/root/doris/be/src/vec/jsonb/serialize.cpp
Line | Count | Source (jump to first uncovered line) |
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 "vec/jsonb/serialize.h" |
19 | | |
20 | | #include <assert.h> |
21 | | |
22 | | #include <algorithm> |
23 | | #include <memory> |
24 | | #include <vector> |
25 | | |
26 | | #include "olap/tablet_schema.h" |
27 | | #include "runtime/descriptors.h" |
28 | | #include "runtime/jsonb_value.h" |
29 | | #include "runtime/primitive_type.h" |
30 | | #include "runtime/types.h" |
31 | | #include "util/jsonb_document.h" |
32 | | #include "util/jsonb_stream.h" |
33 | | #include "util/jsonb_writer.h" |
34 | | #include "vec/columns/column.h" |
35 | | #include "vec/columns/column_string.h" |
36 | | #include "vec/common/arena.h" |
37 | | #include "vec/common/string_ref.h" |
38 | | #include "vec/core/column_with_type_and_name.h" |
39 | | #include "vec/core/columns_with_type_and_name.h" |
40 | | #include "vec/data_types/data_type.h" |
41 | | #include "vec/data_types/serde/data_type_serde.h" |
42 | | #include "vec/io/reader_buffer.h" |
43 | | |
44 | | namespace doris::vectorized { |
45 | | |
46 | | void JsonbSerializeUtil::block_to_jsonb(const TabletSchema& schema, const Block& block, |
47 | | ColumnString& dst, int num_cols, |
48 | 4 | const DataTypeSerDeSPtrs& serdes) { |
49 | 4 | auto num_rows = block.rows(); |
50 | 4 | Arena pool; |
51 | 4 | assert(num_cols <= block.columns()); |
52 | 1.03k | for (int i = 0; i < num_rows; ++i) { |
53 | 1.03k | JsonbWriterT<JsonbOutStream> jsonb_writer; |
54 | 1.03k | jsonb_writer.writeStartObject(); |
55 | 10.2k | for (int j = 0; j < num_cols; ++j) { |
56 | 9.22k | const auto& column = block.get_by_position(j).column; |
57 | 9.22k | const auto& tablet_column = *schema.columns()[j]; |
58 | 9.22k | if (tablet_column.is_row_store_column()) { |
59 | | // ignore dst row store column |
60 | 0 | continue; |
61 | 0 | } |
62 | 9.22k | serdes[j]->write_one_cell_to_jsonb(*column, jsonb_writer, &pool, |
63 | 9.22k | tablet_column.unique_id(), i); |
64 | 9.22k | } |
65 | 1.03k | jsonb_writer.writeEndObject(); |
66 | 1.03k | dst.insert_data(jsonb_writer.getOutput()->getBuffer(), jsonb_writer.getOutput()->getSize()); |
67 | 1.03k | } |
68 | 4 | } |
69 | | |
70 | | // batch rows |
71 | | void JsonbSerializeUtil::jsonb_to_block(const DataTypeSerDeSPtrs& serdes, |
72 | | const ColumnString& jsonb_column, |
73 | | const std::unordered_map<uint32_t, uint32_t>& col_id_to_idx, |
74 | | Block& dst, |
75 | 4 | const std::vector<std::string>& default_values) { |
76 | 1.03k | for (int i = 0; i < jsonb_column.size(); ++i) { |
77 | 1.03k | StringRef jsonb_data = jsonb_column.get_data_at(i); |
78 | 1.03k | jsonb_to_block(serdes, jsonb_data.data, jsonb_data.size, col_id_to_idx, dst, |
79 | 1.03k | default_values); |
80 | 1.03k | } |
81 | 4 | } |
82 | | |
83 | | // single row |
84 | | void JsonbSerializeUtil::jsonb_to_block(const DataTypeSerDeSPtrs& serdes, const char* data, |
85 | | size_t size, |
86 | | const std::unordered_map<uint32_t, uint32_t>& col_id_to_idx, |
87 | | Block& dst, |
88 | 1.03k | const std::vector<std::string>& default_values) { |
89 | 1.03k | auto pdoc = JsonbDocument::createDocument(data, size); |
90 | 1.03k | JsonbDocument& doc = *pdoc; |
91 | 1.03k | size_t num_rows = dst.rows(); |
92 | 1.03k | size_t filled_columns = 0; |
93 | 10.2k | for (auto it = doc->begin(); it != doc->end(); ++it) { |
94 | 9.22k | auto col_it = col_id_to_idx.find(it->getKeyId()); |
95 | 9.22k | if (col_it != col_id_to_idx.end()) { |
96 | 9.22k | MutableColumnPtr dst_column = |
97 | 9.22k | dst.get_by_position(col_it->second).column->assume_mutable(); |
98 | 9.22k | serdes[col_it->second]->read_one_cell_from_jsonb(*dst_column, it->value()); |
99 | 9.22k | ++filled_columns; |
100 | 9.22k | } |
101 | 9.22k | } |
102 | 1.03k | if (filled_columns < dst.columns()) { |
103 | | // fill missing slot |
104 | 0 | for (int i = 0; i < dst.columns(); ++i) { |
105 | 0 | const auto& column_type_name = dst.get_by_position(i); |
106 | 0 | MutableColumnPtr col = column_type_name.column->assume_mutable(); |
107 | 0 | if (col->size() < num_rows + 1) { |
108 | 0 | DCHECK(col->size() == num_rows); |
109 | 0 | if (default_values[i].empty()) { |
110 | 0 | col->insert_default(); |
111 | 0 | } else { |
112 | 0 | Slice value(default_values[i].data(), default_values[i].size()); |
113 | 0 | DataTypeSerDe::FormatOptions opt; |
114 | 0 | opt.converted_from_string = true; |
115 | 0 | static_cast<void>(serdes[i]->deserialize_one_cell_from_json(*col, value, opt)); |
116 | 0 | } |
117 | 0 | } |
118 | 0 | DCHECK(col->size() == num_rows + 1); |
119 | 0 | } |
120 | 0 | } |
121 | 1.03k | } |
122 | | |
123 | | } // namespace doris::vectorized |