Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Buffered serialization #2225

Merged
merged 1 commit into from
Oct 18, 2023
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion src/catalog/catalog.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,6 @@
#include "catalog/node_table_schema.h"
#include "catalog/rel_table_group_schema.h"
#include "catalog/rel_table_schema.h"
#include "common/ser_deser.h"
#include "storage/wal/wal.h"
#include "transaction/transaction_action.h"

Expand Down
48 changes: 25 additions & 23 deletions src/catalog/catalog_content.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,9 @@
#include "catalog/rel_table_schema.h"
#include "common/exception/catalog.h"
#include "common/exception/runtime.h"
#include "common/ser_deser.h"
#include "common/serializer/buffered_file.h"
#include "common/serializer/deserializer.h"
#include "common/serializer/serializer.h"
#include "common/string_format.h"
#include "common/string_utils.h"
#include "storage/storage_utils.h"
Expand Down Expand Up @@ -154,39 +156,39 @@ void CatalogContent::renameTable(table_id_t tableID, const std::string& newName)

void CatalogContent::saveToFile(const std::string& directory, DBFileType dbFileType) {
auto catalogPath = StorageUtils::getCatalogFilePath(directory, dbFileType);
auto fileInfo = FileUtils::openFile(catalogPath, O_WRONLY | O_CREAT);
uint64_t offset = 0;
writeMagicBytes(fileInfo.get(), offset);
SerDeser::serializeValue(StorageVersionInfo::getStorageVersion(), fileInfo.get(), offset);
SerDeser::serializeValue(tableSchemas.size(), fileInfo.get(), offset);
Serializer serializer(
std::make_unique<BufferedFileWriter>(FileUtils::openFile(catalogPath, O_WRONLY | O_CREAT)));
writeMagicBytes(serializer);
serializer.serializeValue(StorageVersionInfo::getStorageVersion());
serializer.serializeValue(tableSchemas.size());
for (auto& [tableID, tableSchema] : tableSchemas) {
SerDeser::serializeValue(tableID, fileInfo.get(), offset);
tableSchema->serialize(fileInfo.get(), offset);
serializer.serializeValue(tableID);
tableSchema->serialize(serializer);
}
SerDeser::serializeValue(nextTableID, fileInfo.get(), offset);
SerDeser::serializeUnorderedMap(macros, fileInfo.get(), offset);
serializer.serializeValue(nextTableID);
serializer.serializeUnorderedMap(macros);
}

void CatalogContent::readFromFile(const std::string& directory, DBFileType dbFileType) {
auto catalogPath = StorageUtils::getCatalogFilePath(directory, dbFileType);
auto fileInfo = FileUtils::openFile(catalogPath, O_RDONLY);
uint64_t offset = 0;
validateMagicBytes(fileInfo.get(), offset);
Deserializer deserializer(
std::make_unique<BufferedFileReader>(FileUtils::openFile(catalogPath, O_RDONLY)));
validateMagicBytes(deserializer);
storage_version_t savedStorageVersion;
SerDeser::deserializeValue(savedStorageVersion, fileInfo.get(), offset);
deserializer.deserializeValue(savedStorageVersion);
validateStorageVersion(savedStorageVersion);
uint64_t numTables;
SerDeser::deserializeValue(numTables, fileInfo.get(), offset);
deserializer.deserializeValue(numTables);
table_id_t tableID;
for (auto i = 0u; i < numTables; i++) {
SerDeser::deserializeValue(tableID, fileInfo.get(), offset);
tableSchemas[tableID] = TableSchema::deserialize(fileInfo.get(), offset);
deserializer.deserializeValue(tableID);
tableSchemas[tableID] = TableSchema::deserialize(deserializer);
}
for (auto& [tableID_, tableSchema] : tableSchemas) {
tableNameToIDMap[tableSchema->tableName] = tableID_;
}
SerDeser::deserializeValue(nextTableID, fileInfo.get(), offset);
SerDeser::deserializeUnorderedMap(macros, fileInfo.get(), offset);
deserializer.deserializeValue(nextTableID);
deserializer.deserializeUnorderedMap(macros);
}

ExpressionType CatalogContent::getFunctionType(const std::string& name) const {
Expand Down Expand Up @@ -239,22 +241,22 @@ void CatalogContent::validateStorageVersion(storage_version_t savedStorageVersio
}
}

void CatalogContent::validateMagicBytes(FileInfo* fileInfo, offset_t& offset) {
void CatalogContent::validateMagicBytes(Deserializer& deserializer) {
auto numMagicBytes = strlen(StorageVersionInfo::MAGIC_BYTES);
uint8_t magicBytes[4];
for (auto i = 0u; i < numMagicBytes; i++) {
SerDeser::deserializeValue<uint8_t>(magicBytes[i], fileInfo, offset);
deserializer.deserializeValue<uint8_t>(magicBytes[i]);
}
if (memcmp(magicBytes, StorageVersionInfo::MAGIC_BYTES, numMagicBytes) != 0) {
throw RuntimeException(
"This is not a valid Kuzu database directory for the current version of Kuzu.");
}
}

void CatalogContent::writeMagicBytes(FileInfo* fileInfo, offset_t& offset) {
void CatalogContent::writeMagicBytes(Serializer& serializer) {
auto numMagicBytes = strlen(StorageVersionInfo::MAGIC_BYTES);
for (auto i = 0u; i < numMagicBytes; i++) {
SerDeser::serializeValue<uint8_t>(StorageVersionInfo::MAGIC_BYTES[i], fileInfo, offset);
serializer.serializeValue<uint8_t>(StorageVersionInfo::MAGIC_BYTES[i]);
}
}

Expand Down
20 changes: 10 additions & 10 deletions src/catalog/node_table_schema.cpp
Original file line number Diff line number Diff line change
@@ -1,26 +1,26 @@
#include "catalog/node_table_schema.h"

#include "common/ser_deser.h"
#include "common/serializer/deserializer.h"
#include "common/serializer/serializer.h"

using namespace kuzu::common;

namespace kuzu {
namespace catalog {

void NodeTableSchema::serializeInternal(FileInfo* fileInfo, uint64_t& offset) {
SerDeser::serializeValue(primaryKeyPropertyID, fileInfo, offset);
SerDeser::serializeUnorderedSet(fwdRelTableIDSet, fileInfo, offset);
SerDeser::serializeUnorderedSet(bwdRelTableIDSet, fileInfo, offset);
void NodeTableSchema::serializeInternal(Serializer& serializer) {
serializer.serializeValue(primaryKeyPropertyID);
serializer.serializeUnorderedSet(fwdRelTableIDSet);
serializer.serializeUnorderedSet(bwdRelTableIDSet);
}

std::unique_ptr<NodeTableSchema> NodeTableSchema::deserialize(
FileInfo* fileInfo, uint64_t& offset) {
std::unique_ptr<NodeTableSchema> NodeTableSchema::deserialize(Deserializer& deserializer) {
property_id_t primaryKeyPropertyID;
std::unordered_set<table_id_t> fwdRelTableIDSet;
std::unordered_set<table_id_t> bwdRelTableIDSet;
SerDeser::deserializeValue(primaryKeyPropertyID, fileInfo, offset);
SerDeser::deserializeUnorderedSet(fwdRelTableIDSet, fileInfo, offset);
SerDeser::deserializeUnorderedSet(bwdRelTableIDSet, fileInfo, offset);
deserializer.deserializeValue(primaryKeyPropertyID);
deserializer.deserializeUnorderedSet(fwdRelTableIDSet);
deserializer.deserializeUnorderedSet(bwdRelTableIDSet);
return std::make_unique<NodeTableSchema>(
primaryKeyPropertyID, fwdRelTableIDSet, bwdRelTableIDSet);
}
Expand Down
23 changes: 12 additions & 11 deletions src/catalog/property.cpp
Original file line number Diff line number Diff line change
@@ -1,27 +1,28 @@
#include "catalog/property.h"

#include "common/ser_deser.h"
#include "common/serializer/deserializer.h"
#include "common/serializer/serializer.h"

using namespace kuzu::common;

namespace kuzu {
namespace catalog {

void Property::serialize(FileInfo* fileInfo, uint64_t& offset) const {
SerDeser::serializeValue(name, fileInfo, offset);
dataType->serialize(fileInfo, offset);
SerDeser::serializeValue(propertyID, fileInfo, offset);
SerDeser::serializeValue(tableID, fileInfo, offset);
void Property::serialize(Serializer& serializer) const {
serializer.serializeValue(name);
dataType->serialize(serializer);
serializer.serializeValue(propertyID);
serializer.serializeValue(tableID);
}

std::unique_ptr<Property> Property::deserialize(FileInfo* fileInfo, uint64_t& offset) {
std::unique_ptr<Property> Property::deserialize(Deserializer& deserializer) {
std::string name;
property_id_t propertyID;
table_id_t tableID;
SerDeser::deserializeValue(name, fileInfo, offset);
auto dataType = LogicalType::deserialize(fileInfo, offset);
SerDeser::deserializeValue(propertyID, fileInfo, offset);
SerDeser::deserializeValue(tableID, fileInfo, offset);
deserializer.deserializeValue(name);
auto dataType = LogicalType::deserialize(deserializer);
deserializer.deserializeValue(propertyID);
deserializer.deserializeValue(tableID);
return std::make_unique<Property>(name, std::move(dataType), propertyID, tableID);
}

Expand Down
15 changes: 8 additions & 7 deletions src/catalog/rdf_graph_schema.cpp
Original file line number Diff line number Diff line change
@@ -1,22 +1,23 @@
#include "catalog/rdf_graph_schema.h"

#include "common/ser_deser.h"
#include "common/serializer/deserializer.h"
#include "common/serializer/serializer.h"

using namespace kuzu::common;

namespace kuzu {
namespace catalog {

void RdfGraphSchema::serializeInternal(FileInfo* fileInfo, uint64_t& offset) {
SerDeser::serializeValue(nodeTableID, fileInfo, offset);
SerDeser::serializeValue(relTableID, fileInfo, offset);
void RdfGraphSchema::serializeInternal(Serializer& serializer) {
serializer.serializeValue(nodeTableID);
serializer.serializeValue(relTableID);
}

std::unique_ptr<RdfGraphSchema> RdfGraphSchema::deserialize(FileInfo* fileInfo, uint64_t& offset) {
std::unique_ptr<RdfGraphSchema> RdfGraphSchema::deserialize(Deserializer& deserializer) {
table_id_t nodeTableID;
table_id_t relTableID;
SerDeser::deserializeValue(nodeTableID, fileInfo, offset);
SerDeser::deserializeValue(relTableID, fileInfo, offset);
deserializer.deserializeValue(nodeTableID);
deserializer.deserializeValue(relTableID);
return std::make_unique<RdfGraphSchema>(nodeTableID, relTableID);
}

Expand Down
12 changes: 6 additions & 6 deletions src/catalog/rel_table_group_schema.cpp
Original file line number Diff line number Diff line change
@@ -1,20 +1,20 @@
#include "catalog/rel_table_group_schema.h"

#include "common/ser_deser.h"
#include "common/serializer/deserializer.h"
#include "common/serializer/serializer.h"

using namespace kuzu::common;

namespace kuzu {
namespace catalog {

void RelTableGroupSchema::serializeInternal(FileInfo* fileInfo, uint64_t& offset) {
SerDeser::serializeVector(relTableIDs, fileInfo, offset);
void RelTableGroupSchema::serializeInternal(Serializer& serializer) {
serializer.serializeVector(relTableIDs);
}

std::unique_ptr<RelTableGroupSchema> RelTableGroupSchema::deserialize(
FileInfo* fileInfo, uint64_t& offset) {
std::unique_ptr<RelTableGroupSchema> RelTableGroupSchema::deserialize(Deserializer& deserializer) {
std::vector<table_id_t> relTableIDs;
SerDeser::deserializeVector(relTableIDs, fileInfo, offset);
deserializer.deserializeVector(relTableIDs);
return std::make_unique<RelTableGroupSchema>(std::move(relTableIDs));
}

Expand Down
19 changes: 10 additions & 9 deletions src/catalog/rel_table_schema.cpp
Original file line number Diff line number Diff line change
@@ -1,7 +1,8 @@
#include "catalog/rel_table_schema.h"

#include "common/exception/catalog.h"
#include "common/ser_deser.h"
#include "common/serializer/deserializer.h"
#include "common/serializer/serializer.h"

using namespace kuzu::common;

Expand Down Expand Up @@ -40,19 +41,19 @@ std::string getRelMultiplicityAsString(RelMultiplicity relMultiplicity) {
}
}

void RelTableSchema::serializeInternal(FileInfo* fileInfo, uint64_t& offset) {
SerDeser::serializeValue(relMultiplicity, fileInfo, offset);
SerDeser::serializeValue(srcTableID, fileInfo, offset);
SerDeser::serializeValue(dstTableID, fileInfo, offset);
void RelTableSchema::serializeInternal(Serializer& serializer) {
serializer.serializeValue(relMultiplicity);
serializer.serializeValue(srcTableID);
serializer.serializeValue(dstTableID);
}

std::unique_ptr<RelTableSchema> RelTableSchema::deserialize(FileInfo* fileInfo, uint64_t& offset) {
std::unique_ptr<RelTableSchema> RelTableSchema::deserialize(Deserializer& deserializer) {
RelMultiplicity relMultiplicity;
table_id_t srcTableID;
table_id_t dstTableID;
SerDeser::deserializeValue(relMultiplicity, fileInfo, offset);
SerDeser::deserializeValue(srcTableID, fileInfo, offset);
SerDeser::deserializeValue(dstTableID, fileInfo, offset);
deserializer.deserializeValue(relMultiplicity);
deserializer.deserializeValue(srcTableID);
deserializer.deserializeValue(dstTableID);
return std::make_unique<RelTableSchema>(relMultiplicity, srcTableID, dstTableID);
}

Expand Down
41 changes: 21 additions & 20 deletions src/catalog/table_schema.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,8 @@
#include "common/exception/internal.h"
#include "common/exception/not_implemented.h"
#include "common/exception/runtime.h"
#include "common/ser_deser.h"
#include "common/serializer/deserializer.h"
#include "common/serializer/serializer.h"
#include "common/string_format.h"
#include "common/string_utils.h"

Expand Down Expand Up @@ -76,42 +77,42 @@ void TableSchema::renameProperty(property_id_t propertyID, const std::string& ne
// LCOV_EXCL_END
}

void TableSchema::serialize(FileInfo* fileInfo, uint64_t& offset) {
SerDeser::serializeValue(tableName, fileInfo, offset);
SerDeser::serializeValue(tableID, fileInfo, offset);
SerDeser::serializeValue(tableType, fileInfo, offset);
SerDeser::serializeVectorOfPtrs(properties, fileInfo, offset);
SerDeser::serializeValue(comment, fileInfo, offset);
SerDeser::serializeValue(nextPropertyID, fileInfo, offset);
serializeInternal(fileInfo, offset);
void TableSchema::serialize(Serializer& serializer) {
serializer.serializeValue(tableName);
serializer.serializeValue(tableID);
serializer.serializeValue(tableType);
serializer.serializeVectorOfPtrs(properties);
serializer.serializeValue(comment);
serializer.serializeValue(nextPropertyID);
serializeInternal(serializer);
}

std::unique_ptr<TableSchema> TableSchema::deserialize(FileInfo* fileInfo, uint64_t& offset) {
std::unique_ptr<TableSchema> TableSchema::deserialize(Deserializer& deserializer) {
std::string tableName;
table_id_t tableID;
TableType tableType;
std::vector<std::unique_ptr<Property>> properties;
std::string comment;
property_id_t nextPropertyID;
SerDeser::deserializeValue(tableName, fileInfo, offset);
SerDeser::deserializeValue(tableID, fileInfo, offset);
SerDeser::deserializeValue(tableType, fileInfo, offset);
SerDeser::deserializeVectorOfPtrs(properties, fileInfo, offset);
SerDeser::deserializeValue(comment, fileInfo, offset);
SerDeser::deserializeValue(nextPropertyID, fileInfo, offset);
deserializer.deserializeValue(tableName);
deserializer.deserializeValue(tableID);
deserializer.deserializeValue(tableType);
deserializer.deserializeVectorOfPtrs(properties);
deserializer.deserializeValue(comment);
deserializer.deserializeValue(nextPropertyID);
std::unique_ptr<TableSchema> result;
switch (tableType) {
case TableType::NODE: {
result = NodeTableSchema::deserialize(fileInfo, offset);
result = NodeTableSchema::deserialize(deserializer);
} break;
case TableType::REL: {
result = RelTableSchema::deserialize(fileInfo, offset);
result = RelTableSchema::deserialize(deserializer);
} break;
case TableType::REL_GROUP: {
result = RelTableGroupSchema::deserialize(fileInfo, offset);
result = RelTableGroupSchema::deserialize(deserializer);
} break;
case TableType::RDF: {
result = RdfGraphSchema::deserialize(fileInfo, offset);
result = RdfGraphSchema::deserialize(deserializer);
} break;
default: {
// LCOV_EXCL_START
Expand Down
4 changes: 2 additions & 2 deletions src/common/CMakeLists.txt
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,7 @@ add_subdirectory(arrow)
add_subdirectory(copier_config)
add_subdirectory(data_chunk)
add_subdirectory(exception)
add_subdirectory(serializer)
add_subdirectory(task_system)
add_subdirectory(types)
add_subdirectory(vector)
Expand All @@ -20,8 +21,7 @@ add_library(kuzu_common
type_utils.cpp
utils.cpp
string_utils.cpp
table_type.cpp
ser_deser.cpp)
table_type.cpp)

target_link_libraries(kuzu_common Glob)

Expand Down
27 changes: 0 additions & 27 deletions src/common/ser_deser.cpp

This file was deleted.

Loading