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 <stddef.h> |
21 | | #include <stdint.h> |
22 | | |
23 | | #include <algorithm> |
24 | | #include <cstdint> |
25 | | #include <memory> |
26 | | #include <string> |
27 | | #include <vector> |
28 | | |
29 | | #include "common/consts.h" |
30 | | #include "core/column/column.h" |
31 | | #include "exprs/aggregate/aggregate_function.h" |
32 | | #include "io/io_common.h" |
33 | | #include "runtime/thread_context.h" |
34 | | #include "storage/olap_common.h" |
35 | | #include "storage/tablet/tablet_schema.h" |
36 | | #include "storage/utils.h" |
37 | | |
38 | | namespace doris { |
39 | | |
40 | | // The class is used to represent row's format in memory. Each row contains |
41 | | // multiple columns, some of which are key-columns (the rest are value-columns). |
42 | | // NOTE: If both key-columns and value-columns exist, then the key-columns |
43 | | // must be placed before value-columns. |
44 | | // |
45 | | // To compare two rows whose schemas are different, but they are from the same origin |
46 | | // we store all column schema maybe accessed here. And default access through column id |
47 | | class Schema; |
48 | | using SchemaSPtr = std::shared_ptr<const Schema>; |
49 | | class Schema { |
50 | | public: |
51 | | // All the columns of one table may exist in the columns param, but col_ids is only a subset. |
52 | 7.40k | Schema(const std::vector<TabletColumnPtr>& columns, const std::vector<ColumnId>& col_ids) { |
53 | 7.40k | size_t num_key_columns = 0; |
54 | 71.5k | for (int i = 0; i < columns.size(); ++i) { |
55 | 64.1k | if (columns[i]->is_key()) { |
56 | 6.58k | ++num_key_columns; |
57 | 6.58k | } |
58 | 64.1k | if (columns[i]->name() == DELETE_SIGN) { |
59 | 1.90k | _delete_sign_idx = i; |
60 | 1.90k | } |
61 | 64.1k | if (columns[i]->name().starts_with(BeConsts::GLOBAL_ROWID_COL)) { |
62 | 0 | _rowid_col_idx = i; |
63 | 0 | } |
64 | 64.1k | if (columns[i]->name() == VERSION_COL) { |
65 | 40 | _version_col_idx = i; |
66 | 40 | } |
67 | 64.1k | if (columns[i]->name() == BINLOG_TSO_COL) { |
68 | 24 | _tso_col_idx = i; |
69 | 24 | } |
70 | 64.1k | if (columns[i]->name() == BINLOG_LSN_COL) { |
71 | 18 | _lsn_col_idx = i; |
72 | 18 | } |
73 | 64.1k | if (columns[i]->name() == BINLOG_OP_COL) { |
74 | 18 | _op_col_idx = i; |
75 | 18 | } |
76 | 64.1k | if (columns[i]->name() == COMMIT_TSO_COL) { |
77 | 4 | _commit_tso_col_idx = i; |
78 | 4 | } |
79 | 64.1k | } |
80 | 7.40k | _init(columns, col_ids, num_key_columns); |
81 | 7.40k | } |
82 | | |
83 | | Schema(const Schema&); |
84 | | Schema& operator=(const Schema& other); |
85 | | |
86 | | ~Schema(); |
87 | | |
88 | | static DataTypePtr get_data_type_ptr(const TabletColumn& column); |
89 | | |
90 | | static IColumn::MutablePtr get_predicate_column_ptr(const DataTypePtr& data_type, |
91 | | const ReaderType reader_type); |
92 | | |
93 | 80.4k | const std::vector<TabletColumnPtr>& columns() const { return _cols; } |
94 | | |
95 | 1.23M | const TabletColumn* column(ColumnId cid) const { return _cols[cid].get(); } |
96 | | |
97 | 2.77k | size_t num_key_columns() const { return _num_key_columns; } |
98 | | |
99 | 16.3k | size_t num_columns() const { return _cols.size(); } |
100 | 1.00M | size_t num_column_ids() const { return _col_ids.size(); } |
101 | 67.6k | const std::vector<ColumnId>& column_ids() const { return _col_ids; } |
102 | 57.5k | ColumnId column_id(size_t index) const { return _col_ids[index]; } |
103 | 29.2k | int column_index(ColumnId cid) const { return _column_id_to_index[cid]; } |
104 | 0 | const std::vector<int>& column_id_to_index() const { return _column_id_to_index; } |
105 | 0 | int32_t delete_sign_idx() const { return _delete_sign_idx; } |
106 | 0 | bool has_sequence_col() const { return _has_sequence_col; } |
107 | 4.44k | int32_t rowid_col_idx() const { return _rowid_col_idx; } |
108 | 12.9k | int32_t version_col_idx() const { return _version_col_idx; } |
109 | 69 | int32_t commit_tso_col_idx() const { return _commit_tso_col_idx; } |
110 | 4 | int32_t tso_col_idx() const { return _tso_col_idx; } |
111 | 0 | int32_t lsn_col_idx() const { return _lsn_col_idx; } |
112 | 0 | int32_t op_col_idx() const { return _op_col_idx; } |
113 | | // Don't use. |
114 | | // TODO: memory size of Schema cannot be accurately tracked. |
115 | | // In some places, temporarily use num_columns() as Schema size. |
116 | 0 | int64_t mem_size() const { return _mem_size; } |
117 | | |
118 | | private: |
119 | | void _init(const std::vector<TabletColumnPtr>& cols, const std::vector<ColumnId>& col_ids, |
120 | | size_t num_key_columns); |
121 | | |
122 | | void _copy_from(const Schema& other); |
123 | | |
124 | | // NOTE: The ColumnId here represents the sequential index number (starting from 0) of |
125 | | // a column in current row, not the unique id-identifier of each column |
126 | | std::vector<ColumnId> _col_ids; |
127 | | // NOTE: _cols[cid] can only be accessed when the cid is |
128 | | // contained in _col_ids |
129 | | std::vector<TabletColumnPtr> _cols; |
130 | | // Tablet column id -> slot index in this Schema. |
131 | | std::vector<int> _column_id_to_index; |
132 | | |
133 | | size_t _num_key_columns; |
134 | | int32_t _delete_sign_idx = -1; |
135 | | bool _has_sequence_col = false; |
136 | | int32_t _rowid_col_idx = -1; |
137 | | int32_t _version_col_idx = -1; |
138 | | int32_t _commit_tso_col_idx = -1; |
139 | | int32_t _tso_col_idx = -1; |
140 | | int32_t _lsn_col_idx = -1; |
141 | | int32_t _op_col_idx = -1; |
142 | | int64_t _mem_size = 0; |
143 | | }; |
144 | | |
145 | | } // namespace doris |