } |
| 87 | const IColumn & getData() const { return *data; } |
| 88 | |
| 89 | IColumn & getOffsetsColumn() { return *offsets; } |
| 90 | const IColumn & getOffsetsColumn() const { return *offsets; } |
| 91 | |
| 92 | Offsets & ALWAYS_INLINE getOffsets() |
| 93 | { |
| 94 | return assert_cast<ColumnOffsets &>(*offsets).getData(); |
| 95 | } |
| 96 | |
| 97 | const Offsets & ALWAYS_INLINE getOffsets() const |
| 98 | { |
| 99 | return assert_cast<const ColumnOffsets &>(*offsets).getData(); |
| 100 | } |
| 101 | |
| 102 | const ColumnPtr & getDataPtr() const { return data; } |
| 103 | ColumnPtr & getDataPtr() { return data; } |
| 104 | |
| 105 | const ColumnPtr & getOffsetsPtr() const { return offsets; } |
| 106 | ColumnPtr & getOffsetsPtr() { return offsets; } |
| 107 | |
| 108 | MutableColumns scatter(ColumnIndex num_columns, const Selector & selector) const override |
| 109 | { |
| 110 | return scatterImpl<ColumnArray>(num_columns, selector); |
| 111 | } |
| 112 | |
| 113 | void gather(ColumnGathererStream & gatherer_stream) override; |
| 114 | |
| 115 | void forEachSubcolumn(ColumnCallback callback) override |
| 116 | { |
| 117 | callback(offsets); |
| 118 | callback(data); |
| 119 | } |
| 120 | |
| 121 | bool structureEquals(const IColumn & rhs) const override |
| 122 | { |
| 123 | if (auto rhs_concrete = typeid_cast<const ColumnArray *>(&rhs)) |
| 124 | return data->structureEquals(*rhs_concrete->data); |
| 125 | return false; |
| 126 | } |
| 127 | |
| 128 | private: |
| 129 | WrappedPtr data; |
| 130 | WrappedPtr offsets; |
| 131 | |
| 132 | size_t ALWAYS_INLINE offsetAt(ssize_t i) const { return getOffsets()[i - 1]; } |
| 133 | size_t ALWAYS_INLINE sizeAt(ssize_t i) const { return getOffsets()[i] - getOffsets()[i - 1]; } |
| 134 | |
| 135 | |
| 136 | /// Multiply values if the nested column is ColumnVector<T>. |
| 137 | template <typename T> |
| 138 | ColumnPtr replicateNumber(const Offsets & replicate_offsets) const; |
| 139 | |
| 140 | /// Multiply the values if the nested column is ColumnString. The code is too complicated. |
| 141 | ColumnPtr replicateString(const Offsets & replicate_offsets) const; |
| 142 | |
| 143 | /** Non-constant arrays of constant values are quite rare. |
| 144 | * Most functions can not work with them, and does not create such columns as a result. |
| 145 | * An exception is the function `replicate` (see FunctionsMiscellaneous.h), which has service meaning for the implementation of lambda functions. |
| 146 | * Only for its sake is the implementation of the `replicate` method for ColumnArray(ColumnConst). |
| 147 | */ |
| 148 | ColumnPtr replicateConst(const Offsets & replicate_offsets) const; |
| 149 | |
| 150 | /** The following is done by simply replicating of nested columns. |
| 151 | */ |
| 152 | ColumnPtr replicateTuple(const Offsets & replicate_offsets) const; |
| 153 | ColumnPtr replicateNullable(const Offsets & replicate_offsets) const; |
| 154 | ColumnPtr replicateGeneric(const Offsets & replicate_offsets) const; |
| 155 | |
| 156 | |
| 157 | /// Specializations for the filter function. |
| 158 | template <typename T> |
| 159 | ColumnPtr filterNumber(const Filter & filt, ssize_t result_size_hint) const; |
| 160 | |
| 161 | ColumnPtr filterString(const Filter & filt, ssize_t result_size_hint) const; |
| 162 | ColumnPtr filterTuple(const Filter & filt, ssize_t result_size_hint) const; |
| 163 | ColumnPtr filterNullable(const Filter & filt, ssize_t result_size_hint) const; |
| 164 | ColumnPtr filterGeneric(const Filter & filt, ssize_t result_size_hint) const; |
| 165 | }; |
| 166 | |
| 167 | |
| 168 | } |
| 169 | |