Packages
rocksdb
0.22.0
3.1.2
3.1.1
3.1.0
3.0.0
2.6.2
2.6.1
retired
2.6.0
retired
2.5.0
2.4.1
2.4.0
2.3.0
2.2.0
2.1.0
2.0.0
1.9.0
1.8.0
1.7.0
1.6.0
1.5.1
1.5.0
1.4.0
1.3.2
1.3.1
1.3.0
1.2.0
1.1.1
1.1.0
1.0.0
0.26.2
0.26.1
0.26.0
0.25.0
0.24.0
0.23.3
0.23.2
0.23.1
0.23.0
0.22.0
0.21.0
0.20.1
0.20.0
0.19.0
0.18.0
0.17.0
0.16.0
0.15.0
0.14.0
0.13.1
0.13.0
0.12.0
0.11.0
0.10.0
0.9.1
0.9.0
0.8.2
0.8.1
0.8.0
0.7.1
0.7.0
0.6.4
0.6.3
0.6.2
0.6.1
0.6.0
RocksDB for Erlang
Current section
Files
Jump to
Current section
Files
deps/rocksdb/table/get_context.cc
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
#include "table/get_context.h"
#include "db/merge_helper.h"
#include "db/pinned_iterators_manager.h"
#include "db/read_callback.h"
#include "monitoring/file_read_sample.h"
#include "monitoring/perf_context_imp.h"
#include "monitoring/statistics.h"
#include "rocksdb/env.h"
#include "rocksdb/merge_operator.h"
#include "rocksdb/statistics.h"
namespace rocksdb {
namespace {
void appendToReplayLog(std::string* replay_log, ValueType type, Slice value) {
#ifndef ROCKSDB_LITE
if (replay_log) {
if (replay_log->empty()) {
// Optimization: in the common case of only one operation in the
// log, we allocate the exact amount of space needed.
replay_log->reserve(1 + VarintLength(value.size()) + value.size());
}
replay_log->push_back(type);
PutLengthPrefixedSlice(replay_log, value);
}
#else
(void)replay_log;
(void)type;
(void)value;
#endif // ROCKSDB_LITE
}
} // namespace
GetContext::GetContext(const Comparator* ucmp,
const MergeOperator* merge_operator, Logger* logger,
Statistics* statistics, GetState init_state,
const Slice& user_key, PinnableSlice* pinnable_val,
bool* value_found, MergeContext* merge_context,
RangeDelAggregator* _range_del_agg, Env* env,
SequenceNumber* seq,
PinnedIteratorsManager* _pinned_iters_mgr,
ReadCallback* callback, bool* is_blob_index)
: ucmp_(ucmp),
merge_operator_(merge_operator),
logger_(logger),
statistics_(statistics),
state_(init_state),
user_key_(user_key),
pinnable_val_(pinnable_val),
value_found_(value_found),
merge_context_(merge_context),
range_del_agg_(_range_del_agg),
env_(env),
seq_(seq),
replay_log_(nullptr),
pinned_iters_mgr_(_pinned_iters_mgr),
callback_(callback),
is_blob_index_(is_blob_index) {
if (seq_) {
*seq_ = kMaxSequenceNumber;
}
sample_ = should_sample_file_read();
}
// Called from TableCache::Get and Table::Get when file/block in which
// key may exist are not there in TableCache/BlockCache respectively. In this
// case we can't guarantee that key does not exist and are not permitted to do
// IO to be certain.Set the status=kFound and value_found=false to let the
// caller know that key may exist but is not there in memory
void GetContext::MarkKeyMayExist() {
state_ = kFound;
if (value_found_ != nullptr) {
*value_found_ = false;
}
}
void GetContext::SaveValue(const Slice& value, SequenceNumber /*seq*/) {
assert(state_ == kNotFound);
appendToReplayLog(replay_log_, kTypeValue, value);
state_ = kFound;
if (LIKELY(pinnable_val_ != nullptr)) {
pinnable_val_->PinSelf(value);
}
}
void GetContext::RecordCounters(Tickers ticker, size_t val) {
if (ticker == Tickers::TICKER_ENUM_MAX) {
return;
}
tickers_value[ticker] += static_cast<uint64_t>(val);
}
bool GetContext::SaveValue(const ParsedInternalKey& parsed_key,
const Slice& value, bool* matched,
Cleanable* value_pinner) {
assert(matched);
assert((state_ != kMerge && parsed_key.type != kTypeMerge) ||
merge_context_ != nullptr);
if (ucmp_->Equal(parsed_key.user_key, user_key_)) {
*matched = true;
// If the value is not in the snapshot, skip it
if (!CheckCallback(parsed_key.sequence)) {
return true; // to continue to the next seq
}
appendToReplayLog(replay_log_, parsed_key.type, value);
if (seq_ != nullptr) {
// Set the sequence number if it is uninitialized
if (*seq_ == kMaxSequenceNumber) {
*seq_ = parsed_key.sequence;
}
}
auto type = parsed_key.type;
// Key matches. Process it
if ((type == kTypeValue || type == kTypeMerge || type == kTypeBlobIndex) &&
range_del_agg_ != nullptr && range_del_agg_->ShouldDelete(parsed_key)) {
type = kTypeRangeDeletion;
}
switch (type) {
case kTypeValue:
case kTypeBlobIndex:
assert(state_ == kNotFound || state_ == kMerge);
if (type == kTypeBlobIndex && is_blob_index_ == nullptr) {
// Blob value not supported. Stop.
state_ = kBlobIndex;
return false;
}
if (kNotFound == state_) {
state_ = kFound;
if (LIKELY(pinnable_val_ != nullptr)) {
if (LIKELY(value_pinner != nullptr)) {
// If the backing resources for the value are provided, pin them
pinnable_val_->PinSlice(value, value_pinner);
} else {
// Otherwise copy the value
pinnable_val_->PinSelf(value);
}
}
} else if (kMerge == state_) {
assert(merge_operator_ != nullptr);
state_ = kFound;
if (LIKELY(pinnable_val_ != nullptr)) {
Status merge_status = MergeHelper::TimedFullMerge(
merge_operator_, user_key_, &value,
merge_context_->GetOperands(), pinnable_val_->GetSelf(),
logger_, statistics_, env_);
pinnable_val_->PinSelf();
if (!merge_status.ok()) {
state_ = kCorrupt;
}
}
}
if (is_blob_index_ != nullptr) {
*is_blob_index_ = (type == kTypeBlobIndex);
}
return false;
case kTypeDeletion:
case kTypeSingleDeletion:
case kTypeRangeDeletion:
// TODO(noetzli): Verify correctness once merge of single-deletes
// is supported
assert(state_ == kNotFound || state_ == kMerge);
if (kNotFound == state_) {
state_ = kDeleted;
} else if (kMerge == state_) {
state_ = kFound;
if (LIKELY(pinnable_val_ != nullptr)) {
Status merge_status = MergeHelper::TimedFullMerge(
merge_operator_, user_key_, nullptr,
merge_context_->GetOperands(), pinnable_val_->GetSelf(),
logger_, statistics_, env_);
pinnable_val_->PinSelf();
if (!merge_status.ok()) {
state_ = kCorrupt;
}
}
}
return false;
case kTypeMerge:
assert(state_ == kNotFound || state_ == kMerge);
state_ = kMerge;
// value_pinner is not set from plain_table_reader.cc for example.
if (pinned_iters_mgr() && pinned_iters_mgr()->PinningEnabled() &&
value_pinner != nullptr) {
value_pinner->DelegateCleanupsTo(pinned_iters_mgr());
merge_context_->PushOperand(value, true /*value_pinned*/);
} else {
merge_context_->PushOperand(value, false);
}
if (merge_operator_ != nullptr &&
merge_operator_->ShouldMerge(merge_context_->GetOperandsDirectionBackward())) {
state_ = kFound;
if (LIKELY(pinnable_val_ != nullptr)) {
Status merge_status = MergeHelper::TimedFullMerge(
merge_operator_, user_key_, nullptr,
merge_context_->GetOperands(), pinnable_val_->GetSelf(),
logger_, statistics_, env_);
pinnable_val_->PinSelf();
if (!merge_status.ok()) {
state_ = kCorrupt;
}
}
return false;
}
return true;
default:
assert(false);
break;
}
}
// state_ could be Corrupt, merge or notfound
return false;
}
void replayGetContextLog(const Slice& replay_log, const Slice& user_key,
GetContext* get_context, Cleanable* value_pinner) {
#ifndef ROCKSDB_LITE
Slice s = replay_log;
while (s.size()) {
auto type = static_cast<ValueType>(*s.data());
s.remove_prefix(1);
Slice value;
bool ret = GetLengthPrefixedSlice(&s, &value);
assert(ret);
(void)ret;
bool dont_care __attribute__((__unused__));
// Since SequenceNumber is not stored and unknown, we will use
// kMaxSequenceNumber.
get_context->SaveValue(
ParsedInternalKey(user_key, kMaxSequenceNumber, type), value,
&dont_care, value_pinner);
}
#else // ROCKSDB_LITE
(void)replay_log;
(void)user_key;
(void)get_context;
(void)value_pinner;
assert(false);
#endif // ROCKSDB_LITE
}
} // namespace rocksdb