1*9507f98cSAndroid Build Coastguard Worker // Copyright (c) 2011 The LevelDB Authors. All rights reserved.
2*9507f98cSAndroid Build Coastguard Worker // Use of this source code is governed by a BSD-style license that can be
3*9507f98cSAndroid Build Coastguard Worker // found in the LICENSE file. See the AUTHORS file for names of contributors.
4*9507f98cSAndroid Build Coastguard Worker //
5*9507f98cSAndroid Build Coastguard Worker // BlockBuilder generates blocks where keys are prefix-compressed:
6*9507f98cSAndroid Build Coastguard Worker //
7*9507f98cSAndroid Build Coastguard Worker // When we store a key, we drop the prefix shared with the previous
8*9507f98cSAndroid Build Coastguard Worker // string. This helps reduce the space requirement significantly.
9*9507f98cSAndroid Build Coastguard Worker // Furthermore, once every K keys, we do not apply the prefix
10*9507f98cSAndroid Build Coastguard Worker // compression and store the entire key. We call this a "restart
11*9507f98cSAndroid Build Coastguard Worker // point". The tail end of the block stores the offsets of all of the
12*9507f98cSAndroid Build Coastguard Worker // restart points, and can be used to do a binary search when looking
13*9507f98cSAndroid Build Coastguard Worker // for a particular key. Values are stored as-is (without compression)
14*9507f98cSAndroid Build Coastguard Worker // immediately following the corresponding key.
15*9507f98cSAndroid Build Coastguard Worker //
16*9507f98cSAndroid Build Coastguard Worker // An entry for a particular key-value pair has the form:
17*9507f98cSAndroid Build Coastguard Worker // shared_bytes: varint32
18*9507f98cSAndroid Build Coastguard Worker // unshared_bytes: varint32
19*9507f98cSAndroid Build Coastguard Worker // value_length: varint32
20*9507f98cSAndroid Build Coastguard Worker // key_delta: char[unshared_bytes]
21*9507f98cSAndroid Build Coastguard Worker // value: char[value_length]
22*9507f98cSAndroid Build Coastguard Worker // shared_bytes == 0 for restart points.
23*9507f98cSAndroid Build Coastguard Worker //
24*9507f98cSAndroid Build Coastguard Worker // The trailer of the block has the form:
25*9507f98cSAndroid Build Coastguard Worker // restarts: uint32[num_restarts]
26*9507f98cSAndroid Build Coastguard Worker // num_restarts: uint32
27*9507f98cSAndroid Build Coastguard Worker // restarts[i] contains the offset within the block of the ith restart point.
28*9507f98cSAndroid Build Coastguard Worker
29*9507f98cSAndroid Build Coastguard Worker #include "table/block_builder.h"
30*9507f98cSAndroid Build Coastguard Worker
31*9507f98cSAndroid Build Coastguard Worker #include <algorithm>
32*9507f98cSAndroid Build Coastguard Worker #include <cassert>
33*9507f98cSAndroid Build Coastguard Worker
34*9507f98cSAndroid Build Coastguard Worker #include "leveldb/comparator.h"
35*9507f98cSAndroid Build Coastguard Worker #include "leveldb/options.h"
36*9507f98cSAndroid Build Coastguard Worker #include "util/coding.h"
37*9507f98cSAndroid Build Coastguard Worker
38*9507f98cSAndroid Build Coastguard Worker namespace leveldb {
39*9507f98cSAndroid Build Coastguard Worker
BlockBuilder(const Options * options)40*9507f98cSAndroid Build Coastguard Worker BlockBuilder::BlockBuilder(const Options* options)
41*9507f98cSAndroid Build Coastguard Worker : options_(options), restarts_(), counter_(0), finished_(false) {
42*9507f98cSAndroid Build Coastguard Worker assert(options->block_restart_interval >= 1);
43*9507f98cSAndroid Build Coastguard Worker restarts_.push_back(0); // First restart point is at offset 0
44*9507f98cSAndroid Build Coastguard Worker }
45*9507f98cSAndroid Build Coastguard Worker
Reset()46*9507f98cSAndroid Build Coastguard Worker void BlockBuilder::Reset() {
47*9507f98cSAndroid Build Coastguard Worker buffer_.clear();
48*9507f98cSAndroid Build Coastguard Worker restarts_.clear();
49*9507f98cSAndroid Build Coastguard Worker restarts_.push_back(0); // First restart point is at offset 0
50*9507f98cSAndroid Build Coastguard Worker counter_ = 0;
51*9507f98cSAndroid Build Coastguard Worker finished_ = false;
52*9507f98cSAndroid Build Coastguard Worker last_key_.clear();
53*9507f98cSAndroid Build Coastguard Worker }
54*9507f98cSAndroid Build Coastguard Worker
CurrentSizeEstimate() const55*9507f98cSAndroid Build Coastguard Worker size_t BlockBuilder::CurrentSizeEstimate() const {
56*9507f98cSAndroid Build Coastguard Worker return (buffer_.size() + // Raw data buffer
57*9507f98cSAndroid Build Coastguard Worker restarts_.size() * sizeof(uint32_t) + // Restart array
58*9507f98cSAndroid Build Coastguard Worker sizeof(uint32_t)); // Restart array length
59*9507f98cSAndroid Build Coastguard Worker }
60*9507f98cSAndroid Build Coastguard Worker
Finish()61*9507f98cSAndroid Build Coastguard Worker Slice BlockBuilder::Finish() {
62*9507f98cSAndroid Build Coastguard Worker // Append restart array
63*9507f98cSAndroid Build Coastguard Worker for (size_t i = 0; i < restarts_.size(); i++) {
64*9507f98cSAndroid Build Coastguard Worker PutFixed32(&buffer_, restarts_[i]);
65*9507f98cSAndroid Build Coastguard Worker }
66*9507f98cSAndroid Build Coastguard Worker PutFixed32(&buffer_, restarts_.size());
67*9507f98cSAndroid Build Coastguard Worker finished_ = true;
68*9507f98cSAndroid Build Coastguard Worker return Slice(buffer_);
69*9507f98cSAndroid Build Coastguard Worker }
70*9507f98cSAndroid Build Coastguard Worker
Add(const Slice & key,const Slice & value)71*9507f98cSAndroid Build Coastguard Worker void BlockBuilder::Add(const Slice& key, const Slice& value) {
72*9507f98cSAndroid Build Coastguard Worker Slice last_key_piece(last_key_);
73*9507f98cSAndroid Build Coastguard Worker assert(!finished_);
74*9507f98cSAndroid Build Coastguard Worker assert(counter_ <= options_->block_restart_interval);
75*9507f98cSAndroid Build Coastguard Worker assert(buffer_.empty() // No values yet?
76*9507f98cSAndroid Build Coastguard Worker || options_->comparator->Compare(key, last_key_piece) > 0);
77*9507f98cSAndroid Build Coastguard Worker size_t shared = 0;
78*9507f98cSAndroid Build Coastguard Worker if (counter_ < options_->block_restart_interval) {
79*9507f98cSAndroid Build Coastguard Worker // See how much sharing to do with previous string
80*9507f98cSAndroid Build Coastguard Worker const size_t min_length = std::min(last_key_piece.size(), key.size());
81*9507f98cSAndroid Build Coastguard Worker while ((shared < min_length) && (last_key_piece[shared] == key[shared])) {
82*9507f98cSAndroid Build Coastguard Worker shared++;
83*9507f98cSAndroid Build Coastguard Worker }
84*9507f98cSAndroid Build Coastguard Worker } else {
85*9507f98cSAndroid Build Coastguard Worker // Restart compression
86*9507f98cSAndroid Build Coastguard Worker restarts_.push_back(buffer_.size());
87*9507f98cSAndroid Build Coastguard Worker counter_ = 0;
88*9507f98cSAndroid Build Coastguard Worker }
89*9507f98cSAndroid Build Coastguard Worker const size_t non_shared = key.size() - shared;
90*9507f98cSAndroid Build Coastguard Worker
91*9507f98cSAndroid Build Coastguard Worker // Add "<shared><non_shared><value_size>" to buffer_
92*9507f98cSAndroid Build Coastguard Worker PutVarint32(&buffer_, shared);
93*9507f98cSAndroid Build Coastguard Worker PutVarint32(&buffer_, non_shared);
94*9507f98cSAndroid Build Coastguard Worker PutVarint32(&buffer_, value.size());
95*9507f98cSAndroid Build Coastguard Worker
96*9507f98cSAndroid Build Coastguard Worker // Add string delta to buffer_ followed by value
97*9507f98cSAndroid Build Coastguard Worker buffer_.append(key.data() + shared, non_shared);
98*9507f98cSAndroid Build Coastguard Worker buffer_.append(value.data(), value.size());
99*9507f98cSAndroid Build Coastguard Worker
100*9507f98cSAndroid Build Coastguard Worker // Update state
101*9507f98cSAndroid Build Coastguard Worker last_key_.resize(shared);
102*9507f98cSAndroid Build Coastguard Worker last_key_.append(key.data() + shared, non_shared);
103*9507f98cSAndroid Build Coastguard Worker assert(Slice(last_key_) == key);
104*9507f98cSAndroid Build Coastguard Worker counter_++;
105*9507f98cSAndroid Build Coastguard Worker }
106*9507f98cSAndroid Build Coastguard Worker
107*9507f98cSAndroid Build Coastguard Worker } // namespace leveldb
108