be/src/exprs/virtual_slot_ref.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 | | #pragma once |
19 | | |
20 | | #include "exprs/vexpr.h" |
21 | | |
22 | | namespace doris { |
23 | | #include "common/compile_check_begin.h" |
24 | | class VirtualSlotRef MOCK_REMOVE(final) : public VExpr { |
25 | | ENABLE_FACTORY_CREATOR(VirtualSlotRef); |
26 | | |
27 | | public: |
28 | | VirtualSlotRef(const TExprNode& node); |
29 | | VirtualSlotRef(const SlotDescriptor* desc); |
30 | | |
31 | | Status prepare(RuntimeState* state, const RowDescriptor& desc, VExprContext* context) override; |
32 | | Status open(RuntimeState* state, VExprContext* context, |
33 | | FunctionContext::FunctionStateScope scope) override; |
34 | | Status execute_column(VExprContext* context, const Block* block, Selector* selector, |
35 | | size_t count, ColumnPtr& result_column) const override; |
36 | | const std::string& expr_name() const override; |
37 | | std::string expr_label() override; |
38 | | std::string debug_string() const override; |
39 | 822 | bool is_constant() const override { return false; } |
40 | 156 | int column_id() const { return _column_id; } |
41 | | int slot_id() const { return _slot_id; } |
42 | | bool equals(const VExpr& other) override; |
43 | 1 | size_t estimate_memory(const size_t rows) override { return 0; } |
44 | 0 | void collect_slot_column_ids(std::set<int>& column_ids) const override { |
45 | 0 | column_ids.insert(_column_id); |
46 | 0 | } |
47 | 894 | std::shared_ptr<VExpr> get_virtual_column_expr() const { return _virtual_column_expr; } |
48 | | |
49 | 0 | Status evaluate_inverted_index(VExprContext* context, uint32_t segment_num_rows) override { |
50 | 0 | return _virtual_column_expr->evaluate_inverted_index(context, segment_num_rows); |
51 | 0 | } |
52 | | |
53 | | /* |
54 | | @brief SQL expression tree patterns for ANN range search optimization. |
55 | | |
56 | | Pattern 1 (should not happen): |
57 | | SELECT * FROM tbl WHERE distance_function(columnA, ArrayLiteral) > 100 |
58 | | VirtualSlotRef |
59 | | | |
60 | | BINARY_PRED |
61 | | |---------------------------------------| |
62 | | | | |
63 | | FUNCTION_CALL(l2_distance_approximate) IntLiteral |
64 | | | |
65 | | |-----------------------| |
66 | | | | |
67 | | SlotRef ArrayLiteral |
68 | | |
69 | | Pattern 2 (optimizable case): |
70 | | SELECT distance_function(columnA, ArrayLiteral) AS dis FROM tbl WHERE dis > 100 |
71 | | BINARY_PRED |
72 | | | |
73 | | |---------------------------------------| |
74 | | | | |
75 | | VIRTUAL_SLOT_REF IntLiteral |
76 | | | |
77 | | FUNCTION_CALL(l2_distance_approximate) |
78 | | | |
79 | | |-----------------------| |
80 | | | | |
81 | | SlotRef ArrayLiteral |
82 | | */ |
83 | | |
84 | | /** |
85 | | * @brief Evaluates ANN range search using index-based optimization. |
86 | | * |
87 | | * This method implements the core logic for ANN range search optimization. |
88 | | * Instead of computing distances for all rows and then filtering, it uses |
89 | | * the ANN index to efficiently find only the rows within the specified range. |
90 | | * |
91 | | * The method: |
92 | | * 1. Extracts query parameters from the range search runtime info |
93 | | * 2. Calls the ANN index to perform range search |
94 | | * 3. Updates the row bitmap with matching results |
95 | | * 4. Collects performance statistics |
96 | | * |
97 | | * @param range_search_runtime Runtime info containing query vector, radius, and metrics |
98 | | * @param cid_to_index_iterators Vector of index iterators for each column |
99 | | * @param idx_to_cid Mapping from index position to column ID |
100 | | * @param column_iterators Vector of column iterators for data access |
101 | | * @param row_bitmap Output bitmap updated with matching row IDs |
102 | | * @param ann_index_stats Statistics collector for performance monitoring |
103 | | * @return Status indicating success or failure of the search operation |
104 | | */ |
105 | | Status evaluate_ann_range_search( |
106 | | const segment_v2::AnnRangeSearchRuntime& range_search_runtime, |
107 | | const std::vector<std::unique_ptr<segment_v2::IndexIterator>>& cid_to_index_iterators, |
108 | | const std::vector<ColumnId>& idx_to_cid, |
109 | | const std::vector<std::unique_ptr<segment_v2::ColumnIterator>>& column_iterators, |
110 | | roaring::Roaring& row_bitmap, segment_v2::AnnIndexStats& ann_index_stats) override; |
111 | | |
112 | | #ifdef BE_TEST |
113 | | // Test-only setter methods for unit testing |
114 | | void set_column_id(int column_id) { _column_id = column_id; } |
115 | | void set_column_name(const std::string* column_name) { _column_name = column_name; } |
116 | | void set_column_data_type(DataTypePtr column_data_type) { |
117 | | _column_data_type = std::move(column_data_type); |
118 | | } |
119 | | void set_virtual_column_expr(std::shared_ptr<VExpr> virtual_column_expr) { |
120 | | _virtual_column_expr = virtual_column_expr; |
121 | | } |
122 | | #endif |
123 | | |
124 | | private: |
125 | | int _column_id; ///< Column ID in the table schema |
126 | | int _slot_id; ///< Slot ID in the expression context |
127 | | const std::string* _column_name; ///< Column name for debugging/logging |
128 | | const std::string _column_label; ///< Column label for display purposes |
129 | | std::shared_ptr<VExpr> _virtual_column_expr; ///< Underlying virtual expression |
130 | | DataTypePtr _column_data_type; ///< Data type of the column |
131 | | }; |
132 | | #include "common/compile_check_end.h" |
133 | | } // namespace doris |