be/src/storage/index/index_file_reader.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 | | // CLucene is third-party code and is not clean under -Wconversion (which |
21 | | // -Wshorten-64-to-32 belongs to). Whether its first expansion lands inside |
22 | | // someone else's suppressed region depends on include order, so suppress it |
23 | | // deliberately here (same pattern as inverted_index_common_impl.h). |
24 | | #ifdef __clang__ |
25 | | #pragma clang diagnostic push |
26 | | #pragma clang diagnostic ignored "-Wconversion" |
27 | | #endif |
28 | | #include <CLucene.h> // IWYU pragma: keep |
29 | | #include <CLucene/store/IndexInput.h> |
30 | | #ifdef __clang__ |
31 | | #pragma clang diagnostic pop |
32 | | #endif |
33 | | |
34 | | #include <gen_cpp/olap_file.pb.h> |
35 | | |
36 | | #include <map> |
37 | | #include <memory> |
38 | | #include <mutex> |
39 | | #include <shared_mutex> |
40 | | #include <string> |
41 | | #include <unordered_map> |
42 | | #include <utility> |
43 | | #include <vector> |
44 | | |
45 | | #include "common/be_mock_util.h" |
46 | | #include "common/config.h" |
47 | | #include "io/fs/file_system.h" |
48 | | #include "storage/index/index_file_writer.h" |
49 | | #include "storage/index/inverted/inverted_index_desc.h" |
50 | | #include "storage/index/snii/reader/logical_index_reader.h" |
51 | | #include "storage/index/snii/reader/snii_segment_reader.h" |
52 | | #include "storage/index/snii/snii_bkd_searcher.h" |
53 | | #include "storage/index/snii/snii_doris_adapter.h" |
54 | | |
55 | | namespace doris { |
56 | | class TabletIndex; |
57 | | namespace segment_v2 { |
58 | | class ReaderFileEntry; |
59 | | class DorisCompoundReader; |
60 | | |
61 | | // A singleton class responsible for reading index files, managing file entries, and providing interfaces to access index data. |
62 | | // The singleton object is at segment level, and it is shared by all threads that read the same segment (even across different queries). |
63 | | // It is created when the first index reader is initialized, and destroyed when the segment is closed. |
64 | | class IndexFileReader { |
65 | | public: |
66 | | // Modern C++ using std::unordered_map with smart pointers for automatic memory management |
67 | | using EntriesType = std::unordered_map<std::string, std::unique_ptr<ReaderFileEntry>>; |
68 | | // Map to hold the file entries for each index ID. |
69 | | using IndicesEntriesMap = |
70 | | std::map<std::pair<int64_t, std::string>, std::unique_ptr<EntriesType>>; |
71 | | |
72 | | IndexFileReader(io::FileSystemSPtr fs, std::string index_path_prefix, |
73 | | InvertedIndexStorageFormatPB storage_format, |
74 | | InvertedIndexFileInfo idx_file_info = InvertedIndexFileInfo(), |
75 | | int64_t tablet_id = -1) |
76 | 3.36k | : _fs(std::move(fs)), |
77 | 3.36k | _index_path_prefix(std::move(index_path_prefix)), |
78 | 3.36k | _storage_format(storage_format), |
79 | 3.36k | _idx_file_info(std::move(idx_file_info)), |
80 | 3.36k | _tablet_id(tablet_id) {} |
81 | 3.36k | virtual ~IndexFileReader() = default; |
82 | | |
83 | | MOCK_FUNCTION Status init(int32_t read_buffer_size = config::inverted_index_read_buffer_size, |
84 | | const io::IOContext* io_ctx = nullptr); |
85 | | MOCK_FUNCTION Result<std::unique_ptr<DorisCompoundReader, DirectoryDeleter>> open( |
86 | | const TabletIndex* index_meta, const io::IOContext* io_ctx = nullptr) const; |
87 | | // Opens one BLOB logical index of kind kBkd: resolves its named sub-files |
88 | | // (bkd_data / bkd_index / bkd_nulls) into extents through the CONTAINER's own |
89 | | // directory and hands back a reader bound to this IndexFileReader's file. |
90 | | // The caller must keep this IndexFileReader alive for the reader's lifetime, |
91 | | // exactly as open_snii_index requires. |
92 | | Result<std::unique_ptr<doris::snii::bkd::BkdSearcher>> open_snii_bkd_index( |
93 | | const TabletIndex* index_meta, const io::IOContext* io_ctx) const; |
94 | | Result<std::unique_ptr<doris::snii::reader::LogicalIndexReader>> open_snii_index( |
95 | | const TabletIndex* index_meta, const io::IOContext* io_ctx = nullptr, |
96 | | doris::snii::reader::LogicalIndexOpenMode open_mode = |
97 | | doris::snii::reader::LogicalIndexOpenMode::kQuery) const; |
98 | | // SNII only: builds the fully validated inheritance view of this container |
99 | | // for a BUILD INDEX rewrite. `segment_doc_count` is the segment's row count; |
100 | | // every kept logical index must agree with it. |
101 | | Status prepare_snii_rewrite_snapshot( |
102 | | const std::vector<doris::snii::reader::LogicalIndexKey>& keep, |
103 | | uint64_t segment_doc_count, doris::snii::reader::SniiRewriteSnapshot* out) const; |
104 | | // SNII only: the raw byte source backing this container, handed to |
105 | | // SniiCompoundWriter::inherit for the sequential prefix copy. |
106 | 3 | doris::snii::io::FileReader* snii_io_reader() const { return _snii_file_reader.get(); } |
107 | | // SNII only: true when the opened container holds a blob logical index |
108 | | // (BKD / ANN). False when this reader opened no SNII container at all. |
109 | 3 | bool snii_has_blob_index() const { |
110 | 3 | std::shared_lock<std::shared_mutex> lock(_mutex); |
111 | 3 | return _snii_segment_reader != nullptr && _snii_segment_reader->has_blob_index(); |
112 | 3 | } |
113 | | void debug_file_entries(); |
114 | | std::string get_index_file_cache_key(const TabletIndex* index_meta) const; |
115 | | std::string get_index_file_path(const TabletIndex* index_meta) const; |
116 | | Status index_file_exist(const TabletIndex* index_meta, bool* res) const; |
117 | | Status has_null(const TabletIndex* index_meta, bool* res) const; |
118 | | Result<InvertedIndexDirectoryMap> get_all_directories(); |
119 | | // open file v2, init _stream |
120 | 2 | int64_t get_inverted_file_size() const { |
121 | 2 | if (_storage_format == InvertedIndexStorageFormatPB::SNII) { |
122 | 2 | return _snii_file_reader == nullptr ? 0 : _snii_file_reader->size(); |
123 | 2 | } |
124 | 0 | return _stream == nullptr ? 0 : _stream->length(); |
125 | 2 | } |
126 | 4 | const std::string& get_index_path_prefix() const { return _index_path_prefix; } |
127 | 2.75k | InvertedIndexStorageFormatPB get_storage_format() const { return _storage_format; } |
128 | | friend IndexFileWriter; |
129 | | |
130 | | protected: |
131 | | Status _init_from(int32_t read_buffer_size, const io::IOContext* io_ctx); |
132 | | Status _init_snii(const io::IOContext* io_ctx); |
133 | | Result<std::unique_ptr<DorisCompoundReader, DirectoryDeleter>> _open( |
134 | | int64_t index_id, const std::string& index_suffix, |
135 | | const io::IOContext* io_ctx = nullptr) const; |
136 | | |
137 | | private: |
138 | | IndicesEntriesMap _indices_entries; |
139 | | std::unique_ptr<CL_NS(store)::IndexInput> _stream = nullptr; |
140 | | std::shared_ptr<snii_doris::DorisSniiFileReader> _snii_file_reader; |
141 | | std::unique_ptr<doris::snii::reader::SniiSegmentReader> _snii_segment_reader; |
142 | | const io::FileSystemSPtr _fs; |
143 | | std::string _index_path_prefix; |
144 | | int32_t _read_buffer_size = -1; |
145 | | InvertedIndexStorageFormatPB _storage_format; |
146 | | mutable std::shared_mutex _mutex; // Use mutable for const read operations |
147 | | bool _inited = false; |
148 | | InvertedIndexFileInfo _idx_file_info; |
149 | | int64_t _tablet_id = -1; |
150 | | }; |
151 | | |
152 | | } // namespace segment_v2 |
153 | | } // namespace doris |