tursom/memgraph
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Separate text index header from implementation

Browse Source
This commit is contained in:
Ante Pušić 2024-01-14 22:53:36 +01:00
parent 29916a9db2
commit 6fb4eb6099
5 changed files with 112 additions and 62 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/storage/v2/CMakeLists.txt
View File

@ -20,6 +20,7 @@ add_library(mg-storage-v2 STATIC
vertex_info_cache.cpp
storage.cpp
indices/indices.cpp
indices/text_index.cpp
all_vertices_iterable.cpp
vertices_iterable.cpp
inmemory/storage.cpp
@ -42,4 +43,4 @@ add_library(mg-storage-v2 STATIC
inmemory/replication/recovery.cpp
)
target_link_libraries(mg-storage-v2 mg::replication Threads::Threads mg-utils gflags absl::flat_hash_map mg-rpc mg-slk mg-events mg-memory)
target_link_libraries(mg-storage-v2 mg::replication Threads::Threads mg-utils mg-flags gflags absl::flat_hash_map mg-rpc mg-slk mg-events mg-memory mgcxx_text_search tantivy_text_search)

81
src/storage/v2/indices/text_index.cpp Normal file
View File

@ -0,0 +1,81 @@
// Copyright 2024 Memgraph Ltd.
//
// Use of this software is governed by the Business Source License
// included in the file licenses/BSL.txt; by using this file, you agree to be bound by the terms of the Business Source
// License, and you may not use this file except in compliance with the Business Source License.
//
// As of the Change Date specified in that file, in accordance with
// the Business Source License, use of this software will be governed
// by the Apache License, Version 2.0, included in the file
// licenses/APL.txt.
#include "storage/v2/indices/text_index.hpp"
#include "query/db_accessor.hpp"
#include "storage/v2/mgcxx_mock.hpp"
namespace memgraph::storage {
void TextIndex::UpdateOnAddLabel(LabelId added_label, Vertex *vertex_after_update, const Transaction &tx) const {}
void TextIndex::UpdateOnRemoveLabel(LabelId removed_label, Vertex *vertex_after_update, const Transaction &tx) const {}
void TextIndex::UpdateOnSetProperty(PropertyId property, const PropertyValue &value, Vertex *vertex,
const Transaction &tx) const {}
std::vector<mgcxx_mock::text_search::IndexContext *> TextIndex::GetApplicableTextIndices(Vertex *vertex) {
std::vector<mgcxx_mock::text_search::IndexContext *> applicable_text_indices;
for (const auto &label : vertex->labels) {
if (label_to_index_.contains(label)) {
applicable_text_indices.push_back(&index_.at(label_to_index_.at(label)));
}
}
return applicable_text_indices;
}
bool TextIndex::CreateIndex(std::string index_name, LabelId label, memgraph::query::DbAccessor *db) {
auto index_config = mgcxx_mock::text_search::IndexConfig{
.mappings = "TODO devise the mapping by reading the indexable nodes' properties"};
auto new_index = mgcxx_mock::text_search::Mock::create_index(index_name, index_config);
index_[index_name] = new_index;
label_to_index_[label] = index_name;
return true;
// TODO add documents (indexable nodes) to index
}
bool TextIndex::DropIndex(std::string index_name) {
memcxx::text_search::drop_index(index_name);
mgcxx_mock::text_search::Mock::drop_index(index_name);
index_.erase(index_name);
std::erase_if(label_to_index_, [index_name](const auto &item) { return item.second == index_name; });
return true;
}
bool TextIndex::IndexExists(std::string index_name) const { return index_.contains(index_name); }
std::vector<Gid> TextIndex::Search(std::string index_name, std::string search_query) const {
auto input = mgcxx_mock::text_search::SearchInput{.search_query = search_query, .return_fields = {"metadata.gid"}};
// Basic check for search fields in the query (Tantivy syntax delimits them with a `:` to the right)I
if (search_query.find(":") == std::string::npos) {
input.search_fields = {"data"};
}
std::vector<Gid> found_nodes;
for (const auto &doc : mgcxx_mock::text_search::Mock::search(index_.at(index_name), input).docs) {
found_nodes.push_back(storage::Gid::FromString(doc.data));
}
return found_nodes;
}
std::vector<std::string> TextIndex::ListIndices() const {
std::vector<std::string> ret;
ret.reserve(index_.size());
for (const auto &item : index_) {
ret.push_back(item.first);
}
return ret;
}
uint64_t TextIndex::ApproximateVertexCount(std::string index_name) const { return 10; }
} // namespace memgraph::storage

80
src/storage/v2/indices/text_index.hpp
View File

@ -9,8 +9,17 @@
// by the Apache License, Version 2.0, included in the file
// licenses/APL.txt.
#include "query/db_accessor.hpp"
#pragma once
#include "storage/v2/id_types.hpp"
#include "storage/v2/mgcxx_mock.hpp"
#include "storage/v2/transaction.hpp"
#include "storage/v2/vertex.hpp"
#include "text_search.hpp"
namespace memgraph::query {
class DbAccessor;
}
namespace memgraph::storage {
@ -25,70 +34,29 @@ class TextIndex {
~TextIndex() = default;
void UpdateOnAddLabel(LabelId added_label, Vertex *vertex_after_update, const Transaction &tx) {}
std::map<std::string, mgcxx_mock::text_search::IndexContext> index_;
std::map<LabelId, std::string> label_to_index_;
void UpdateOnRemoveLabel(LabelId removed_label, Vertex *vertex_after_update, const Transaction &tx) {}
void UpdateOnAddLabel(LabelId added_label, Vertex *vertex_after_update, const Transaction &tx) const;
void UpdateOnSetProperty(PropertyId property, const PropertyValue &value, Vertex *vertex, const Transaction &tx) {}
void UpdateOnRemoveLabel(LabelId removed_label, Vertex *vertex_after_update, const Transaction &tx) const;
std::vector<mgcxx_mock::text_search::IndexContext *> GetApplicableTextIndices(Vertex *vertex, Storage *storage) {
std::vector<mgcxx_mock::text_search::IndexContext *> applicable_text_indices;
for (const auto &label : vertex->labels) {
if (label_to_index_.contains(label)) {
applicable_text_indices.push_back(&index_.at(label_to_index_.at(label)));
}
}
return applicable_text_indices;
}
void UpdateOnSetProperty(PropertyId property, const PropertyValue &value, Vertex *vertex,
const Transaction &tx) const;
bool CreateIndex(std::string index_name, LabelId label, memgraph::query::DbAccessor *db) {
auto index_config = mgcxx_mock::text_search::IndexConfig{
.mappings = "TODO devise the mapping by reading the indexable nodes' properties"};
auto new_index = mgcxx_mock::text_search::Mock::create_index(index_name, index_config);
index_[index_name] = new_index;
label_to_index_[label] = index_name;
return true;
std::vector<mgcxx_mock::text_search::IndexContext *> GetApplicableTextIndices(Vertex *vertex);
// TODO add documents (indexable nodes) to index
}
bool CreateIndex(std::string index_name, LabelId label, memgraph::query::DbAccessor *db);
bool DropIndex(std::string index_name) {
mgcxx_mock::text_search::Mock::drop_index(index_name);
index_.erase(index_name);
std::erase_if(label_to_index_, [index_name](const auto &item) { return item.second == index_name; });
return true;
}
bool DropIndex(std::string index_name);
bool IndexExists(std::string index_name) { return index_.contains(index_name); }
bool IndexExists(std::string index_name) const;
std::vector<Gid> Search(std::string index_name, std::string search_query) {
auto input = mgcxx_mock::text_search::SearchInput {
.search_query = search_query, .return_fields = {{"metadata.gid"}};
// Basic check for search fields in the query (Tantivy syntax delimits them with a `:` to the right)
if (search_query.find(":") == std::string::npos) {
input.search_fields = {"data"};
}
std::vector<Gid> Search(std::string index_name, std::string search_query) const;
std::vector<Gid> found_nodes;
for (const auto &doc : mgcxx_mock::text_search::Mock::search(index_.at(index_name), input).docs) {
found_nodes.push_back(storage::Gid::FromString(doc.data));
}
return found_nodes;
}
std::vector<std::string> ListIndices() const;
std::vector<std::string> ListIndices() {
std::vector<std::string> ret;
ret.reserve(index_.size());
for (const auto &item : index_) {
ret.push_back(item.first);
}
return ret;
}
uint64_t ApproximateVertexCount(std::string index_name) { return 10; }
std::map<std::string, mgcxx_mock::text_search::IndexContext> index_;
std::map<LabelId, std::string> label_to_index_;
};
uint64_t ApproximateVertexCount(std::string index_name) const;
};
} // namespace memgraph::storage

2
src/storage/v2/storage.hpp
View File

@ -269,7 +269,7 @@ class Storage {
virtual utils::BasicResult<StorageIndexDefinitionError, void> DropIndex(LabelId label, PropertyId property) = 0;
virtual utils::BasicResult<StorageIndexDefinitionError, void> CreateTextIndex(std::string index_name, LabelId label,
DbAccessor *db) {
query::DbAccessor *db) {
storage_->indices_.text_index_->CreateIndex(index_name, label, db);
return {};
}

8
src/storage/v2/vertex_accessor.cpp
View File

@ -277,7 +277,7 @@ Result<PropertyValue> VertexAccessor::SetProperty(PropertyId property, const Pro
[transaction = transaction_, storage = storage_, vertex = vertex_, &value, &property, &current_value]() {
CreateAndLinkDelta(transaction, vertex, Delta::SetPropertyTag(), property, current_value);
if (flags::run_time::GetTextSearchEnabled()) {
for (const auto *index_context : storage->indices_.text_index_->GetApplicableTextIndices(vertex, storage)) {
for (const auto *index_context : storage->indices_.text_index_->GetApplicableTextIndices(vertex)) {
auto search_input = mgcxx_mock::text_search::SearchInput{
.search_query = fmt::format("metadata.gid:{}", vertex->gid.AsInt()), .return_fields = {"data"}};
@ -323,7 +323,7 @@ Result<bool> VertexAccessor::InitProperties(const std::map<storage::PropertyId,
return;
}
if (flags::run_time::GetTextSearchEnabled()) {
for (const auto *index_context : storage->indices_.text_index_->GetApplicableTextIndices(vertex, storage)) {
for (const auto *index_context : storage->indices_.text_index_->GetApplicableTextIndices(vertex)) {
auto new_properties_document =
mgcxx_mock::text_search::DocumentInput{}; // TODO (pending real Tantivy operation): create a JSON, set
// properties and convert to string
@ -366,7 +366,7 @@ Result<std::vector<std::tuple<PropertyId, PropertyValue, PropertyValue>>> Vertex
[storage = storage_, transaction = transaction_, vertex = vertex_, &properties, &id_old_new_change]() {
id_old_new_change.emplace(vertex->properties.UpdateProperties(properties));
if (flags::run_time::GetTextSearchEnabled()) {
for (const auto *index_context : storage->indices_.text_index_->GetApplicableTextIndices(vertex, storage)) {
for (const auto *index_context : storage->indices_.text_index_->GetApplicableTextIndices(vertex)) {
auto search_input = mgcxx_mock::text_search::SearchInput{
.search_query = fmt::format("metadata.gid:{}", vertex->gid.AsInt()), .return_fields = {"data"}};
@ -425,7 +425,7 @@ Result<std::map<PropertyId, PropertyValue>> VertexAccessor::ClearProperties() {
vertex->properties.ClearProperties();
if (flags::run_time::GetTextSearchEnabled()) {
for (const auto *index_context : storage->indices_.text_index_->GetApplicableTextIndices(vertex, storage)) {
for (const auto *index_context : storage->indices_.text_index_->GetApplicableTextIndices(vertex)) {
auto search_input =
mgcxx_mock::text_search::SearchInput{.search_query = fmt::format("metadata.gid:{}", vertex->gid.AsInt())};
mgcxx_mock::text_search::Mock::delete_document(*index_context, search_input, true);