/bitcoin/src/leveldb/table/block_builder.cc
Line | Count | Source |
1 | | // Copyright (c) 2011 The LevelDB Authors. All rights reserved. |
2 | | // Use of this source code is governed by a BSD-style license that can be |
3 | | // found in the LICENSE file. See the AUTHORS file for names of contributors. |
4 | | // |
5 | | // BlockBuilder generates blocks where keys are prefix-compressed: |
6 | | // |
7 | | // When we store a key, we drop the prefix shared with the previous |
8 | | // string. This helps reduce the space requirement significantly. |
9 | | // Furthermore, once every K keys, we do not apply the prefix |
10 | | // compression and store the entire key. We call this a "restart |
11 | | // point". The tail end of the block stores the offsets of all of the |
12 | | // restart points, and can be used to do a binary search when looking |
13 | | // for a particular key. Values are stored as-is (without compression) |
14 | | // immediately following the corresponding key. |
15 | | // |
16 | | // An entry for a particular key-value pair has the form: |
17 | | // shared_bytes: varint32 |
18 | | // unshared_bytes: varint32 |
19 | | // value_length: varint32 |
20 | | // key_delta: char[unshared_bytes] |
21 | | // value: char[value_length] |
22 | | // shared_bytes == 0 for restart points. |
23 | | // |
24 | | // The trailer of the block has the form: |
25 | | // restarts: uint32[num_restarts] |
26 | | // num_restarts: uint32 |
27 | | // restarts[i] contains the offset within the block of the ith restart point. |
28 | | |
29 | | #include "table/block_builder.h" |
30 | | |
31 | | #include <assert.h> |
32 | | |
33 | | #include <algorithm> |
34 | | |
35 | | #include "leveldb/comparator.h" |
36 | | #include "leveldb/options.h" |
37 | | #include "util/coding.h" |
38 | | |
39 | | namespace leveldb { |
40 | | |
41 | | BlockBuilder::BlockBuilder(const Options* options) |
42 | 72 | : options_(options), restarts_(), counter_(0), finished_(false) { |
43 | 72 | assert(options->block_restart_interval >= 1); Branch (43:3): [True: 72, False: 0]
|
44 | 72 | restarts_.push_back(0); // First restart point is at offset 0 |
45 | 72 | } |
46 | | |
47 | 673 | void BlockBuilder::Reset() { |
48 | 673 | buffer_.clear(); |
49 | 673 | restarts_.clear(); |
50 | 673 | restarts_.push_back(0); // First restart point is at offset 0 |
51 | 673 | counter_ = 0; |
52 | 673 | finished_ = false; |
53 | 673 | last_key_.clear(); |
54 | 673 | } |
55 | | |
56 | 11.7k | size_t BlockBuilder::CurrentSizeEstimate() const { |
57 | 11.7k | return (buffer_.size() + // Raw data buffer |
58 | 11.7k | restarts_.size() * sizeof(uint32_t) + // Restart array |
59 | 11.7k | sizeof(uint32_t)); // Restart array length |
60 | 11.7k | } |
61 | | |
62 | 673 | Slice BlockBuilder::Finish() { |
63 | | // Append restart array |
64 | 2.41k | for (size_t i = 0; i < restarts_.size(); i++) { Branch (64:22): [True: 1.73k, False: 673]
|
65 | 1.73k | PutFixed32(&buffer_, restarts_[i]); |
66 | 1.73k | } |
67 | 673 | PutFixed32(&buffer_, restarts_.size()); |
68 | 673 | finished_ = true; |
69 | 673 | return Slice(buffer_); |
70 | 673 | } |
71 | | |
72 | 12.3k | void BlockBuilder::Add(const Slice& key, const Slice& value) { |
73 | 12.3k | Slice last_key_piece(last_key_); |
74 | 12.3k | assert(!finished_); Branch (74:3): [True: 12.3k, False: 0]
|
75 | 12.3k | assert(counter_ <= options_->block_restart_interval); Branch (75:3): [True: 12.3k, False: 0]
|
76 | 12.3k | assert(buffer_.empty() // No values yet? Branch (76:3): [True: 673, False: 11.7k]
Branch (76:3): [True: 11.7k, False: 0]
Branch (76:3): [True: 12.3k, False: 0]
|
77 | 12.3k | || options_->comparator->Compare(key, last_key_piece) > 0); |
78 | 12.3k | size_t shared = 0; |
79 | 12.3k | if (counter_ < options_->block_restart_interval) { Branch (79:7): [True: 11.3k, False: 1.06k]
|
80 | | // See how much sharing to do with previous string |
81 | 11.3k | const size_t min_length = std::min(last_key_piece.size(), key.size()); |
82 | 43.3k | while ((shared < min_length) && (last_key_piece[shared] == key[shared])) { Branch (82:12): [True: 42.6k, False: 673]
Branch (82:37): [True: 32.0k, False: 10.6k]
|
83 | 32.0k | shared++; |
84 | 32.0k | } |
85 | 11.3k | } else { |
86 | | // Restart compression |
87 | 1.06k | restarts_.push_back(buffer_.size()); |
88 | 1.06k | counter_ = 0; |
89 | 1.06k | } |
90 | 12.3k | const size_t non_shared = key.size() - shared; |
91 | | |
92 | | // Add "<shared><non_shared><value_size>" to buffer_ |
93 | 12.3k | PutVarint32(&buffer_, shared); |
94 | 12.3k | PutVarint32(&buffer_, non_shared); |
95 | 12.3k | PutVarint32(&buffer_, value.size()); |
96 | | |
97 | | // Add string delta to buffer_ followed by value |
98 | 12.3k | buffer_.append(key.data() + shared, non_shared); |
99 | 12.3k | buffer_.append(value.data(), value.size()); |
100 | | |
101 | | // Update state |
102 | 12.3k | last_key_.resize(shared); |
103 | 12.3k | last_key_.append(key.data() + shared, non_shared); |
104 | 12.3k | assert(Slice(last_key_) == key); Branch (104:3): [True: 12.3k, False: 0]
|
105 | 12.3k | counter_++; |
106 | 12.3k | } |
107 | | |
108 | | } // namespace leveldb |