Compare commits

...

92 Commits
master ... E118

Author SHA1 Message Date
gvolfing
451da803dc Add another test to ExpandOne 2022-09-20 18:40:23 +02:00
gvolfing
d1f51aa3f9 Add checkf for edge types in the ExpandOne handle 2022-09-20 18:04:32 +02:00
gvolfing
aec3bc3d05 Fix ExpandOne behavior 2022-09-20 16:39:52 +02:00
gvolfing
7fe5eb2c49 Clean-up tests 2022-09-20 09:52:57 +02:00
gvolfing
26e57d1988 Add the rest of the implementation for ExpandOne handle, fix segfault issue 2022-09-19 19:58:20 +02:00
gvolfing
f066d4f0b8 Save for tomorrow 2022-09-18 13:29:31 +02:00
gvolfing
359ac507a3 Merge branch 'T0919-MG-implement-message-based-actions' into T0919-MG-save-branch-for-outcommented-messages 2022-09-15 19:22:34 +02:00
gvolfing
4530e8fe50 Make ConvertPropertyMap take its argument as rvalue& 2022-09-15 17:18:23 +02:00
gvolfing
ef205dce4d Fix clang-tidy warnings 2022-09-15 13:36:32 +02:00
gvolfing
fd1668403a Restructure EdgeId message struct 2022-09-15 11:57:57 +02:00
gvolfing
b8bca320c8 Make enum types CamelCase, remove dead comments 2022-09-15 11:34:40 +02:00
gvolfing
3140f81d61 Apply suggestions from code review
Co-authored-by: Jure Bajic <jure.bajic@memgraph.com>
2022-09-15 11:07:46 +02:00
gvolfing
59bb54c26e React to PR comments 3 2022-09-14 17:12:43 +02:00
gvolfing
c6bc9d39ab Merge branch 'E118-MG-lexicographically-ordered-storage' into T0919-MG-implement-message-based-actions 2022-09-14 16:51:28 +02:00
gvolfing
d0930f66c8 React to PR comments 2 2022-09-14 16:48:07 +02:00
gvolfing
7db0f9ee23 React to PR comments 1 2022-09-14 14:19:44 +02:00
János Benjamin Antal
8e1f83acc9 Reference vertices by id in edges (#532) 2022-09-14 13:11:14 +02:00
János Benjamin Antal
3a956d2266 Fix random hangs in simulator 2022-09-13 13:16:26 +02:00
János Benjamin Antal
9284ba0805 Small fixes 2022-09-13 11:09:32 +02:00
gvolfing
4c53713217 Remove outcommented code 2022-09-13 11:08:54 +02:00
gvolfing
ae04e717f9 Enable back the all of the simulator tests 2022-09-13 10:56:14 +02:00
gvolfing
1706be3341 Enable back the rest of the simulator tests 2022-09-13 10:53:46 +02:00
gvolfing
4d53d53859 Clean up used messages and their handles 2022-09-13 10:34:13 +02:00
gvolfing
f5d2bcd979 Cleanup 2022-09-12 18:09:28 +02:00
gvolfing
16028f50e8 Finish impl of ScanAll handle and its test, uncomment non-priority cases, general cleanup 2022-09-12 17:13:54 +02:00
János Benjamin Antal
bfdc69fdbd Fix Value special member functions 2022-09-12 14:06:30 +02:00
János Benjamin Antal
f49c962c5e Fix promise timeout handling 2022-09-12 14:06:02 +02:00
gvolfing
c71f16d9aa Save progress 2022-09-12 13:35:36 +02:00
gvolfing
2751d1fc94 Add test for DeleteEdges message handle 2022-09-11 21:04:29 +02:00
gvolfing
0d728d8a0a Test AddEdges message handle 2022-09-11 20:41:45 +02:00
gvolfing
0cfd0703ab Test UpdateVertices message handle 2022-09-11 19:56:02 +02:00
gvolfing
8ca0c0df53 Add test for DeleteVertices message handle 2022-09-11 15:54:46 +02:00
gvolfing
3feaa79688 Make Test and CreateVertices message handle work 2022-09-11 09:59:58 +02:00
gvolfing
a119fa3f73 Merge branch 'E118-MG-lexicographically-ordered-storage' into T0919-MG-implement-message-based-actions 2022-09-09 13:26:23 +02:00
gvolfing
f50d0f3a75 Add additional constructors to Value and modify test for message handlers 2022-09-09 13:20:13 +02:00
jbajic
f910cb770c Merge branch 'project-pineapples' into E118-MG-lexicographically-ordered-storage 2022-09-09 12:39:28 +02:00
gvolfing
4d650aa595 Add test file 2022-09-08 17:09:53 +02:00
gvolfing
1316a90f8a Add additional dependencies and delete unused headers and type aliases 2022-09-08 16:46:21 +02:00
gvolfing
8bb6e00317 Add files for tests 2022-09-08 14:01:34 +02:00
gvolfing
c690b349f7 Add message routing capabilites to ShardRsm 2022-09-08 13:09:21 +02:00
gvolfing
9c2ecbd0a1 Add definition for ScanVerticesHandle 2022-09-08 12:47:57 +02:00
gvolfing
5b31ddc4fa Add move ctors where needed 2022-09-08 11:09:02 +02:00
gvolfing
a308bf16a2 Uncomment parts that require dinamically allocated resources from the union 2022-09-08 07:53:03 +02:00
gvolfing
3b5a736a97 Add partial implementation of ScanAll 2022-09-07 21:57:00 +02:00
Kostas Kyrimis
38d0b89b04 Implement expression evaluator library (#486) 2022-09-07 18:15:32 +03:00
gvolfing
8ebd85b6ea Update messages 2022-09-07 14:03:10 +02:00
Jure Bajic
a2a6a3855b Remove primary label from vertex (#529)
- Remove the primary label from vertex
- Pass vertex_validator instead of schema validator
- Fix vertex ctors
2022-09-07 13:36:10 +02:00
Jure Bajic
e9f0360fb3 Enable indices for storage v3 (#528)
- Enable indices in storage v3
- Add new test cases
- Change `CreateVertexAndValidate` to call `SetProperty` and `AddLabel`
2022-09-06 22:48:25 +02:00
gvolfing
db190baeef Add handle declaration for ScanVertices request 2022-09-06 20:27:57 +02:00
gvolfing
a1ec8d19d1 TODO 2022-09-06 15:31:54 +02:00
gvolfing
c576d02974 Add definition for UpdateEdge handle, modify its message 2022-09-06 15:26:14 +02:00
gvolfing
941ef01b88 Add definition for UpdateVertexHandle 2022-09-06 15:26:14 +02:00
gvolfing
6a7b2e586c Add implementation for DeleteEdges handle 2022-09-06 15:26:14 +02:00
gvolfing
10aa6b977c Modify existing messages, add new one for updates, more cleanup 2022-09-06 15:26:14 +02:00
gvolfing
c8ca234fec General cleanup 2022-09-06 15:26:14 +02:00
gvolfing
3e05b690a9 Make handle implementations exit the loops on the first encountered error 2022-09-06 15:26:14 +02:00
gvolfing
316fbfda52 Add implementation to CreateEdge handle, and modify messages 2022-09-06 15:26:14 +02:00
gvolfing
81b1173336 Add message types related to edges and the declarations for their handles 2022-09-06 15:26:14 +02:00
gvolfing
88240a39d9 Modify DeleteVertices message type, add implementation for its handle 2022-09-06 15:26:14 +02:00
gvolfing
31581ce4aa Refactor bridging helper functions 2022-09-06 15:26:14 +02:00
gvolfing
df60eae124 Change the underlying datastructure of property map in the defined messages, and bridge it to the VertexAccessor 2022-09-06 15:26:14 +02:00
gvolfing
950338833e Add debug messages and refactor the CreateVertices handle 2022-09-06 15:26:14 +02:00
gvolfing
921dc8b740 Add other message types relating to vertices 2022-09-06 15:26:14 +02:00
gvolfing
02f63fd68d Rename class ShardMessageHandler to ShardRsm 2022-09-06 15:26:14 +02:00
gvolfing
b1feee5b08 Rename file shard_message_handler to shard_rsm 2022-09-06 15:26:14 +02:00
gvolfing
d7c89d9cd5 Add CreateVerticies skeleton 2022-09-06 15:26:14 +02:00
János Benjamin Antal
fe4955447e Merge branch 'project-pineapples' into E118-MG-lexicographically-ordered-storage 2022-09-05 14:59:01 +02:00
Tyler Neely
1631c20df2 In memory shard map (#505) 2022-09-05 15:50:54 +03:00
János Benjamin Antal
947baedbe6 Merge remote-tracking branch 'origin/project-pineapples' into E118-MG-lexicographically-ordered-storage 2022-09-02 11:41:03 +02:00
Kostas Kyrimis
c6447eb48b Add shard requests responses (#526) 2022-09-01 18:54:47 +03:00
Jure Bajic
7e84744d07 Split storage and shards (#519)
- Rename storage to shard
- Add primary label and range for shard
- Remove id_mapper functionality from shard
- Adapt tests
2022-09-01 09:10:40 +02:00
Tyler Neely
c0d03888f4 Implement basic raft version (#498) 2022-08-30 15:07:34 +02:00
János Benjamin Antal
efb3c8d03d Remove multi-threaded related logic and variables (#460)
* Remove logic that was necessary for optimal multi-threaded performance, such
  as accumulating deleted object in local containers and appending them to a
  global one, handling overlapping locking.
* Remove background GC thread.
* Remove mutexes, locks and atomics throughout storage.
2022-08-30 13:41:53 +02:00
Jure Bajic
95dbc022c0 Integrate schema and keystore (#497)
- Integrate schema and keystore on vertex creation
- Add GC test for storage v3
- Add tests for accessors
- Fix all tests related to this except for query v2
- Fix labels not returning primary label
2022-08-29 14:38:25 +02:00
Tyler Neely
a40403e3ce Add local transport (#512)
* Create LocalTransport Io provider for sending messages to components on the same machine
* Move src/io/simulation/message_conversion.hpp to src/io/message_conversion.hpp for use in other Io providers
2022-08-29 13:49:51 +02:00
Tyler Neely
14c9e68456 Transport prototype (#466) 2022-08-12 08:24:32 +02:00
jbajic
68b26275a3 Merge branch 'project-pineapples' into E118-MG-lexicographically-ordered-storage 2022-08-08 11:34:46 +02:00
jbajic
5012824e05 Address review comments 2022-08-04 11:45:16 +02:00
Jure Bajic
a12a1ea358 Move schema to storage v3 and query v2
* Move schema to storage v3

* Remove schema from v2

* Move schema to query v2

* Remove schema from query v1

* Make glue v2

* Move schema related tests to newer versions of query and storage

* Fix typo in CMake

* Fix interpreter test

* Fix clang tidy errors

* Change temp dir name
2022-08-04 09:50:02 +02:00
János Benjamin Antal
2891041468 Make storage use KeyStore (#455) 2022-08-03 18:10:58 +02:00
János Benjamin Antal
cc5ee6a496 Merge remote-tracking branch 'origin/project-pineapples' into E118-MG-lexicographically-ordered-storage 2022-08-02 08:19:43 +02:00
jbajic
f57f30c8cf Merge branch 'project-pineapples' into E112-MG-implement-partial-schema 2022-08-01 10:46:11 +02:00
Jure Bajic
462daf3a2b Enforce schema on vertex creation
- Separating schema definition from schema validation
- Updating vertex_accessor and db_accessors with necessary methods
- Adding a primary label to Vertex
- Adding schema tests
- Updating existing tests for storage v3, and deprecating old:
  - interpreter => interpreter_v2
  - query_plan_accumulate_aggregate => storage_v3_query_plan_accumulate_aggregate
  - query_plan_create_set_remove_delete => storage_v3_query_plan_create_set_remove_delete
  - query_plan_bag_semantics => storage_v3_query_plan_bag_semantics
  - query_plan_edge_cases => storage_v3_query_plan_edge_cases
  - query_plan_v2_create_set_remove_delete => storage_v3_query_plan_v2_create_set_remove_delete
  - query_plan_match_filter_return => storage_v3_query_plan_match_filter_return
2022-07-29 13:38:17 +02:00
Marko Budiselic
2009fefc8a Merge branch 'master' into project-pineapples 2022-07-28 15:36:51 +02:00
Marko Budiselić
eb3f96d1f6 Bring changes from master to project-pineapples (#477)
* Fix aggregation functions on `null` and group-by inputs (#448)
* Upgrade Antrl to 4.10.1 and remove antlr_lock (#441)
* Update clang-tidy job (#476)
* Add parser stress test (#463)

NOTE: Doing this to have buildable comments on the project-pineapples branch

Co-authored-by: gvolfing <107616712+gvolfing@users.noreply.github.com>
Co-authored-by: Jure Bajic <jure.bajic@memgraph.com>
2022-07-28 15:36:17 +02:00
jbajic
264b233053 Merge branch 'project-pineapples' into E112-MG-implement-partial-schema 2022-07-22 11:48:45 +02:00
Jure Bajic
2ceaf59767 Create query engine v2 (#444)
Create version v2 of the query engine.
Adjust CMake and lisp files
Connect query engine v2 with storage engine v3
2022-07-19 12:28:19 +02:00
János Benjamin Antal
c0bee760bf git 2022-07-18 08:21:04 +02:00
Jure Bajic
3f4f66b57f Create schema DDL expressions
* Add initial schema implementation

* Add index to schema

* List schemas and enable multiple properties

* Implement SchemaTypes

* Apply suggestions from code review

Co-authored-by: Jeremy B <97525434+42jeremy@users.noreply.github.com>
Co-authored-by: János Benjamin Antal <antaljanosbenjamin@users.noreply.github.com>

* Address review comments

* Remove Map and List

* Add schema operations in storage

* Add create and show schema queries

* Add privileges for schema

* Add missing keywords into lexer

* Add drop schema query

* Add schema visitors

* Update metadata

* Add PrepareSchemaQuery function

* Implement show schemas

* Add show schema query

* Fix schema visitor

* Add common schema type

* Fix grammar

* Temporary create ddl logic

* Fix naming for schemaproperty type to schema type

* Rename schemaproperty to schemapropertytype

* Enable Create schema ddl

* Override visitPropertyType

* Add initial schema implementation

* Add initial schema implementation

* Add index to schema

* List schemas and enable multiple properties

* Implement SchemaTypes

* Apply suggestions from code review

Co-authored-by: Jeremy B <97525434+42jeremy@users.noreply.github.com>
Co-authored-by: János Benjamin Antal <antaljanosbenjamin@users.noreply.github.com>

* Address review comments

* Remove Map and List

* Apply suggestions from code review

Co-authored-by: Kostas Kyrimis  <kostaskyrim@gmail.com>

Co-authored-by: Jeremy B <97525434+42jeremy@users.noreply.github.com>
Co-authored-by: János Benjamin Antal <antaljanosbenjamin@users.noreply.github.com>
Co-authored-by: Kostas Kyrimis  <kostaskyrim@gmail.com>

* Add verification on creation and deletion

* Rename DeleteSchema to DropSchema

* Remove list and map from lexer

* Fix grammar with schemaTypeMap

* Add privilege and cypher visitor tests

* Catch repeating type name in schema definition

* Fix conflicting keywords

* Add notifications

* Drop float support

* Finish interpreter tests

* Fix tests

* Fix clang tidy errors

* Fix GetSchema

* Replace for with transfrom

* Add cloning og schema_property_map

* Address review comments

* Rename SchemaPropertyType to SchemaType

* Remove inline

* Assert of schema properties

Co-authored-by: Jeremy B <97525434+42jeremy@users.noreply.github.com>
Co-authored-by: János Benjamin Antal <antaljanosbenjamin@users.noreply.github.com>
Co-authored-by: Kostas Kyrimis  <kostaskyrim@gmail.com>
2022-07-11 09:20:15 +02:00
Jure Bajic
2998f92595 Add initial schema implementation
* Add initial schema implementation

* Add index to schema

* List schemas and enable multiple properties

* Implement SchemaTypes

* Apply suggestions from code review

Co-authored-by: Jeremy B <97525434+42jeremy@users.noreply.github.com>
Co-authored-by: János Benjamin Antal <antaljanosbenjamin@users.noreply.github.com>

* Address review comments

* Remove Map and List

* Apply suggestions from code review

Co-authored-by: Kostas Kyrimis  <kostaskyrim@gmail.com>

Co-authored-by: Jeremy B <97525434+42jeremy@users.noreply.github.com>
Co-authored-by: János Benjamin Antal <antaljanosbenjamin@users.noreply.github.com>
Co-authored-by: Kostas Kyrimis  <kostaskyrim@gmail.com>
2022-07-08 10:33:43 +02:00
János Benjamin Antal
1bdc32ba5d Copy storage v2 to create storage v3 (#416)
* Copy storage v2 to v3

* Integrate v3 to cmake

* Fix clang-tidy warnings

* Add dummy unit test for storage-v3 to trigger build for code analysis builds
2022-07-05 08:20:59 +02:00
János Benjamin Antal
21870a0e7e Merge branch 'master' into project-pineapples 2022-06-23 14:49:14 +02:00
255 changed files with 89042 additions and 40 deletions

View File

@@ -6,6 +6,7 @@ Checks: '*,
-altera-unroll-loops,
-android-*,
-cert-err58-cpp,
-cert-str34-c,
-cppcoreguidelines-avoid-c-arrays,
-cppcoreguidelines-avoid-goto,
-cppcoreguidelines-avoid-magic-numbers,
@@ -49,6 +50,7 @@ Checks: '*,
-misc-non-private-member-variables-in-classes,
-modernize-avoid-c-arrays,
-modernize-concat-nested-namespaces,
-modernize-loop-convert,
-modernize-pass-by-value,
-modernize-use-equals-default,
-modernize-use-nodiscard,

View File

@@ -130,7 +130,7 @@ jobs:
source /opt/toolchain-v4/activate
# Restrict clang-tidy results only to the modified parts
git diff -U0 ${{ env.BASE_BRANCH }}... -- src | ./tools/github/clang-tidy/clang-tidy-diff.py -p 1 -j $THREADS -path build | tee ./build/clang_tidy_output.txt
git diff -U0 ${{ env.BASE_BRANCH }}... -- src | ./tools/github/clang-tidy/clang-tidy-diff.py -p 1 -j $THREADS -extra-arg="-DMG_CLANG_TIDY_CHECK" -path build | tee ./build/clang_tidy_output.txt
# Fail if any warning is reported
! cat ./build/clang_tidy_output.txt | ./tools/github/clang-tidy/grep_error_lines.sh > /dev/null
@@ -171,7 +171,7 @@ jobs:
# Run leftover CTest tests (all except unit and benchmark tests).
cd build
ctest -E "(memgraph__unit|memgraph__benchmark)" --output-on-failure
ctest -E "(memgraph__unit|memgraph__benchmark|memgraph__simulation)" --output-on-failure
- name: Run drivers tests
run: |
@@ -262,6 +262,15 @@ jobs:
cd build
ctest -R memgraph__unit --output-on-failure -j$THREADS
- name: Run simulation tests
run: |
# Activate toolchain.
source /opt/toolchain-v4/activate
# Run unit tests.
cd build
ctest -R memgraph__simulation --output-on-failure -j$THREADS
- name: Run e2e tests
run: |
# TODO(gitbuda): Setup mgclient and pymgclient properly.

View File

@@ -39,7 +39,7 @@ jobs:
source /opt/toolchain-v4/activate
# The results are also written to standard output in order to retain them in the logs
./tools/github/clang-tidy/run-clang-tidy.py -p build -j $THREADS -clang-tidy-binary=/opt/toolchain-v4/bin/clang-tidy "$PWD/src/*" |
./tools/github/clang-tidy/run-clang-tidy.py -p build -j $THREADS -extra-arg="-DMG_CLANG_TIDY_CHECK" -clang-tidy-binary=/opt/toolchain-v4/bin/clang-tidy "$PWD/src/*" |
tee ./build/full_clang_tidy_output.txt
- name: Summarize clang-tidy results

8
.gitignore vendored
View File

@@ -23,6 +23,8 @@ cmake-build-*
cmake/DownloadProject/
dist/
src/query/frontend/opencypher/generated/
src/query/v2/frontend/opencypher/generated/
src/parser/opencypher/generated
tags
ve/
ve3/
@@ -50,15 +52,21 @@ src/distributed/pull_produce_rpc_messages.hpp
src/distributed/storage_gc_rpc_messages.hpp
src/distributed/token_sharing_rpc_messages.hpp
src/distributed/updates_rpc_messages.hpp
src/query/v2/frontend/ast/ast.hpp
src/query/frontend/ast/ast.hpp
src/query/distributed/frontend/ast/ast_serialization.hpp
src/query/v2/distributed/frontend/ast/ast_serialization.hpp
src/durability/distributed/state_delta.hpp
src/durability/single_node/state_delta.hpp
src/durability/single_node_ha/state_delta.hpp
src/query/frontend/semantic/symbol.hpp
src/query/v2/frontend/semantic/symbol.hpp
src/query/distributed/frontend/semantic/symbol_serialization.hpp
src/query/v2/distributed/frontend/semantic/symbol_serialization.hpp
src/query/distributed/plan/ops.hpp
src/query/v2/distributed/plan/ops.hpp
src/query/plan/operator.hpp
src/query/v2/plan/operator.hpp
src/raft/log_entry.hpp
src/raft/raft_rpc_messages.hpp
src/raft/snapshot_metadata.hpp

View File

@@ -5,16 +5,22 @@ add_subdirectory(lisp)
add_subdirectory(utils)
add_subdirectory(requests)
add_subdirectory(io)
add_subdirectory(io/simulator)
add_subdirectory(kvstore)
add_subdirectory(telemetry)
add_subdirectory(communication)
add_subdirectory(memory)
add_subdirectory(storage/v2)
add_subdirectory(storage/v3)
add_subdirectory(integrations)
add_subdirectory(query)
add_subdirectory(query/v2)
add_subdirectory(slk)
add_subdirectory(rpc)
add_subdirectory(auth)
add_subdirectory(parser)
add_subdirectory(expr)
add_subdirectory(coordinator)
if (MG_ENTERPRISE)
add_subdirectory(audit)

View File

@@ -37,7 +37,7 @@ const std::vector<Permission> kPermissionsAll = {
Permission::CONSTRAINT, Permission::DUMP, Permission::AUTH, Permission::REPLICATION,
Permission::DURABILITY, Permission::READ_FILE, Permission::FREE_MEMORY, Permission::TRIGGER,
Permission::CONFIG, Permission::STREAM, Permission::MODULE_READ, Permission::MODULE_WRITE,
Permission::WEBSOCKET};
Permission::WEBSOCKET, Permission::SCHEMA};
} // namespace
std::string PermissionToString(Permission permission) {
@@ -84,6 +84,8 @@ std::string PermissionToString(Permission permission) {
return "MODULE_WRITE";
case Permission::WEBSOCKET:
return "WEBSOCKET";
case Permission::SCHEMA:
return "SCHEMA";
}
}

View File

@@ -38,7 +38,8 @@ enum class Permission : uint64_t {
STREAM = 1U << 17U,
MODULE_READ = 1U << 18U,
MODULE_WRITE = 1U << 19U,
WEBSOCKET = 1U << 20U
WEBSOCKET = 1U << 20U,
SCHEMA = 1U << 21U
};
// clang-format on

19
src/common/types.hpp Normal file
View File

@@ -0,0 +1,19 @@
// Copyright 2022 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.
#pragma once
#include <cstdint>
namespace memgraph::common {
enum class SchemaType : uint8_t { BOOL, INT, STRING, DATE, LOCALTIME, LOCALDATETIME, DURATION };
} // namespace memgraph::common

View File

@@ -0,0 +1,10 @@
set(coordinator_src_files
coordinator.hpp
shard_map.hpp
hybrid_logical_clock.hpp)
find_package(fmt REQUIRED)
find_package(Threads REQUIRED)
add_library(mg-coordinator STATIC ${coordinator_src_files})
target_link_libraries(mg-coordinator stdc++fs Threads::Threads fmt::fmt mg-utils)

View File

@@ -0,0 +1,259 @@
// Copyright 2022 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.
#pragma once
#include <optional>
#include <string>
#include <unordered_set>
#include <variant>
#include <vector>
#include "coordinator/hybrid_logical_clock.hpp"
#include "coordinator/shard_map.hpp"
#include "io/simulator/simulator.hpp"
#include "io/time.hpp"
#include "io/transport.hpp"
#include "storage/v3/id_types.hpp"
#include "storage/v3/schemas.hpp"
namespace memgraph::coordinator {
using memgraph::storage::v3::LabelId;
using memgraph::storage::v3::PropertyId;
using Address = memgraph::io::Address;
using SimT = memgraph::io::simulator::SimulatorTransport;
using memgraph::storage::v3::SchemaProperty;
struct HlcRequest {
Hlc last_shard_map_version;
};
struct HlcResponse {
Hlc new_hlc;
std::optional<ShardMap> fresher_shard_map;
};
struct GetShardMapRequest {
// No state
};
struct GetShardMapResponse {
ShardMap shard_map;
};
struct AllocateHlcBatchResponse {
bool success;
Hlc low;
Hlc high;
};
struct AllocateEdgeIdBatchRequest {
size_t batch_size;
};
struct AllocateEdgeIdBatchResponse {
uint64_t low;
uint64_t high;
};
struct AllocatePropertyIdsRequest {
std::vector<std::string> property_names;
};
struct AllocatePropertyIdsResponse {
std::map<std::string, PropertyId> property_ids;
};
struct SplitShardRequest {
Hlc previous_shard_map_version;
LabelId label_id;
CompoundKey split_key;
};
struct SplitShardResponse {
bool success;
};
struct RegisterStorageEngineRequest {
Address address;
};
struct RegisterStorageEngineResponse {
bool success;
};
struct DeregisterStorageEngineRequest {
Address address;
};
struct DeregisterStorageEngineResponse {
bool success;
};
struct InitializeLabelRequest {
std::string label_name;
std::vector<SchemaProperty> schema;
Hlc last_shard_map_version;
};
struct InitializeLabelResponse {
bool success;
std::optional<ShardMap> fresher_shard_map;
};
struct HeartbeatRequest {};
struct HeartbeatResponse {};
using CoordinatorWriteRequests =
std::variant<HlcRequest, AllocateEdgeIdBatchRequest, SplitShardRequest, RegisterStorageEngineRequest,
DeregisterStorageEngineRequest, InitializeLabelRequest, AllocatePropertyIdsRequest>;
using CoordinatorWriteResponses =
std::variant<HlcResponse, AllocateEdgeIdBatchResponse, SplitShardResponse, RegisterStorageEngineResponse,
DeregisterStorageEngineResponse, InitializeLabelResponse, AllocatePropertyIdsResponse>;
using CoordinatorReadRequests = std::variant<GetShardMapRequest, HeartbeatRequest>;
using CoordinatorReadResponses = std::variant<GetShardMapResponse, HeartbeatResponse>;
class Coordinator {
public:
explicit Coordinator(ShardMap sm) : shard_map_{std::move(sm)} {}
// NOLINTNEXTLINE(readability-convert-member-functions-to-static
CoordinatorReadResponses Read(CoordinatorReadRequests requests) {
return std::visit([&](auto &&request) { return HandleRead(std::forward<decltype(request)>(request)); },
std::move(requests)); // NOLINT(hicpp-move-const-arg,performance-move-const-arg)
}
// NOLINTNEXTLINE(readability-convert-member-functions-to-static
CoordinatorWriteResponses Apply(CoordinatorWriteRequests requests) {
return std::visit([&](auto &&request) mutable { return ApplyWrite(std::forward<decltype(request)>(request)); },
std::move(requests));
}
private:
ShardMap shard_map_;
uint64_t highest_allocated_timestamp_;
/// Query engines need to periodically request batches of unique edge IDs.
uint64_t highest_allocated_edge_id_;
static CoordinatorReadResponses HandleRead(HeartbeatRequest && /* heartbeat_request */) {
return HeartbeatResponse{};
}
CoordinatorReadResponses HandleRead(GetShardMapRequest && /* get_shard_map_request */) {
GetShardMapResponse res;
res.shard_map = shard_map_;
return res;
}
CoordinatorWriteResponses ApplyWrite(HlcRequest &&hlc_request) {
HlcResponse res{};
auto hlc_shard_map = shard_map_.GetHlc();
MG_ASSERT(!(hlc_request.last_shard_map_version.logical_id > hlc_shard_map.logical_id));
res.new_hlc = Hlc{
.logical_id = ++highest_allocated_timestamp_,
// TODO(tyler) probably pass some more context to the Coordinator here
// so that we can use our wall clock and enforce monotonicity.
// .coordinator_wall_clock = io_.Now(),
};
// Allways return fresher shard_map for now.
res.fresher_shard_map = std::make_optional(shard_map_);
return res;
}
CoordinatorWriteResponses ApplyWrite(AllocateEdgeIdBatchRequest &&ahr) {
AllocateEdgeIdBatchResponse res{};
uint64_t low = highest_allocated_edge_id_;
highest_allocated_edge_id_ += ahr.batch_size;
uint64_t high = highest_allocated_edge_id_;
res.low = low;
res.high = high;
return res;
}
/// This splits the shard immediately beneath the provided
/// split key, keeping the assigned peers identical for now,
/// but letting them be gradually migrated over time.
CoordinatorWriteResponses ApplyWrite(SplitShardRequest &&split_shard_request) {
SplitShardResponse res{};
if (split_shard_request.previous_shard_map_version != shard_map_.shard_map_version) {
res.success = false;
} else {
res.success = shard_map_.SplitShard(split_shard_request.previous_shard_map_version, split_shard_request.label_id,
split_shard_request.split_key);
}
return res;
}
/// This adds the provided storage engine to the standby storage engine pool,
/// which can be used to rebalance storage over time.
static CoordinatorWriteResponses ApplyWrite(RegisterStorageEngineRequest && /* register_storage_engine_request */) {
RegisterStorageEngineResponse res{};
// TODO
return res;
}
/// This begins the process of draining the provided storage engine from all raft
/// clusters that it might be participating in.
static CoordinatorWriteResponses ApplyWrite(DeregisterStorageEngineRequest && /* register_storage_engine_request */) {
DeregisterStorageEngineResponse res{};
// TODO
// const Address &address = register_storage_engine_request.address;
// storage_engine_pool_.erase(address);
// res.success = true;
return res;
}
CoordinatorWriteResponses ApplyWrite(InitializeLabelRequest &&initialize_label_request) {
InitializeLabelResponse res{};
bool success = shard_map_.InitializeNewLabel(initialize_label_request.label_name, initialize_label_request.schema,
initialize_label_request.last_shard_map_version);
if (success) {
res.fresher_shard_map = shard_map_;
res.success = false;
} else {
res.fresher_shard_map = std::nullopt;
res.success = true;
}
return res;
}
CoordinatorWriteResponses ApplyWrite(AllocatePropertyIdsRequest &&allocate_property_ids_request) {
AllocatePropertyIdsResponse res{};
auto property_ids = shard_map_.AllocatePropertyIds(allocate_property_ids_request.property_names);
res.property_ids = property_ids;
return res;
}
};
} // namespace memgraph::coordinator

View File

@@ -0,0 +1,25 @@
// Copyright 2022 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.
#pragma once
#include "coordinator/coordinator.hpp"
#include "io/rsm/rsm_client.hpp"
namespace memgraph::coordinator {
using memgraph::io::rsm::RsmClient;
template <typename IoImpl>
using CoordinatorClient = RsmClient<IoImpl, CoordinatorWriteRequests, CoordinatorWriteResponses,
CoordinatorReadRequests, CoordinatorReadResponses>;
} // namespace memgraph::coordinator

View File

@@ -0,0 +1,23 @@
// Copyright 2022 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.
#pragma once
#include "coordinator/coordinator.hpp"
#include "io/rsm/raft.hpp"
namespace memgraph::coordinator {
template <typename IoImpl>
using CoordinatorRsm = memgraph::io::rsm::Raft<IoImpl, Coordinator, CoordinatorWriteRequests, CoordinatorWriteResponses,
CoordinatorReadRequests, CoordinatorReadResponses>;
} // namespace memgraph::coordinator

View File

@@ -0,0 +1,28 @@
// Copyright 2022 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.
#pragma once
#include "io/time.hpp"
namespace memgraph::coordinator {
using Time = memgraph::io::Time;
/// Hybrid-logical clock
struct Hlc {
uint64_t logical_id;
Time coordinator_wall_clock;
bool operator==(const Hlc &other) const = default;
};
} // namespace memgraph::coordinator

View File

@@ -0,0 +1,189 @@
// Copyright 2022 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.
#pragma once
#include <map>
#include <vector>
#include "coordinator/hybrid_logical_clock.hpp"
#include "io/address.hpp"
#include "storage/v3/id_types.hpp"
#include "storage/v3/property_value.hpp"
#include "storage/v3/schemas.hpp"
namespace memgraph::coordinator {
using memgraph::io::Address;
using memgraph::storage::v3::LabelId;
using memgraph::storage::v3::PropertyId;
using memgraph::storage::v3::SchemaProperty;
enum class Status : uint8_t {
CONSENSUS_PARTICIPANT,
INITIALIZING,
// TODO(tyler) this will possibly have more states,
// depending on the reconfiguration protocol that we
// implement.
};
struct AddressAndStatus {
memgraph::io::Address address;
Status status;
};
using CompoundKey = std::vector<memgraph::storage::v3::PropertyValue>;
using Shard = std::vector<AddressAndStatus>;
using Shards = std::map<CompoundKey, Shard>;
using LabelName = std::string;
using PropertyName = std::string;
using PropertyMap = std::map<PropertyName, PropertyId>;
struct LabelSpace {
std::vector<SchemaProperty> schema;
std::map<CompoundKey, Shard> shards;
};
struct ShardMap {
Hlc shard_map_version;
uint64_t max_property_id;
std::map<PropertyName, PropertyId> properties;
uint64_t max_label_id;
std::map<LabelName, LabelId> labels;
std::map<LabelId, LabelSpace> label_spaces;
std::map<LabelId, std::vector<SchemaProperty>> schemas;
// TODO(gabor) later we will want to update the wallclock time with
// the given Io<impl>'s time as well
Hlc IncrementShardMapVersion() noexcept {
++shard_map_version.logical_id;
return shard_map_version;
}
Hlc GetHlc() const noexcept { return shard_map_version; }
bool SplitShard(Hlc previous_shard_map_version, LabelId label_id, const CompoundKey &key) {
if (previous_shard_map_version != shard_map_version) {
return false;
}
auto &label_space = label_spaces.at(label_id);
auto &shards_in_map = label_space.shards;
MG_ASSERT(!shards_in_map.contains(key));
MG_ASSERT(label_spaces.contains(label_id));
// Finding the Shard that the new CompoundKey should map to.
auto prev = std::prev(shards_in_map.upper_bound(key));
Shard duplicated_shard = prev->second;
// Apply the split
shards_in_map[key] = duplicated_shard;
return true;
}
bool InitializeNewLabel(std::string label_name, std::vector<SchemaProperty> schema, Hlc last_shard_map_version) {
if (shard_map_version != last_shard_map_version || labels.contains(label_name)) {
return false;
}
const LabelId label_id = LabelId::FromUint(++max_label_id);
labels.emplace(std::move(label_name), label_id);
LabelSpace label_space{
.schema = std::move(schema),
.shards = Shards{},
};
label_spaces.emplace(label_id, label_space);
IncrementShardMapVersion();
return true;
}
void AddServer(Address server_address) {
// Find a random place for the server to plug in
}
Shards GetShardsForRange(const LabelName &label_name, const CompoundKey &start_key, const CompoundKey &end_key) const {
MG_ASSERT(start_key <= end_key);
MG_ASSERT(labels.contains(label_name));
LabelId label_id = labels.at(label_name);
const auto &label_space = label_spaces.at(label_id);
const auto &shards_for_label = label_space.shards;
MG_ASSERT(shards_for_label.begin()->first <= start_key,
"the ShardMap must always contain a minimal key that is less than or equal to any requested key");
auto it = std::prev(shards_for_label.upper_bound(start_key));
const auto end_it = shards_for_label.upper_bound(end_key);
Shards shards{};
std::copy(it, end_it, std::inserter(shards, shards.end()));
return shards;
}
Shard GetShardForKey(const LabelName &label_name, const CompoundKey &key) const {
MG_ASSERT(labels.contains(label_name));
LabelId label_id = labels.at(label_name);
const auto &label_space = label_spaces.at(label_id);
MG_ASSERT(label_space.shards.begin()->first <= key,
"the ShardMap must always contain a minimal key that is less than or equal to any requested key");
return std::prev(label_space.shards.upper_bound(key))->second;
}
PropertyMap AllocatePropertyIds(const std::vector<PropertyName> &new_properties) {
PropertyMap ret{};
bool mutated = false;
for (const auto &property_name : new_properties) {
if (properties.contains(property_name)) {
auto property_id = properties.at(property_name);
ret.emplace(property_name, property_id);
} else {
mutated = true;
const PropertyId property_id = PropertyId::FromUint(++max_property_id);
ret.emplace(property_name, property_id);
properties.emplace(property_name, property_id);
}
}
if (mutated) {
IncrementShardMapVersion();
}
return ret;
}
std::optional<PropertyId> GetPropertyId(const std::string &property_name) const {
if (properties.contains(property_name)) {
return properties.at(property_name);
}
return std::nullopt;
}
};
} // namespace memgraph::coordinator

20
src/expr/CMakeLists.txt Normal file
View File

@@ -0,0 +1,20 @@
define_add_lcp(add_lcp_expr lcp_expr_cpp_files generated_lcp_expr_files)
add_lcp_expr(semantic/symbol.lcp)
add_custom_target(generate_lcp_expr DEPENDS ${generated_lcp_expr_files})
set(mg_expr_sources
${lcp_expr_cpp_files}
parsing.cpp)
find_package(Boost REQUIRED)
add_library(mg-expr STATIC ${mg_expr_sources})
add_dependencies(mg-expr generate_lcp_expr)
target_include_directories(mg-expr PUBLIC ${CMAKE_SOURCE_DIR}/include)
target_include_directories(mg-expr PUBLIC ${CMAKE_CURRENT_SOURCE_DIR})
target_include_directories(mg-expr PUBLIC ${CMAKE_CURRENT_SOURCE_DIR}/ast)
target_include_directories(mg-expr PUBLIC ${CMAKE_CURRENT_SOURCE_DIR}/interpret)
target_include_directories(mg-expr PUBLIC ${CMAKE_CURRENT_SOURCE_DIR}/semantic)
target_link_libraries(mg-expr cppitertools Boost::headers mg-utils mg-parser)

35
src/expr/ast.hpp Normal file
View File

@@ -0,0 +1,35 @@
// Copyright 2022 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.
#pragma once
#ifndef MG_AST_INCLUDE_PATH
#ifdef MG_CLANG_TIDY_CHECK
// NOLINTNEXTLINE(cppcoreguidelines-macro-usage)
#define MG_AST_INCLUDE_PATH "query/v2/frontend/ast/ast.hpp"
#else
#error Missing AST include path
#endif
#endif
#ifndef MG_INJECTED_NAMESPACE_NAME
#ifdef MG_CLANG_TIDY_CHECK
// NOLINTNEXTLINE(cppcoreguidelines-macro-usage)
#define MG_INJECTED_NAMESPACE_NAME memgraph::query::v2
#else
#error Missing AST namespace
#endif
#endif
#include MG_AST_INCLUDE_PATH
namespace memgraph::expr {
using namespace MG_INJECTED_NAMESPACE_NAME; // NOLINT(google-build-using-namespace)
} // namespace memgraph::expr

View File

@@ -0,0 +1,135 @@
// Copyright 2022 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.
#pragma once
#include "utils/visitor.hpp"
namespace MG_INJECTED_NAMESPACE_NAME {
// Forward declares for Tree visitors.
class CypherQuery;
class SingleQuery;
class CypherUnion;
class NamedExpression;
class Identifier;
class PropertyLookup;
class LabelsTest;
class Aggregation;
class Function;
class Reduce;
class Coalesce;
class Extract;
class All;
class Single;
class Any;
class None;
class ParameterLookup;
class CallProcedure;
class Create;
class Match;
class Return;
class With;
class Pattern;
class NodeAtom;
class EdgeAtom;
class PrimitiveLiteral;
class ListLiteral;
class MapLiteral;
class OrOperator;
class XorOperator;
class AndOperator;
class NotOperator;
class AdditionOperator;
class SubtractionOperator;
class MultiplicationOperator;
class DivisionOperator;
class ModOperator;
class UnaryPlusOperator;
class UnaryMinusOperator;
class IsNullOperator;
class NotEqualOperator;
class EqualOperator;
class LessOperator;
class GreaterOperator;
class LessEqualOperator;
class GreaterEqualOperator;
class InListOperator;
class SubscriptOperator;
class ListSlicingOperator;
class IfOperator;
class Delete;
class Where;
class SetProperty;
class SetProperties;
class SetLabels;
class RemoveProperty;
class RemoveLabels;
class Merge;
class Unwind;
class AuthQuery;
class ExplainQuery;
class ProfileQuery;
class IndexQuery;
class InfoQuery;
class ConstraintQuery;
class RegexMatch;
class DumpQuery;
class ReplicationQuery;
class LockPathQuery;
class LoadCsv;
class FreeMemoryQuery;
class TriggerQuery;
class IsolationLevelQuery;
class CreateSnapshotQuery;
class StreamQuery;
class SettingQuery;
class VersionQuery;
class Foreach;
class SchemaQuery;
using TreeCompositeVisitor = memgraph::utils::CompositeVisitor<
SingleQuery, CypherUnion, NamedExpression, OrOperator, XorOperator, AndOperator, NotOperator, AdditionOperator,
SubtractionOperator, MultiplicationOperator, DivisionOperator, ModOperator, NotEqualOperator, EqualOperator,
LessOperator, GreaterOperator, LessEqualOperator, GreaterEqualOperator, InListOperator, SubscriptOperator,
ListSlicingOperator, IfOperator, UnaryPlusOperator, UnaryMinusOperator, IsNullOperator, ListLiteral, MapLiteral,
PropertyLookup, LabelsTest, Aggregation, Function, Reduce, Coalesce, Extract, All, Single, Any, None, CallProcedure,
Create, Match, Return, With, Pattern, NodeAtom, EdgeAtom, Delete, Where, SetProperty, SetProperties, SetLabels,
RemoveProperty, RemoveLabels, Merge, Unwind, RegexMatch, LoadCsv, Foreach>;
using TreeLeafVisitor = memgraph::utils::LeafVisitor<Identifier, PrimitiveLiteral, ParameterLookup>;
class HierarchicalTreeVisitor : public TreeCompositeVisitor, public TreeLeafVisitor {
public:
using TreeCompositeVisitor::PostVisit;
using TreeCompositeVisitor::PreVisit;
using TreeLeafVisitor::Visit;
using typename TreeLeafVisitor::ReturnType;
};
template <class TResult>
class ExpressionVisitor
: public memgraph::utils::Visitor<
TResult, NamedExpression, OrOperator, XorOperator, AndOperator, NotOperator, AdditionOperator,
SubtractionOperator, MultiplicationOperator, DivisionOperator, ModOperator, NotEqualOperator, EqualOperator,
LessOperator, GreaterOperator, LessEqualOperator, GreaterEqualOperator, InListOperator, SubscriptOperator,
ListSlicingOperator, IfOperator, UnaryPlusOperator, UnaryMinusOperator, IsNullOperator, ListLiteral,
MapLiteral, PropertyLookup, LabelsTest, Aggregation, Function, Reduce, Coalesce, Extract, All, Single, Any,
None, ParameterLookup, Identifier, PrimitiveLiteral, RegexMatch> {};
template <class TResult>
class QueryVisitor
: public memgraph::utils::Visitor<TResult, CypherQuery, ExplainQuery, ProfileQuery, IndexQuery, AuthQuery,
InfoQuery, ConstraintQuery, DumpQuery, ReplicationQuery, LockPathQuery,
FreeMemoryQuery, TriggerQuery, IsolationLevelQuery, CreateSnapshotQuery,
StreamQuery, SettingQuery, VersionQuery, SchemaQuery> {};
} // namespace MG_INJECTED_NAMESPACE_NAME

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,271 @@
// Copyright 2022 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.
#pragma once
#include <iostream>
#include <type_traits>
#include "expr/ast.hpp"
#include "expr/typed_value.hpp"
#include "utils/algorithm.hpp"
#include "utils/logging.hpp"
#include "utils/string.hpp"
namespace memgraph::expr {
namespace detail {
template <typename T>
void PrintObject(std::ostream *out, const T &arg) {
static_assert(!std::is_convertible<T, Expression *>::value,
"This overload shouldn't be called with pointers convertible "
"to Expression *. This means your other PrintObject overloads aren't "
"being called for certain AST nodes when they should (or perhaps such "
"overloads don't exist yet).");
*out << arg;
}
inline void PrintObject(std::ostream *out, const std::string &str) { *out << utils::Escape(str); }
inline void PrintObject(std::ostream *out, Aggregation::Op op) { *out << Aggregation::OpToString(op); }
inline void PrintObject(std::ostream *out, Expression *expr);
inline void PrintObject(std::ostream *out, Identifier *expr) { PrintObject(out, static_cast<Expression *>(expr)); }
template <typename T>
void PrintObject(std::ostream *out, const std::vector<T> &vec) {
*out << "[";
utils::PrintIterable(*out, vec, ", ", [](auto &stream, const auto &item) { PrintObject(&stream, item); });
*out << "]";
}
template <typename T>
void PrintObject(std::ostream *out, const std::vector<T, utils::Allocator<T>> &vec) {
*out << "[";
utils::PrintIterable(*out, vec, ", ", [](auto &stream, const auto &item) { PrintObject(&stream, item); });
*out << "]";
}
template <typename K, typename V>
void PrintObject(std::ostream *out, const std::map<K, V> &map) {
*out << "{";
utils::PrintIterable(*out, map, ", ", [](auto &stream, const auto &item) {
PrintObject(&stream, item.first);
stream << ": ";
PrintObject(&stream, item.second);
});
*out << "}";
}
template <typename T>
void PrintObject(std::ostream *out, const utils::pmr::map<utils::pmr::string, T> &map) {
*out << "{";
utils::PrintIterable(*out, map, ", ", [](auto &stream, const auto &item) {
PrintObject(&stream, item.first);
stream << ": ";
PrintObject(&stream, item.second);
});
*out << "}";
}
template <typename T1, typename T2, typename T3>
inline void PrintObject(std::ostream *out, const TypedValueT<T1, T2, T3> &value) {
using TypedValue = TypedValueT<T1, T2, T3>;
switch (value.type()) {
case TypedValue::Type::Null:
*out << "null";
break;
case TypedValue::Type::String:
PrintObject(out, value.ValueString());
break;
case TypedValue::Type::Bool:
*out << (value.ValueBool() ? "true" : "false");
break;
case TypedValue::Type::Int:
PrintObject(out, value.ValueInt());
break;
case TypedValue::Type::Double:
PrintObject(out, value.ValueDouble());
break;
case TypedValue::Type::List:
PrintObject(out, value.ValueList());
break;
case TypedValue::Type::Map:
PrintObject(out, value.ValueMap());
break;
case TypedValue::Type::Date:
PrintObject(out, value.ValueDate());
break;
case TypedValue::Type::Duration:
PrintObject(out, value.ValueDuration());
break;
case TypedValue::Type::LocalTime:
PrintObject(out, value.ValueLocalTime());
break;
case TypedValue::Type::LocalDateTime:
PrintObject(out, value.ValueLocalDateTime());
break;
default:
MG_ASSERT(false, "PrintObject(std::ostream *out, const TypedValue &value) should not reach here");
}
}
template <typename T>
void PrintOperatorArgs(std::ostream *out, const T &arg) {
*out << " ";
PrintObject(out, arg);
*out << ")";
}
template <typename T, typename... Ts>
void PrintOperatorArgs(std::ostream *out, const T &arg, const Ts &...args) {
*out << " ";
PrintObject(out, arg);
PrintOperatorArgs(out, args...);
}
template <typename... Ts>
void PrintOperator(std::ostream *out, const std::string &name, const Ts &...args) {
*out << "(" << name;
PrintOperatorArgs(out, args...);
}
} // namespace detail
class ExpressionPrettyPrinter : public ExpressionVisitor<void> {
public:
explicit ExpressionPrettyPrinter(std::ostream *out) : out_(out) {}
// Unary operators
// NOLINTNEXTLINE(cppcoreguidelines-macro-usage)
#define UNARY_OPERATOR_VISIT(OP_NODE, OP_STR) \
/* NOLINTNEXTLINE(bugprone-macro-parentheses) */ \
void Visit(OP_NODE &op) override { detail::PrintOperator(out_, OP_STR, op.expression_); }
UNARY_OPERATOR_VISIT(NotOperator, "Not");
UNARY_OPERATOR_VISIT(UnaryPlusOperator, "+");
UNARY_OPERATOR_VISIT(UnaryMinusOperator, "-");
UNARY_OPERATOR_VISIT(IsNullOperator, "IsNull");
#undef UNARY_OPERATOR_VISIT
// Binary operators
// NOLINTNEXTLINE(cppcoreguidelines-macro-usage)
#define BINARY_OPERATOR_VISIT(OP_NODE, OP_STR) \
/* NOLINTNEXTLINE(bugprone-macro-parentheses) */ \
void Visit(OP_NODE &op) override { detail::PrintOperator(out_, OP_STR, op.expression1_, op.expression2_); }
BINARY_OPERATOR_VISIT(OrOperator, "Or");
BINARY_OPERATOR_VISIT(XorOperator, "Xor");
BINARY_OPERATOR_VISIT(AndOperator, "And");
BINARY_OPERATOR_VISIT(AdditionOperator, "+");
BINARY_OPERATOR_VISIT(SubtractionOperator, "-");
BINARY_OPERATOR_VISIT(MultiplicationOperator, "*");
BINARY_OPERATOR_VISIT(DivisionOperator, "/");
BINARY_OPERATOR_VISIT(ModOperator, "%");
BINARY_OPERATOR_VISIT(NotEqualOperator, "!=");
BINARY_OPERATOR_VISIT(EqualOperator, "==");
BINARY_OPERATOR_VISIT(LessOperator, "<");
BINARY_OPERATOR_VISIT(GreaterOperator, ">");
BINARY_OPERATOR_VISIT(LessEqualOperator, "<=");
BINARY_OPERATOR_VISIT(GreaterEqualOperator, ">=");
BINARY_OPERATOR_VISIT(InListOperator, "In");
BINARY_OPERATOR_VISIT(SubscriptOperator, "Subscript");
#undef BINARY_OPERATOR_VISIT
// Other
void Visit(ListSlicingOperator &op) override {
detail::PrintOperator(out_, "ListSlicing", op.list_, op.lower_bound_, op.upper_bound_);
}
void Visit(IfOperator &op) override {
detail::PrintOperator(out_, "If", op.condition_, op.then_expression_, op.else_expression_);
}
void Visit(ListLiteral &op) override { detail::PrintOperator(out_, "ListLiteral", op.elements_); }
void Visit(MapLiteral &op) override {
std::map<std::string, Expression *> map;
for (const auto &kv : op.elements_) {
map[kv.first.name] = kv.second;
}
detail::PrintObject(out_, map);
}
void Visit(LabelsTest &op) override { detail::PrintOperator(out_, "LabelsTest", op.expression_); }
void Visit(Aggregation &op) override { detail::PrintOperator(out_, "Aggregation", op.op_); }
void Visit(Function &op) override { detail::PrintOperator(out_, "Function", op.function_name_, op.arguments_); }
void Visit(Reduce &op) override {
detail::PrintOperator(out_, "Reduce", op.accumulator_, op.initializer_, op.identifier_, op.list_, op.expression_);
}
void Visit(Coalesce &op) override { detail::PrintOperator(out_, "Coalesce", op.expressions_); }
void Visit(Extract &op) override { detail::PrintOperator(out_, "Extract", op.identifier_, op.list_, op.expression_); }
void Visit(All &op) override {
detail::PrintOperator(out_, "All", op.identifier_, op.list_expression_, op.where_->expression_);
}
void Visit(Single &op) override {
detail::PrintOperator(out_, "Single", op.identifier_, op.list_expression_, op.where_->expression_);
}
void Visit(Any &op) override {
detail::PrintOperator(out_, "Any", op.identifier_, op.list_expression_, op.where_->expression_);
}
void Visit(None &op) override {
detail::PrintOperator(out_, "None", op.identifier_, op.list_expression_, op.where_->expression_);
}
void Visit(Identifier &op) override { detail::PrintOperator(out_, "Identifier", op.name_); }
void Visit(PrimitiveLiteral &op) override { detail::PrintObject(out_, op.value_); }
void Visit(PropertyLookup &op) override {
detail::PrintOperator(out_, "PropertyLookup", op.expression_, op.property_.name);
}
void Visit(ParameterLookup &op) override { detail::PrintOperator(out_, "ParameterLookup", op.token_position_); }
void Visit(NamedExpression &op) override { detail::PrintOperator(out_, "NamedExpression", op.name_, op.expression_); }
void Visit(RegexMatch &op) override { detail::PrintOperator(out_, "=~", op.string_expr_, op.regex_); }
private:
std::ostream *out_;
};
namespace detail {
inline void PrintObject(std::ostream *out, Expression *expr) {
if (expr) {
ExpressionPrettyPrinter printer{out};
expr->Accept(printer);
} else {
*out << "<null>";
}
}
} // namespace detail
inline void PrintExpression(Expression *expr, std::ostream *out) {
ExpressionPrettyPrinter printer{out};
expr->Accept(printer);
}
inline void PrintExpression(NamedExpression *expr, std::ostream *out) {
ExpressionPrettyPrinter printer{out};
expr->Accept(printer);
}
} // namespace memgraph::expr

35
src/expr/exceptions.hpp Normal file
View File

@@ -0,0 +1,35 @@
// Copyright 2022 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.
#pragma once
#include "utils/exceptions.hpp"
namespace memgraph::expr {
class SyntaxException : public utils::BasicException {
public:
using utils::BasicException::BasicException;
SyntaxException() : SyntaxException("") {}
};
class SemanticException : public utils::BasicException {
public:
using utils::BasicException::BasicException;
SemanticException() : BasicException("") {}
};
class ExpressionRuntimeException : public utils::BasicException {
public:
using utils::BasicException::BasicException;
};
} // namespace memgraph::expr

790
src/expr/interpret/eval.hpp Normal file
View File

@@ -0,0 +1,790 @@
// Copyright 2022 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.
/// @file
#pragma once
#include <algorithm>
#include <limits>
#include <map>
#include <optional>
#include <regex>
#include <vector>
#include "expr/ast.hpp"
#include "expr/exceptions.hpp"
#include "expr/interpret/frame.hpp"
#include "expr/semantic/symbol_table.hpp"
#include "utils/exceptions.hpp"
namespace memgraph::expr {
template <typename TypedValue, typename EvaluationContext, typename DbAccessor, typename StorageView, typename LabelId,
typename PropertyValue, typename ConvFunction, typename Error>
class ExpressionEvaluator : public ExpressionVisitor<TypedValue> {
public:
ExpressionEvaluator(Frame<TypedValue> *frame, const SymbolTable &symbol_table, const EvaluationContext &ctx,
DbAccessor *dba, StorageView view)
: frame_(frame), symbol_table_(&symbol_table), ctx_(&ctx), dba_(dba), view_(view) {}
using ExpressionVisitor<TypedValue>::Visit;
utils::MemoryResource *GetMemoryResource() const { return ctx_->memory; }
TypedValue Visit(NamedExpression &named_expression) override {
const auto &symbol = symbol_table_->at(named_expression);
auto value = named_expression.expression_->Accept(*this);
frame_->at(symbol) = value;
return value;
}
TypedValue Visit(Identifier &ident) override {
return TypedValue(frame_->at(symbol_table_->at(ident)), ctx_->memory);
}
// NOLINTNEXTLINE(cppcoreguidelines-macro-usage)
#define BINARY_OPERATOR_VISITOR(OP_NODE, CPP_OP, CYPHER_OP) \
/* NOLINTNEXTLINE(bugprone-macro-parentheses) */ \
TypedValue Visit(OP_NODE &op) override { \
auto val1 = op.expression1_->Accept(*this); \
auto val2 = op.expression2_->Accept(*this); \
try { \
return val1 CPP_OP val2; \
} catch (const TypedValueException &) { \
throw ExpressionRuntimeException("Invalid types: {} and {} for '{}'.", val1.type(), val2.type(), #CYPHER_OP); \
} \
}
// NOLINTNEXTLINE(cppcoreguidelines-macro-usage)
#define UNARY_OPERATOR_VISITOR(OP_NODE, CPP_OP, CYPHER_OP) \
/* NOLINTNEXTLINE(bugprone-macro-parentheses) */ \
TypedValue Visit(OP_NODE &op) override { \
auto val = op.expression_->Accept(*this); \
try { \
return CPP_OP val; \
} catch (const TypedValueException &) { \
throw ExpressionRuntimeException("Invalid type {} for '{}'.", val.type(), #CYPHER_OP); \
} \
}
BINARY_OPERATOR_VISITOR(OrOperator, ||, OR);
BINARY_OPERATOR_VISITOR(XorOperator, ^, XOR);
BINARY_OPERATOR_VISITOR(AdditionOperator, +, +);
BINARY_OPERATOR_VISITOR(SubtractionOperator, -, -);
BINARY_OPERATOR_VISITOR(MultiplicationOperator, *, *);
BINARY_OPERATOR_VISITOR(DivisionOperator, /, /);
BINARY_OPERATOR_VISITOR(ModOperator, %, %);
BINARY_OPERATOR_VISITOR(NotEqualOperator, !=, <>);
BINARY_OPERATOR_VISITOR(EqualOperator, ==, =);
BINARY_OPERATOR_VISITOR(LessOperator, <, <);
BINARY_OPERATOR_VISITOR(GreaterOperator, >, >);
BINARY_OPERATOR_VISITOR(LessEqualOperator, <=, <=);
BINARY_OPERATOR_VISITOR(GreaterEqualOperator, >=, >=);
UNARY_OPERATOR_VISITOR(NotOperator, !, NOT);
UNARY_OPERATOR_VISITOR(UnaryPlusOperator, +, +);
UNARY_OPERATOR_VISITOR(UnaryMinusOperator, -, -);
#undef BINARY_OPERATOR_VISITOR
#undef UNARY_OPERATOR_VISITOR
TypedValue Visit(AndOperator &op) override {
auto value1 = op.expression1_->Accept(*this);
if (value1.IsBool() && !value1.ValueBool()) {
// If first expression is false, don't evaluate the second one.
return value1;
}
auto value2 = op.expression2_->Accept(*this);
try {
return value1 && value2;
} catch (const TypedValueException &) {
throw ExpressionRuntimeException("Invalid types: {} and {} for AND.", value1.type(), value2.type());
}
}
TypedValue Visit(IfOperator &if_operator) override {
auto condition = if_operator.condition_->Accept(*this);
if (condition.IsNull()) {
return if_operator.else_expression_->Accept(*this);
}
if (condition.type() != TypedValue::Type::Bool) {
// At the moment IfOperator is used only in CASE construct.
throw ExpressionRuntimeException("CASE expected boolean expression, got {}.", condition.type());
}
if (condition.ValueBool()) {
return if_operator.then_expression_->Accept(*this);
}
return if_operator.else_expression_->Accept(*this);
}
TypedValue Visit(InListOperator &in_list) override {
auto literal = in_list.expression1_->Accept(*this);
auto _list = in_list.expression2_->Accept(*this);
if (_list.IsNull()) {
return TypedValue(ctx_->memory);
}
// Exceptions have higher priority than returning nulls when list expression
// is not null.
if (_list.type() != TypedValue::Type::List) {
throw ExpressionRuntimeException("IN expected a list, got {}.", _list.type());
}
const auto &list = _list.ValueList();
// If literal is NULL there is no need to try to compare it with every
// element in the list since result of every comparison will be NULL. There
// is one special case that we must test explicitly: if list is empty then
// result is false since no comparison will be performed.
if (list.empty()) return TypedValue(false, ctx_->memory);
if (literal.IsNull()) return TypedValue(ctx_->memory);
auto has_null = false;
for (const auto &element : list) {
auto result = literal == element;
if (result.IsNull()) {
has_null = true;
} else if (result.ValueBool()) {
return TypedValue(true, ctx_->memory);
}
}
if (has_null) {
return TypedValue(ctx_->memory);
}
return TypedValue(false, ctx_->memory);
}
TypedValue Visit(SubscriptOperator &list_indexing) override {
auto lhs = list_indexing.expression1_->Accept(*this);
auto index = list_indexing.expression2_->Accept(*this);
if (!lhs.IsList() && !lhs.IsMap() && !lhs.IsVertex() && !lhs.IsEdge() && !lhs.IsNull())
throw ExpressionRuntimeException(
"Expected a list, a map, a node or an edge to index with '[]', got "
"{}.",
lhs.type());
if (lhs.IsNull() || index.IsNull()) return TypedValue(ctx_->memory);
if (lhs.IsList()) {
if (!index.IsInt())
throw ExpressionRuntimeException("Expected an integer as a list index, got {}.", index.type());
auto index_int = index.ValueInt();
// NOTE: Take non-const reference to list, so that we can move out the
// indexed element as the result.
auto &list = lhs.ValueList();
if (index_int < 0) {
index_int += static_cast<int64_t>(list.size());
}
if (index_int >= static_cast<int64_t>(list.size()) || index_int < 0) return TypedValue(ctx_->memory);
// NOTE: Explicit move is needed, so that we return the move constructed
// value and preserve the correct MemoryResource.
return std::move(list[index_int]);
}
if (lhs.IsMap()) {
if (!index.IsString())
throw ExpressionRuntimeException("Expected a string as a map index, got {}.", index.type());
// NOTE: Take non-const reference to map, so that we can move out the
// looked-up element as the result.
auto &map = lhs.ValueMap();
auto found = map.find(index.ValueString());
if (found == map.end()) return TypedValue(ctx_->memory);
// NOTE: Explicit move is needed, so that we return the move constructed
// value and preserve the correct MemoryResource.
return std::move(found->second);
}
if (lhs.IsVertex()) {
if (!index.IsString())
throw ExpressionRuntimeException("Expected a string as a property name, got {}.", index.type());
return TypedValue(GetProperty(lhs.ValueVertex(), index.ValueString()), ctx_->memory);
}
if (lhs.IsEdge()) {
if (!index.IsString())
throw ExpressionRuntimeException("Expected a string as a property name, got {}.", index.type());
return TypedValue(GetProperty(lhs.ValueEdge(), index.ValueString()), ctx_->memory);
}
// lhs is Null
return TypedValue(ctx_->memory);
}
TypedValue Visit(ListSlicingOperator &op) override {
// If some type is null we can't return null, because throwing exception
// on illegal type has higher priority.
auto is_null = false;
auto get_bound = [&](Expression *bound_expr, int64_t default_value) {
if (bound_expr) {
auto bound = bound_expr->Accept(*this);
if (bound.type() == TypedValue::Type::Null) {
is_null = true;
} else if (bound.type() != TypedValue::Type::Int) {
throw ExpressionRuntimeException("Expected an integer for a bound in list slicing, got {}.", bound.type());
}
return bound;
}
return TypedValue(default_value, ctx_->memory);
};
auto _upper_bound = get_bound(op.upper_bound_, std::numeric_limits<int64_t>::max());
auto _lower_bound = get_bound(op.lower_bound_, 0);
auto _list = op.list_->Accept(*this);
if (_list.type() == TypedValue::Type::Null) {
is_null = true;
} else if (_list.type() != TypedValue::Type::List) {
throw ExpressionRuntimeException("Expected a list to slice, got {}.", _list.type());
}
if (is_null) {
return TypedValue(ctx_->memory);
}
const auto &list = _list.ValueList();
auto normalise_bound = [&](int64_t bound) {
if (bound < 0) {
bound = static_cast<int64_t>(list.size()) + bound;
}
return std::max(static_cast<int64_t>(0), std::min(bound, static_cast<int64_t>(list.size())));
};
auto lower_bound = normalise_bound(_lower_bound.ValueInt());
auto upper_bound = normalise_bound(_upper_bound.ValueInt());
if (upper_bound <= lower_bound) {
return TypedValue(typename TypedValue::TVector(ctx_->memory), ctx_->memory);
}
return TypedValue(
typename TypedValue::TVector(list.begin() + lower_bound, list.begin() + upper_bound, ctx_->memory));
}
TypedValue Visit(IsNullOperator &is_null) override {
auto value = is_null.expression_->Accept(*this);
return TypedValue(value.IsNull(), ctx_->memory);
}
TypedValue Visit(PropertyLookup &property_lookup) override {
auto expression_result = property_lookup.expression_->Accept(*this);
auto maybe_date = [this](const auto &date, const auto &prop_name) -> std::optional<TypedValue> {
if (prop_name == "year") {
return TypedValue(date.year, ctx_->memory);
}
if (prop_name == "month") {
return TypedValue(date.month, ctx_->memory);
}
if (prop_name == "day") {
return TypedValue(date.day, ctx_->memory);
}
return std::nullopt;
};
auto maybe_local_time = [this](const auto &lt, const auto &prop_name) -> std::optional<TypedValue> {
if (prop_name == "hour") {
return TypedValue(lt.hour, ctx_->memory);
}
if (prop_name == "minute") {
return TypedValue(lt.minute, ctx_->memory);
}
if (prop_name == "second") {
return TypedValue(lt.second, ctx_->memory);
}
if (prop_name == "millisecond") {
return TypedValue(lt.millisecond, ctx_->memory);
}
if (prop_name == "microsecond") {
return TypedValue(lt.microsecond, ctx_->memory);
}
return std::nullopt;
};
auto maybe_duration = [this](const auto &dur, const auto &prop_name) -> std::optional<TypedValue> {
if (prop_name == "day") {
return TypedValue(dur.Days(), ctx_->memory);
}
if (prop_name == "hour") {
return TypedValue(dur.SubDaysAsHours(), ctx_->memory);
}
if (prop_name == "minute") {
return TypedValue(dur.SubDaysAsMinutes(), ctx_->memory);
}
if (prop_name == "second") {
return TypedValue(dur.SubDaysAsSeconds(), ctx_->memory);
}
if (prop_name == "millisecond") {
return TypedValue(dur.SubDaysAsMilliseconds(), ctx_->memory);
}
if (prop_name == "microsecond") {
return TypedValue(dur.SubDaysAsMicroseconds(), ctx_->memory);
}
if (prop_name == "nanosecond") {
return TypedValue(dur.SubDaysAsNanoseconds(), ctx_->memory);
}
return std::nullopt;
};
switch (expression_result.type()) {
case TypedValue::Type::Null:
return TypedValue(ctx_->memory);
case TypedValue::Type::Vertex:
return GetProperty(expression_result.ValueVertex(), property_lookup.property_);
case TypedValue::Type::Edge:
return GetProperty(expression_result.ValueEdge(), property_lookup.property_);
case TypedValue::Type::Map: {
// NOTE: Take non-const reference to map, so that we can move out the
// looked-up element as the result.
auto &map = expression_result.ValueMap();
auto found = map.find(property_lookup.property_.name.c_str());
if (found == map.end()) return TypedValue(ctx_->memory);
// NOTE: Explicit move is needed, so that we return the move constructed
// value and preserve the correct MemoryResource.
return std::move(found->second);
}
case TypedValue::Type::Duration: {
const auto &prop_name = property_lookup.property_.name;
const auto &dur = expression_result.ValueDuration();
if (auto dur_field = maybe_duration(dur, prop_name); dur_field) {
return std::move(*dur_field);
}
throw ExpressionRuntimeException("Invalid property name {} for Duration", prop_name);
}
case TypedValue::Type::Date: {
const auto &prop_name = property_lookup.property_.name;
const auto &date = expression_result.ValueDate();
if (auto date_field = maybe_date(date, prop_name); date_field) {
return std::move(*date_field);
}
throw ExpressionRuntimeException("Invalid property name {} for Date", prop_name);
}
case TypedValue::Type::LocalTime: {
const auto &prop_name = property_lookup.property_.name;
const auto &lt = expression_result.ValueLocalTime();
if (auto lt_field = maybe_local_time(lt, prop_name); lt_field) {
return std::move(*lt_field);
}
throw ExpressionRuntimeException("Invalid property name {} for LocalTime", prop_name);
}
case TypedValue::Type::LocalDateTime: {
const auto &prop_name = property_lookup.property_.name;
const auto &ldt = expression_result.ValueLocalDateTime();
if (auto date_field = maybe_date(ldt.date, prop_name); date_field) {
return std::move(*date_field);
}
if (auto lt_field = maybe_local_time(ldt.local_time, prop_name); lt_field) {
return std::move(*lt_field);
}
throw ExpressionRuntimeException("Invalid property name {} for LocalDateTime", prop_name);
}
default:
throw ExpressionRuntimeException("Only nodes, edges, maps and temporal types have properties to be looked-up.");
}
}
TypedValue Visit(LabelsTest &labels_test) override {
auto expression_result = labels_test.expression_->Accept(*this);
switch (expression_result.type()) {
case TypedValue::Type::Null:
return TypedValue(ctx_->memory);
case TypedValue::Type::Vertex: {
const auto &vertex = expression_result.ValueVertex();
for (const auto &label : labels_test.labels_) {
auto has_label = vertex.HasLabel(view_, GetLabel(label));
if (has_label.HasError() && has_label.GetError() == Error::NONEXISTENT_OBJECT) {
// This is a very nasty and temporary hack in order to make MERGE
// work. The old storage had the following logic when returning an
// `OLD` view: `return old ? old : new`. That means that if the
// `OLD` view didn't exist, it returned the NEW view. With this hack
// we simulate that behavior.
// TODO (mferencevic, teon.banek): Remove once MERGE is
// reimplemented.
has_label = vertex.HasLabel(StorageView::NEW, GetLabel(label));
}
if (has_label.HasError()) {
switch (has_label.GetError()) {
case Error::DELETED_OBJECT:
throw ExpressionRuntimeException("Trying to access labels on a deleted node.");
case Error::NONEXISTENT_OBJECT:
throw ExpressionRuntimeException("Trying to access labels from a node that doesn't exist.");
case Error::SERIALIZATION_ERROR:
case Error::VERTEX_HAS_EDGES:
case Error::PROPERTIES_DISABLED:
throw ExpressionRuntimeException("Unexpected error when accessing labels.");
}
}
if (!*has_label) {
return TypedValue(false, ctx_->memory);
}
}
return TypedValue(true, ctx_->memory);
}
default:
throw ExpressionRuntimeException("Only nodes have labels.");
}
}
TypedValue Visit(PrimitiveLiteral &literal) override {
// TODO: no need to evaluate constants, we can write it to frame in one
// of the previous phases.
return TypedValue(literal.value_, ctx_->memory);
}
TypedValue Visit(ListLiteral &literal) override {
typename TypedValue::TVector result(ctx_->memory);
result.reserve(literal.elements_.size());
for (const auto &expression : literal.elements_) result.emplace_back(expression->Accept(*this));
return TypedValue(result, ctx_->memory);
}
TypedValue Visit(MapLiteral &literal) override {
typename TypedValue::TMap result(ctx_->memory);
for (const auto &pair : literal.elements_) result.emplace(pair.first.name, pair.second->Accept(*this));
return TypedValue(result, ctx_->memory);
}
TypedValue Visit(Aggregation &aggregation) override {
return TypedValue(frame_->at(symbol_table_->at(aggregation)), ctx_->memory);
}
TypedValue Visit(Coalesce &coalesce) override {
auto &exprs = coalesce.expressions_;
if (exprs.size() == 0) {
throw ExpressionRuntimeException("'coalesce' requires at least one argument.");
}
for (int64_t i = 0; i < exprs.size(); ++i) {
TypedValue val(exprs[i]->Accept(*this), ctx_->memory);
if (!val.IsNull()) {
return val;
}
}
return TypedValue(ctx_->memory);
}
TypedValue Visit(Function &function) override {
FunctionContext function_ctx{dba_, ctx_->memory, ctx_->timestamp, &ctx_->counters, view_};
// Stack allocate evaluated arguments when there's a small number of them.
if (function.arguments_.size() <= 8) {
TypedValue arguments[8] = {TypedValue(ctx_->memory), TypedValue(ctx_->memory), TypedValue(ctx_->memory),
TypedValue(ctx_->memory), TypedValue(ctx_->memory), TypedValue(ctx_->memory),
TypedValue(ctx_->memory), TypedValue(ctx_->memory)};
for (size_t i = 0; i < function.arguments_.size(); ++i) {
arguments[i] = function.arguments_[i]->Accept(*this);
}
auto res = function.function_(arguments, function.arguments_.size(), function_ctx);
MG_ASSERT(res.GetMemoryResource() == ctx_->memory);
return res;
} else {
typename TypedValue::TVector arguments(ctx_->memory);
arguments.reserve(function.arguments_.size());
for (const auto &argument : function.arguments_) {
arguments.emplace_back(argument->Accept(*this));
}
auto res = function.function_(arguments.data(), arguments.size(), function_ctx);
MG_ASSERT(res.GetMemoryResource() == ctx_->memory);
return res;
}
}
TypedValue Visit(Reduce &reduce) override {
auto list_value = reduce.list_->Accept(*this);
if (list_value.IsNull()) {
return TypedValue(ctx_->memory);
}
if (list_value.type() != TypedValue::Type::List) {
throw ExpressionRuntimeException("REDUCE expected a list, got {}.", list_value.type());
}
const auto &list = list_value.ValueList();
const auto &element_symbol = symbol_table_->at(*reduce.identifier_);
const auto &accumulator_symbol = symbol_table_->at(*reduce.accumulator_);
auto accumulator = reduce.initializer_->Accept(*this);
for (const auto &element : list) {
frame_->at(accumulator_symbol) = accumulator;
frame_->at(element_symbol) = element;
accumulator = reduce.expression_->Accept(*this);
}
return accumulator;
}
TypedValue Visit(Extract &extract) override {
auto list_value = extract.list_->Accept(*this);
if (list_value.IsNull()) {
return TypedValue(ctx_->memory);
}
if (list_value.type() != TypedValue::Type::List) {
throw ExpressionRuntimeException("EXTRACT expected a list, got {}.", list_value.type());
}
const auto &list = list_value.ValueList();
const auto &element_symbol = symbol_table_->at(*extract.identifier_);
typename TypedValue::TVector result(ctx_->memory);
result.reserve(list.size());
for (const auto &element : list) {
if (element.IsNull()) {
result.emplace_back();
} else {
frame_->at(element_symbol) = element;
result.emplace_back(extract.expression_->Accept(*this));
}
}
return TypedValue(result, ctx_->memory);
}
TypedValue Visit(All &all) override {
auto list_value = all.list_expression_->Accept(*this);
if (list_value.IsNull()) {
return TypedValue(ctx_->memory);
}
if (list_value.type() != TypedValue::Type::List) {
throw ExpressionRuntimeException("ALL expected a list, got {}.", list_value.type());
}
const auto &list = list_value.ValueList();
const auto &symbol = symbol_table_->at(*all.identifier_);
bool has_null_elements = false;
bool has_value = false;
for (const auto &element : list) {
frame_->at(symbol) = element;
auto result = all.where_->expression_->Accept(*this);
if (!result.IsNull() && result.type() != TypedValue::Type::Bool) {
throw ExpressionRuntimeException("Predicate of ALL must evaluate to boolean, got {}.", result.type());
}
if (!result.IsNull()) {
has_value = true;
if (!result.ValueBool()) {
return TypedValue(false, ctx_->memory);
}
} else {
has_null_elements = true;
}
}
if (!has_value) {
return TypedValue(ctx_->memory);
}
if (has_null_elements) {
return TypedValue(false, ctx_->memory);
} else {
return TypedValue(true, ctx_->memory);
}
}
TypedValue Visit(Single &single) override {
auto list_value = single.list_expression_->Accept(*this);
if (list_value.IsNull()) {
return TypedValue(ctx_->memory);
}
if (list_value.type() != TypedValue::Type::List) {
throw ExpressionRuntimeException("SINGLE expected a list, got {}.", list_value.type());
}
const auto &list = list_value.ValueList();
const auto &symbol = symbol_table_->at(*single.identifier_);
bool has_value = false;
bool predicate_satisfied = false;
for (const auto &element : list) {
frame_->at(symbol) = element;
auto result = single.where_->expression_->Accept(*this);
if (!result.IsNull() && result.type() != TypedValue::Type::Bool) {
throw ExpressionRuntimeException("Predicate of SINGLE must evaluate to boolean, got {}.", result.type());
}
if (result.type() == TypedValue::Type::Bool) {
has_value = true;
}
if (result.IsNull() || !result.ValueBool()) {
continue;
}
// Return false if more than one element satisfies the predicate.
if (predicate_satisfied) {
return TypedValue(false, ctx_->memory);
} else {
predicate_satisfied = true;
}
}
if (!has_value) {
return TypedValue(ctx_->memory);
} else {
return TypedValue(predicate_satisfied, ctx_->memory);
}
}
TypedValue Visit(Any &any) override {
auto list_value = any.list_expression_->Accept(*this);
if (list_value.IsNull()) {
return TypedValue(ctx_->memory);
}
if (list_value.type() != TypedValue::Type::List) {
throw ExpressionRuntimeException("ANY expected a list, got {}.", list_value.type());
}
const auto &list = list_value.ValueList();
const auto &symbol = symbol_table_->at(*any.identifier_);
bool has_value = false;
for (const auto &element : list) {
frame_->at(symbol) = element;
auto result = any.where_->expression_->Accept(*this);
if (!result.IsNull() && result.type() != TypedValue::Type::Bool) {
throw ExpressionRuntimeException("Predicate of ANY must evaluate to boolean, got {}.", result.type());
}
if (!result.IsNull()) {
has_value = true;
if (result.ValueBool()) {
return TypedValue(true, ctx_->memory);
}
}
}
// Return Null if all elements are Null
if (!has_value) {
return TypedValue(ctx_->memory);
} else {
return TypedValue(false, ctx_->memory);
}
}
TypedValue Visit(None &none) override {
auto list_value = none.list_expression_->Accept(*this);
if (list_value.IsNull()) {
return TypedValue(ctx_->memory);
}
if (list_value.type() != TypedValue::Type::List) {
throw ExpressionRuntimeException("NONE expected a list, got {}.", list_value.type());
}
const auto &list = list_value.ValueList();
const auto &symbol = symbol_table_->at(*none.identifier_);
bool has_value = false;
for (const auto &element : list) {
frame_->at(symbol) = element;
auto result = none.where_->expression_->Accept(*this);
if (!result.IsNull() && result.type() != TypedValue::Type::Bool) {
throw ExpressionRuntimeException("Predicate of NONE must evaluate to boolean, got {}.", result.type());
}
if (!result.IsNull()) {
has_value = true;
if (result.ValueBool()) {
return TypedValue(false, ctx_->memory);
}
}
}
// Return Null if all elements are Null
if (!has_value) {
return TypedValue(ctx_->memory);
} else {
return TypedValue(true, ctx_->memory);
}
}
TypedValue Visit(ParameterLookup &param_lookup) override {
return TypedValue(conv_(ctx_->parameters.AtTokenPosition(param_lookup.token_position_)), ctx_->memory);
}
TypedValue Visit(RegexMatch &regex_match) override {
auto target_string_value = regex_match.string_expr_->Accept(*this);
auto regex_value = regex_match.regex_->Accept(*this);
if (target_string_value.IsNull() || regex_value.IsNull()) {
return TypedValue(ctx_->memory);
}
if (regex_value.type() != TypedValue::Type::String) {
throw ExpressionRuntimeException("Regular expression must evaluate to a string, got {}.", regex_value.type());
}
if (target_string_value.type() != TypedValue::Type::String) {
// Instead of error, we return Null which makes it compatible in case we
// use indexed lookup which filters out any non-string properties.
// Assuming a property lookup is the target_string_value.
return TypedValue(ctx_->memory);
}
const auto &target_string = target_string_value.ValueString();
try {
std::regex regex(regex_value.ValueString());
return TypedValue(std::regex_match(target_string, regex), ctx_->memory);
} catch (const std::regex_error &e) {
throw ExpressionRuntimeException("Regex error in '{}': {}", regex_value.ValueString(), e.what());
}
}
private:
template <class TRecordAccessor>
TypedValue GetProperty(const TRecordAccessor &record_accessor, PropertyIx prop) {
auto maybe_prop = record_accessor.GetProperty(view_, ctx_->properties[prop.ix]);
if (maybe_prop.HasError() && maybe_prop.GetError() == Error::NONEXISTENT_OBJECT) {
// This is a very nasty and temporary hack in order to make MERGE work.
// The old storage had the following logic when returning an `OLD` view:
// `return old ? old : new`. That means that if the `OLD` view didn't
// exist, it returned the NEW view. With this hack we simulate that
// behavior.
// TODO (mferencevic, teon.banek): Remove once MERGE is reimplemented.
maybe_prop = record_accessor.GetProperty(StorageView::NEW, ctx_->properties[prop.ix]);
}
if (maybe_prop.HasError()) {
switch (maybe_prop.GetError()) {
case Error::DELETED_OBJECT:
throw ExpressionRuntimeException("Trying to get a property from a deleted object.");
case Error::NONEXISTENT_OBJECT:
throw ExpressionRuntimeException("Trying to get a property from an object that doesn't exist.");
case Error::SERIALIZATION_ERROR:
case Error::VERTEX_HAS_EDGES:
case Error::PROPERTIES_DISABLED:
throw ExpressionRuntimeException("Unexpected error when getting a property.");
}
}
return conv_(*maybe_prop);
}
template <class TRecordAccessor>
TypedValue GetProperty(const TRecordAccessor &record_accessor, const std::string_view name) {
auto maybe_prop = record_accessor.GetProperty(view_, dba_->NameToProperty(name));
if (maybe_prop.HasError() && maybe_prop.GetError() == Error::NONEXISTENT_OBJECT) {
// This is a very nasty and temporary hack in order to make MERGE work.
// The old storage had the following logic when returning an `OLD` view:
// `return old ? old : new`. That means that if the `OLD` view didn't
// exist, it returned the NEW view. With this hack we simulate that
// behavior.
// TODO (mferencevic, teon.banek): Remove once MERGE is reimplemented.
maybe_prop = record_accessor.GetProperty(view_, dba_->NameToProperty(name));
}
if (maybe_prop.HasError()) {
switch (maybe_prop.GetError()) {
case Error::DELETED_OBJECT:
throw ExpressionRuntimeException("Trying to get a property from a deleted object.");
case Error::NONEXISTENT_OBJECT:
throw ExpressionRuntimeException("Trying to get a property from an object that doesn't exist.");
case Error::SERIALIZATION_ERROR:
case Error::VERTEX_HAS_EDGES:
case Error::PROPERTIES_DISABLED:
throw ExpressionRuntimeException("Unexpected error when getting a property.");
}
}
return conv_(*maybe_prop);
}
LabelId GetLabel(LabelIx label) { return ctx_->labels[label.ix]; }
Frame<TypedValue> *frame_;
const SymbolTable *symbol_table_;
const EvaluationContext *ctx_;
DbAccessor *dba_;
// which switching approach should be used when evaluating
StorageView view_;
ConvFunction conv_;
};
/// A helper function for evaluating an expression that's an int.
///
/// @param what - Name of what's getting evaluated. Used for user feedback (via
/// exception) when the evaluated value is not an int.
/// @throw ExpressionRuntimeException if expression doesn't evaluate to an int.
template <typename ExpressionEvaluator>
int64_t EvaluateInt(ExpressionEvaluator *evaluator, Expression *expr, const std::string &what) {
TypedValue value = expr->Accept(*evaluator);
try {
return value.ValueInt();
} catch (TypedValueException &e) {
throw ExpressionRuntimeException(what + " must be an int");
}
}
template <typename ExpressionEvaluator>
std::optional<size_t> EvaluateMemoryLimit(ExpressionEvaluator *eval, Expression *memory_limit, size_t memory_scale) {
if (!memory_limit) return std::nullopt;
auto limit_value = memory_limit->Accept(*eval);
if (!limit_value.IsInt() || limit_value.ValueInt() <= 0)
throw ExpressionRuntimeException("Memory limit must be a non-negative integer.");
size_t limit = limit_value.ValueInt();
if (std::numeric_limits<size_t>::max() / memory_scale < limit)
throw ExpressionRuntimeException("Memory limit overflow.");
return limit * memory_scale;
}
} // namespace memgraph::expr

View File

@@ -0,0 +1,45 @@
// Copyright 2022 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.
#pragma once
#include <vector>
#include "expr/semantic/symbol_table.hpp"
#include "utils/logging.hpp"
#include "utils/memory.hpp"
#include "utils/pmr/vector.hpp"
namespace memgraph::expr {
template <typename TypedValue>
class Frame {
public:
/// Create a Frame of given size backed by a utils::NewDeleteResource()
explicit Frame(int64_t size) : elems_(size, utils::NewDeleteResource()) { MG_ASSERT(size >= 0); }
Frame(int64_t size, utils::MemoryResource *memory) : elems_(size, memory) { MG_ASSERT(size >= 0); }
TypedValue &operator[](const Symbol &symbol) { return elems_[symbol.position()]; }
const TypedValue &operator[](const Symbol &symbol) const { return elems_[symbol.position()]; }
TypedValue &at(const Symbol &symbol) { return elems_.at(symbol.position()); }
const TypedValue &at(const Symbol &symbol) const { return elems_.at(symbol.position()); }
auto &elems() { return elems_; }
utils::MemoryResource *GetMemoryResource() const { return elems_.get_allocator().GetMemoryResource(); }
private:
utils::pmr::vector<TypedValue> elems_;
};
} // namespace memgraph::expr

184
src/expr/parsing.cpp Normal file
View File

@@ -0,0 +1,184 @@
// Copyright 2022 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 "expr/parsing.hpp"
#include <cctype>
#include <codecvt>
#include <locale>
#include <stdexcept>
#include "expr/exceptions.hpp"
#include "utils/logging.hpp"
#include "utils/string.hpp"
namespace memgraph::expr {
int64_t ParseIntegerLiteral(const std::string &s) {
try {
// Not really correct since long long can have a bigger range than int64_t.
return static_cast<int64_t>(std::stoll(s, 0, 0));
} catch (const std::out_of_range &) {
throw SemanticException("Integer literal exceeds 64 bits.");
}
}
std::string ParseStringLiteral(const std::string &s) {
// These functions is declared as lambda since its semantics is highly
// specific for this conxtext and shouldn't be used elsewhere.
auto EncodeEscapedUnicodeCodepointUtf32 = [](const std::string &s, int &i) {
const int kLongUnicodeLength = 8;
int j = i + 1;
while (j < static_cast<int>(s.size()) - 1 && j < i + kLongUnicodeLength + 1 && isxdigit(s[j])) {
++j;
}
if (j - i == kLongUnicodeLength + 1) {
char32_t t = stoi(s.substr(i + 1, kLongUnicodeLength), 0, 16);
i += kLongUnicodeLength;
std::wstring_convert<std::codecvt_utf8<char32_t>, char32_t> converter;
return converter.to_bytes(t);
}
throw SyntaxException(
"Expected 8 hex digits as unicode codepoint started with \\U. "
"Use \\u for 4 hex digits format.");
};
auto EncodeEscapedUnicodeCodepointUtf16 = [](const std::string &s, int &i) {
const int kShortUnicodeLength = 4;
int j = i + 1;
while (j < static_cast<int>(s.size()) - 1 && j < i + kShortUnicodeLength + 1 && isxdigit(s[j])) {
++j;
}
if (j - i >= kShortUnicodeLength + 1) {
char16_t t = stoi(s.substr(i + 1, kShortUnicodeLength), 0, 16);
if (t >= 0xD800 && t <= 0xDBFF) {
// t is high surrogate pair. Expect one more utf16 codepoint.
j = i + kShortUnicodeLength + 1;
if (j >= static_cast<int>(s.size()) - 1 || s[j] != '\\') {
throw SemanticException("Invalid UTF codepoint.");
}
++j;
if (j >= static_cast<int>(s.size()) - 1 || (s[j] != 'u' && s[j] != 'U')) {
throw SemanticException("Invalid UTF codepoint.");
}
++j;
int k = j;
while (k < static_cast<int>(s.size()) - 1 && k < j + kShortUnicodeLength && isxdigit(s[k])) {
++k;
}
if (k != j + kShortUnicodeLength) {
throw SemanticException("Invalid UTF codepoint.");
}
char16_t surrogates[3] = {t, static_cast<char16_t>(stoi(s.substr(j, kShortUnicodeLength), 0, 16)), 0};
i += kShortUnicodeLength + 2 + kShortUnicodeLength;
std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t> converter;
return converter.to_bytes(surrogates);
} else {
i += kShortUnicodeLength;
std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t> converter;
return converter.to_bytes(t);
}
}
throw SyntaxException(
"Expected 4 hex digits as unicode codepoint started with \\u. "
"Use \\U for 8 hex digits format.");
};
std::string unescaped;
bool escape = false;
// First and last char is quote, we don't need to look at them.
for (int i = 1; i < static_cast<int>(s.size()) - 1; ++i) {
if (escape) {
switch (s[i]) {
case '\\':
unescaped += '\\';
break;
case '\'':
unescaped += '\'';
break;
case '"':
unescaped += '"';
break;
case 'B':
case 'b':
unescaped += '\b';
break;
case 'F':
case 'f':
unescaped += '\f';
break;
case 'N':
case 'n':
unescaped += '\n';
break;
case 'R':
case 'r':
unescaped += '\r';
break;
case 'T':
case 't':
unescaped += '\t';
break;
case 'U':
try {
unescaped += EncodeEscapedUnicodeCodepointUtf32(s, i);
} catch (const std::range_error &) {
throw SemanticException("Invalid UTF codepoint.");
}
break;
case 'u':
try {
unescaped += EncodeEscapedUnicodeCodepointUtf16(s, i);
} catch (const std::range_error &) {
throw SemanticException("Invalid UTF codepoint.");
}
break;
default:
// This should never happen, except grammar changes and we don't
// notice change in this production.
DLOG_FATAL("can't happen");
throw std::exception();
}
escape = false;
} else if (s[i] == '\\') {
escape = true;
} else {
unescaped += s[i];
}
}
return unescaped;
}
double ParseDoubleLiteral(const std::string &s) {
try {
return utils::ParseDouble(s);
} catch (const utils::BasicException &) {
throw SemanticException("Couldn't parse string to double.");
}
}
std::string ParseParameter(const std::string &s) {
DMG_ASSERT(s[0] == '$', "Invalid string passed as parameter name");
if (s[1] != '`') return s.substr(1);
// If parameter name is escaped symbolic name then symbolic name should be
// unescaped and leading and trailing backquote should be removed.
DMG_ASSERT(s.size() > 3U && s.back() == '`', "Invalid string passed as parameter name");
std::string out;
for (int i = 2; i < static_cast<int>(s.size()) - 1; ++i) {
if (s[i] == '`') {
++i;
}
out.push_back(s[i]);
}
return out;
}
} // namespace memgraph::expr

27
src/expr/parsing.hpp Normal file
View File

@@ -0,0 +1,27 @@
// Copyright 2022 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.
/// @file
#pragma once
#include <cstdint>
#include <string>
namespace memgraph::expr {
// These are the functions for parsing literals and parameter names from
// opencypher query.
int64_t ParseIntegerLiteral(const std::string &s);
std::string ParseStringLiteral(const std::string &s);
double ParseDoubleLiteral(const std::string &s);
std::string ParseParameter(const std::string &s);
} // namespace memgraph::expr

View File

@@ -0,0 +1,87 @@
;; Copyright 2022 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.
#>cpp
#pragma once
#include <string>
#include "utils/typeinfo.hpp"
cpp<#
(lcp:namespace memgraph)
(lcp:namespace expr)
(lcp:define-class symbol ()
((name "std::string" :scope :public)
(position :int64_t :scope :public)
(user-declared :bool :initval "true" :scope :public)
(type "Type" :initval "Type::ANY" :scope :public)
(token-position :int64_t :initval "-1" :scope :public))
(:public
;; This is similar to TypedValue::Type, but this has `Any` type.
;; TODO: Make a better Type structure which can store a generic List.
(lcp:define-enum type (any vertex edge path number edge-list)
(:serialize))
#>cpp
// TODO: Generate enum to string conversion from LCP. Note, that this is
// displayed to the end user, so we may want to have a pretty name of each
// value.
static std::string TypeToString(Type type) {
const char *enum_string[] = {"Any", "Vertex", "Edge",
"Path", "Number", "EdgeList"};
return enum_string[static_cast<int>(type)];
}
Symbol() {}
Symbol(const std::string &name, int position, bool user_declared,
Type type = Type::ANY, int token_position = -1)
: name_(name),
position_(position),
user_declared_(user_declared),
type_(type),
token_position_(token_position) {}
bool operator==(const Symbol &other) const {
return position_ == other.position_ && name_ == other.name_ &&
type_ == other.type_;
}
bool operator!=(const Symbol &other) const { return !operator==(other); }
// TODO: Remove these since members are public
const auto &name() const { return name_; }
int position() const { return position_; }
Type type() const { return type_; }
bool user_declared() const { return user_declared_; }
int token_position() const { return token_position_; }
cpp<#)
(:serialize (:slk)))
(lcp:pop-namespace) ;; expr
(lcp:pop-namespace) ;; memgraph
#>cpp
namespace std {
template <>
struct hash<memgraph::expr::Symbol> {
size_t operator()(const memgraph::expr::Symbol &symbol) const {
size_t prime = 265443599u;
size_t hash = std::hash<int>{}(symbol.position());
hash ^= prime * std::hash<std::string>{}(symbol.name());
hash ^= prime * std::hash<int>{}(static_cast<int>(symbol.type()));
return hash;
}
};
} // namespace std
cpp<#

View File

@@ -0,0 +1,712 @@
// Copyright 2022 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.
// Copyright 2017 Memgraph
//
// Created by Teon Banek on 11-03-2017
#pragma once
#include <algorithm>
#include <optional>
#include <ranges>
#include <unordered_map>
#include <unordered_set>
#include <variant>
#include <vector>
#include "expr/ast.hpp"
#include "expr/ast/ast_visitor.hpp"
#include "expr/exceptions.hpp"
#include "expr/semantic/symbol_table.hpp"
namespace memgraph::expr {
namespace detail {
inline std::unordered_map<std::string, Identifier *> GeneratePredefinedIdentifierMap(
const std::vector<Identifier *> &predefined_identifiers) {
std::unordered_map<std::string, Identifier *> identifier_map;
for (const auto &identifier : predefined_identifiers) {
identifier_map.emplace(identifier->name_, identifier);
}
return identifier_map;
}
} // namespace detail
/// Visits the AST and generates symbols for variables.
///
/// During the process of symbol generation, simple semantic checks are
/// performed. Such as, redeclaring a variable or conflicting expectations of
/// variable types.
class SymbolGenerator : public HierarchicalTreeVisitor {
public:
explicit SymbolGenerator(SymbolTable *symbol_table, const std::vector<Identifier *> &predefined_identifiers)
: symbol_table_(symbol_table),
predefined_identifiers_{detail::GeneratePredefinedIdentifierMap(predefined_identifiers)},
scopes_(1, Scope()) {}
using HierarchicalTreeVisitor::PostVisit;
using HierarchicalTreeVisitor::PreVisit;
using HierarchicalTreeVisitor::Visit;
using typename HierarchicalTreeVisitor::ReturnType;
// Query
bool PreVisit(SingleQuery & /*unused*/) override {
prev_return_names_ = curr_return_names_;
curr_return_names_.clear();
return true;
}
// Union
bool PreVisit(CypherUnion & /*unused*/) override {
scopes_.back() = Scope();
return true;
}
bool PostVisit(CypherUnion &cypher_union) override {
if (prev_return_names_ != curr_return_names_) {
throw SemanticException("All subqueries in an UNION must have the same column names.");
}
// create new symbols for the result of the union
for (const auto &name : curr_return_names_) {
auto symbol = CreateSymbol(name, false);
cypher_union.union_symbols_.push_back(symbol);
}
return true;
}
// Clauses
bool PreVisit(Create & /*unused*/) override {
scopes_.back().in_create = true;
return true;
}
bool PostVisit(Create & /*unused*/) override {
scopes_.back().in_create = false;
return true;
}
bool PreVisit(CallProcedure &call_proc) override {
for (auto *expr : call_proc.arguments_) {
expr->Accept(*this);
}
return false;
}
bool PostVisit(CallProcedure &call_proc) override {
for (auto *ident : call_proc.result_identifiers_) {
if (HasSymbolLocalScope(ident->name_)) {
throw RedeclareVariableError(ident->name_);
}
ident->MapTo(CreateSymbol(ident->name_, true));
}
return true;
}
bool PreVisit(LoadCsv & /*unused*/) override { return false; }
bool PostVisit(LoadCsv &load_csv) override {
if (HasSymbolLocalScope(load_csv.row_var_->name_)) {
throw RedeclareVariableError(load_csv.row_var_->name_);
}
load_csv.row_var_->MapTo(CreateSymbol(load_csv.row_var_->name_, true));
return true;
}
bool PreVisit(Return &ret) override {
auto &scope = scopes_.back();
scope.in_return = true;
VisitReturnBody(ret.body_);
scope.in_return = false;
return false; // We handled the traversal ourselves.
}
bool PostVisit(Return & /*unused*/) override {
for (const auto &name_symbol : scopes_.back().symbols) curr_return_names_.insert(name_symbol.first);
return true;
}
bool PreVisit(With &with) override {
auto &scope = scopes_.back();
scope.in_with = true;
VisitReturnBody(with.body_, with.where_);
scope.in_with = false;
return false; // We handled the traversal ourselves.
}
bool PreVisit(Where & /*unused*/) override {
scopes_.back().in_where = true;
return true;
}
bool PostVisit(Where & /*unused*/) override {
scopes_.back().in_where = false;
return true;
}
bool PreVisit(Merge & /*unused*/) override {
scopes_.back().in_merge = true;
return true;
}
bool PostVisit(Merge & /*unused*/) override {
scopes_.back().in_merge = false;
return true;
}
bool PostVisit(Unwind &unwind) override {
const auto &name = unwind.named_expression_->name_;
if (HasSymbolLocalScope(name)) {
throw RedeclareVariableError(name);
}
unwind.named_expression_->MapTo(CreateSymbol(name, true));
return true;
}
bool PreVisit(Match & /*unused*/) override {
scopes_.back().in_match = true;
return true;
}
bool PostVisit(Match & /*unused*/) override {
auto &scope = scopes_.back();
scope.in_match = false;
// Check variables in property maps after visiting Match, so that they can
// reference symbols out of bind order.
for (auto &ident : scope.identifiers_in_match) {
if (!HasSymbolLocalScope(ident->name_) && !ConsumePredefinedIdentifier(ident->name_))
throw UnboundVariableError(ident->name_);
ident->MapTo(scope.symbols[ident->name_]);
}
scope.identifiers_in_match.clear();
return true;
}
bool PreVisit(Foreach &for_each) override {
const auto &name = for_each.named_expression_->name_;
scopes_.emplace_back(Scope());
scopes_.back().in_foreach = true;
for_each.named_expression_->MapTo(
CreateSymbol(name, true, Symbol::Type::ANY, for_each.named_expression_->token_position_));
return true;
}
bool PostVisit(Foreach & /*unused*/) override {
scopes_.pop_back();
return true;
}
// Expressions
ReturnType Visit(Identifier &ident) override {
auto &scope = scopes_.back();
if (scope.in_skip || scope.in_limit) {
throw SemanticException("Variables are not allowed in {}.", scope.in_skip ? "SKIP" : "LIMIT");
}
Symbol symbol;
if (scope.in_pattern && !(scope.in_node_atom || scope.visiting_edge)) {
// If we are in the pattern, and outside of a node or an edge, the
// identifier is the pattern name.
symbol = GetOrCreateSymbolLocalScope(ident.name_, ident.user_declared_, Symbol::Type::PATH);
} else if (scope.in_pattern && scope.in_pattern_atom_identifier) {
// Patterns used to create nodes and edges cannot redeclare already
// established bindings. Declaration only happens in single node
// patterns and in edge patterns. OpenCypher example,
// `MATCH (n) CREATE (n)` should throw an error that `n` is already
// declared. While `MATCH (n) CREATE (n) -[:R]-> (n)` is allowed,
// since `n` now references the bound node instead of declaring it.
if ((scope.in_create_node || scope.in_create_edge) && HasSymbolLocalScope(ident.name_)) {
throw RedeclareVariableError(ident.name_);
}
auto type = Symbol::Type::VERTEX;
if (scope.visiting_edge) {
// Edge referencing is not allowed (like in Neo4j):
// `MATCH (n) - [r] -> (n) - [r] -> (n) RETURN r` is not allowed.
if (HasSymbolLocalScope(ident.name_)) {
throw RedeclareVariableError(ident.name_);
}
type = scope.visiting_edge->IsVariable() ? Symbol::Type::EDGE_LIST : Symbol::Type::EDGE;
}
symbol = GetOrCreateSymbolLocalScope(ident.name_, ident.user_declared_, type);
} else if (scope.in_pattern && !scope.in_pattern_atom_identifier && scope.in_match) {
if (scope.in_edge_range && scope.visiting_edge->identifier_->name_ == ident.name_) {
// Prevent variable path bounds to reference the identifier which is bound
// by the variable path itself.
throw UnboundVariableError(ident.name_);
}
// Variables in property maps or bounds of variable length path during MATCH
// can reference symbols bound later in the same MATCH. We collect them
// here, so that they can be checked after visiting Match.
scope.identifiers_in_match.emplace_back(&ident);
} else {
// Everything else references a bound symbol.
if (!HasSymbol(ident.name_) && !ConsumePredefinedIdentifier(ident.name_)) throw UnboundVariableError(ident.name_);
symbol = GetOrCreateSymbol(ident.name_, ident.user_declared_, Symbol::Type::ANY);
}
ident.MapTo(symbol);
return true;
}
ReturnType Visit(PrimitiveLiteral & /*unused*/) override { return true; }
ReturnType Visit(ParameterLookup & /*unused*/) override { return true; }
bool PreVisit(Aggregation &aggr) override {
auto &scope = scopes_.back();
// Check if the aggregation can be used in this context. This check should
// probably move to a separate phase, which checks if the query is well
// formed.
if ((!scope.in_return && !scope.in_with) || scope.in_order_by || scope.in_skip || scope.in_limit ||
scope.in_where) {
throw SemanticException("Aggregation functions are only allowed in WITH and RETURN.");
}
if (scope.in_aggregation) {
throw SemanticException(
"Using aggregation functions inside aggregation functions is not "
"allowed.");
}
if (scope.num_if_operators) {
// Neo allows aggregations here and produces very interesting behaviors.
// To simplify implementation at this moment we decided to completely
// disallow aggregations inside of the CASE.
// However, in some cases aggregation makes perfect sense, for example:
// CASE count(n) WHEN 10 THEN "YES" ELSE "NO" END.
// TODO: Rethink of allowing aggregations in some parts of the CASE
// construct.
throw SemanticException("Using aggregation functions inside of CASE is not allowed.");
}
// Create a virtual symbol for aggregation result.
// Currently, we only have aggregation operators which return numbers.
auto aggr_name = Aggregation::OpToString(aggr.op_) + std::to_string(aggr.symbol_pos_);
aggr.MapTo(CreateSymbol(aggr_name, false, Symbol::Type::NUMBER));
scope.in_aggregation = true;
scope.has_aggregation = true;
return true;
}
bool PostVisit(Aggregation & /*unused*/) override {
scopes_.back().in_aggregation = false;
return true;
}
bool PreVisit(IfOperator & /*unused*/) override {
++scopes_.back().num_if_operators;
return true;
}
bool PostVisit(IfOperator & /*unused*/) override {
--scopes_.back().num_if_operators;
return true;
}
bool PreVisit(All &all) override {
all.list_expression_->Accept(*this);
VisitWithIdentifiers(all.where_->expression_, {all.identifier_});
return false;
}
bool PreVisit(Single &single) override {
single.list_expression_->Accept(*this);
VisitWithIdentifiers(single.where_->expression_, {single.identifier_});
return false;
}
bool PreVisit(Any &any) override {
any.list_expression_->Accept(*this);
VisitWithIdentifiers(any.where_->expression_, {any.identifier_});
return false;
}
bool PreVisit(None &none) override {
none.list_expression_->Accept(*this);
VisitWithIdentifiers(none.where_->expression_, {none.identifier_});
return false;
}
bool PreVisit(Reduce &reduce) override {
reduce.initializer_->Accept(*this);
reduce.list_->Accept(*this);
VisitWithIdentifiers(reduce.expression_, {reduce.accumulator_, reduce.identifier_});
return false;
}
bool PreVisit(Extract &extract) override {
extract.list_->Accept(*this);
VisitWithIdentifiers(extract.expression_, {extract.identifier_});
return false;
}
// Pattern and its subparts.
bool PreVisit(Pattern &pattern) override {
auto &scope = scopes_.back();
scope.in_pattern = true;
if ((scope.in_create || scope.in_merge) && pattern.atoms_.size() == 1U) {
MG_ASSERT(utils::IsSubtype(*pattern.atoms_[0], NodeAtom::kType), "Expected a single NodeAtom in Pattern");
scope.in_create_node = true;
}
return true;
}
bool PostVisit(Pattern & /*unused*/) override {
auto &scope = scopes_.back();
scope.in_pattern = false;
scope.in_create_node = false;
return true;
}
bool PreVisit(NodeAtom &node_atom) override {
auto &scope = scopes_.back();
auto check_node_semantic = [&node_atom, &scope, this](const bool props_or_labels) {
const auto &node_name = node_atom.identifier_->name_;
if ((scope.in_create || scope.in_merge) && props_or_labels && HasSymbolLocalScope(node_name)) {
throw SemanticException("Cannot create node '" + node_name +
"' with labels or properties, because it is already declared.");
}
scope.in_pattern_atom_identifier = true;
node_atom.identifier_->Accept(*this);
scope.in_pattern_atom_identifier = false;
};
scope.in_node_atom = true;
if (auto *properties = std::get_if<std::unordered_map<PropertyIx, Expression *>>(&node_atom.properties_)) {
bool props_or_labels = !properties->empty() || !node_atom.labels_.empty();
check_node_semantic(props_or_labels);
for (auto kv : *properties) {
kv.second->Accept(*this);
}
return false;
}
auto &properties_parameter = std::get<ParameterLookup *>(node_atom.properties_);
bool props_or_labels = !properties_parameter || !node_atom.labels_.empty();
check_node_semantic(props_or_labels);
properties_parameter->Accept(*this);
return false;
}
bool PostVisit(NodeAtom & /*unused*/) override {
scopes_.back().in_node_atom = false;
return true;
}
bool PreVisit(EdgeAtom &edge_atom) override {
auto &scope = scopes_.back();
scope.visiting_edge = &edge_atom;
if (scope.in_create || scope.in_merge) {
scope.in_create_edge = true;
if (edge_atom.edge_types_.size() != 1U) {
throw SemanticException(
"A single relationship type must be specified "
"when creating an edge.");
}
if (scope.in_create && // Merge allows bidirectionality
edge_atom.direction_ == EdgeAtom::Direction::BOTH) {
throw SemanticException(
"Bidirectional relationship are not supported "
"when creating an edge");
}
if (edge_atom.IsVariable()) {
throw SemanticException(
"Variable length relationships are not supported when creating an "
"edge.");
}
}
if (auto *properties = std::get_if<std::unordered_map<PropertyIx, Expression *>>(&edge_atom.properties_)) {
for (auto kv : *properties) {
kv.second->Accept(*this);
}
} else {
std::get<ParameterLookup *>(edge_atom.properties_)->Accept(*this);
}
if (edge_atom.IsVariable()) {
scope.in_edge_range = true;
if (edge_atom.lower_bound_) {
edge_atom.lower_bound_->Accept(*this);
}
if (edge_atom.upper_bound_) {
edge_atom.upper_bound_->Accept(*this);
}
scope.in_edge_range = false;
scope.in_pattern = false;
if (edge_atom.filter_lambda_.expression) {
VisitWithIdentifiers(edge_atom.filter_lambda_.expression,
{edge_atom.filter_lambda_.inner_edge, edge_atom.filter_lambda_.inner_node});
} else {
// Create inner symbols, but don't bind them in scope, since they are to
// be used in the missing filter expression.
auto *inner_edge = edge_atom.filter_lambda_.inner_edge;
inner_edge->MapTo(
symbol_table_->CreateSymbol(inner_edge->name_, inner_edge->user_declared_, Symbol::Type::EDGE));
auto *inner_node = edge_atom.filter_lambda_.inner_node;
inner_node->MapTo(
symbol_table_->CreateSymbol(inner_node->name_, inner_node->user_declared_, Symbol::Type::VERTEX));
}
if (edge_atom.weight_lambda_.expression) {
VisitWithIdentifiers(edge_atom.weight_lambda_.expression,
{edge_atom.weight_lambda_.inner_edge, edge_atom.weight_lambda_.inner_node});
}
scope.in_pattern = true;
}
scope.in_pattern_atom_identifier = true;
edge_atom.identifier_->Accept(*this);
scope.in_pattern_atom_identifier = false;
if (edge_atom.total_weight_) {
if (HasSymbolLocalScope(edge_atom.total_weight_->name_)) {
throw RedeclareVariableError(edge_atom.total_weight_->name_);
}
edge_atom.total_weight_->MapTo(GetOrCreateSymbolLocalScope(
edge_atom.total_weight_->name_, edge_atom.total_weight_->user_declared_, Symbol::Type::NUMBER));
}
return false;
}
bool PostVisit(EdgeAtom & /*unused*/) override {
auto &scope = scopes_.back();
scope.visiting_edge = nullptr;
scope.in_create_edge = false;
return true;
}
private:
// Scope stores the state of where we are when visiting the AST and a map of
// names to symbols.
struct Scope {
bool in_pattern{false};
bool in_merge{false};
bool in_create{false};
// in_create_node is true if we are creating or merging *only* a node.
// Therefore, it is *not* equivalent to (in_create || in_merge) &&
// in_node_atom.
bool in_create_node{false};
// True if creating an edge;
// shortcut for (in_create || in_merge) && visiting_edge.
bool in_create_edge{false};
bool in_node_atom{false};
EdgeAtom *visiting_edge{nullptr};
bool in_aggregation{false};
bool in_return{false};
bool in_with{false};
bool in_skip{false};
bool in_limit{false};
bool in_order_by{false};
bool in_where{false};
bool in_match{false};
bool in_foreach{false};
// True when visiting a pattern atom (node or edge) identifier, which can be
// reused or created in the pattern itself.
bool in_pattern_atom_identifier{false};
// True when visiting range bounds of a variable path.
bool in_edge_range{false};
// True if the return/with contains an aggregation in any named expression.
bool has_aggregation{false};
// Map from variable names to symbols.
std::map<std::string, Symbol> symbols;
// Identifiers found in property maps of patterns or as variable length path
// bounds in a single Match clause. They need to be checked after visiting
// Match. Identifiers created by naming vertices, edges and paths are *not*
// stored in here.
std::vector<Identifier *> identifiers_in_match;
// Number of nested IfOperators.
int num_if_operators{0};
};
inline static std::optional<Symbol> FindSymbolInScope(const std::string &name, const Scope &scope,
Symbol::Type type) {
if (auto it = scope.symbols.find(name); it != scope.symbols.end()) {
const auto &symbol = it->second;
// Unless we have `ANY` type, check that types match.
if (type != Symbol::Type::ANY && symbol.type() != Symbol::Type::ANY && type != symbol.type()) {
throw TypeMismatchError(name, Symbol::TypeToString(symbol.type()), Symbol::TypeToString(type));
}
return symbol;
}
return std::nullopt;
}
bool HasSymbol(const std::string &name) const {
return std::ranges::any_of(scopes_, [&name](const auto &scope) { return scope.symbols.contains(name); });
}
bool HasSymbolLocalScope(const std::string &name) const { return scopes_.back().symbols.contains(name); }
// @return true if it added a predefined identifier with that name
bool ConsumePredefinedIdentifier(const std::string &name) {
auto it = predefined_identifiers_.find(name);
if (it == predefined_identifiers_.end()) {
return false;
}
// we can only use the predefined identifier in a single scope so we remove it after creating
// a symbol for it
auto &identifier = it->second;
MG_ASSERT(!identifier->user_declared_, "Predefined symbols cannot be user declared!");
identifier->MapTo(CreateSymbol(identifier->name_, identifier->user_declared_));
predefined_identifiers_.erase(it);
return true;
}
// Returns a freshly generated symbol. Previous mapping of the same name to a
// different symbol is replaced with the new one.
Symbol CreateSymbol(const std::string &name, bool user_declared, Symbol::Type type = Symbol::Type::ANY,
int token_position = -1) {
auto symbol = symbol_table_->CreateSymbol(name, user_declared, type, token_position);
scopes_.back().symbols[name] = symbol;
return symbol;
}
Symbol GetOrCreateSymbol(const std::string &name, bool user_declared, Symbol::Type type = Symbol::Type::ANY) {
// NOLINTNEXTLINE
for (auto scope = scopes_.rbegin(); scope != scopes_.rend(); ++scope) {
if (auto maybe_symbol = FindSymbolInScope(name, *scope, type); maybe_symbol) {
return *maybe_symbol;
}
}
return CreateSymbol(name, user_declared, type);
}
// Returns the symbol by name. If the mapping already exists, checks if the
// types match. Otherwise, returns a new symbol.
Symbol GetOrCreateSymbolLocalScope(const std::string &name, bool user_declared,
Symbol::Type type = Symbol::Type::ANY) {
auto &scope = scopes_.back();
if (auto maybe_symbol = FindSymbolInScope(name, scope, type); maybe_symbol) {
return *maybe_symbol;
}
return CreateSymbol(name, user_declared, type);
}
void VisitReturnBody(ReturnBody &body, Where *where = nullptr) {
auto &scope = scopes_.back();
for (auto &expr : body.named_expressions) {
expr->Accept(*this);
}
std::vector<Symbol> user_symbols;
if (body.all_identifiers) {
// Carry over user symbols because '*' appeared.
for (const auto &sym_pair : scope.symbols) {
if (!sym_pair.second.user_declared()) {
continue;
}
user_symbols.emplace_back(sym_pair.second);
}
if (user_symbols.empty()) {
throw SemanticException("There are no variables in scope to use for '*'.");
}
}
// WITH/RETURN clause removes declarations of all the previous variables and
// declares only those established through named expressions. New declarations
// must not be visible inside named expressions themselves.
bool removed_old_names = false;
if ((!where && body.order_by.empty()) || scope.has_aggregation) {
// WHERE and ORDER BY need to see both the old and new symbols, unless we
// have an aggregation. Therefore, we can clear the symbols immediately if
// there is neither ORDER BY nor WHERE, or we have an aggregation.
scope.symbols.clear();
removed_old_names = true;
}
// Create symbols for named expressions.
std::unordered_set<std::string> new_names;
for (const auto &user_sym : user_symbols) {
new_names.insert(user_sym.name());
scope.symbols[user_sym.name()] = user_sym;
}
for (auto &named_expr : body.named_expressions) {
const auto &name = named_expr->name_;
if (!new_names.insert(name).second) {
throw SemanticException("Multiple results with the same name '{}' are not allowed.", name);
}
// An improvement would be to infer the type of the expression, so that the
// new symbol would have a more specific type.
named_expr->MapTo(CreateSymbol(name, true, Symbol::Type::ANY, named_expr->token_position_));
}
scope.in_order_by = true;
for (const auto &order_pair : body.order_by) {
order_pair.expression->Accept(*this);
}
scope.in_order_by = false;
if (body.skip) {
scope.in_skip = true;
body.skip->Accept(*this);
scope.in_skip = false;
}
if (body.limit) {
scope.in_limit = true;
body.limit->Accept(*this);
scope.in_limit = false;
}
if (where) where->Accept(*this);
if (!removed_old_names) {
// We have an ORDER BY or WHERE, but no aggregation, which means we didn't
// clear the old symbols, so do it now. We cannot just call clear, because
// we've added new symbols.
for (auto sym_it = scope.symbols.begin(); sym_it != scope.symbols.end();) {
if (new_names.find(sym_it->first) == new_names.end()) {
sym_it = scope.symbols.erase(sym_it);
} else {
sym_it++;
}
}
}
scopes_.back().has_aggregation = false;
}
void VisitWithIdentifiers(Expression *expr, const std::vector<Identifier *> &identifiers) {
auto &scope = scopes_.back();
std::vector<std::pair<std::optional<Symbol>, Identifier *>> prev_symbols;
// Collect previous symbols if they exist.
for (const auto &identifier : identifiers) {
std::optional<Symbol> prev_symbol;
auto prev_symbol_it = scope.symbols.find(identifier->name_);
if (prev_symbol_it != scope.symbols.end()) {
prev_symbol = prev_symbol_it->second;
}
identifier->MapTo(CreateSymbol(identifier->name_, identifier->user_declared_));
prev_symbols.emplace_back(prev_symbol, identifier);
}
// Visit the expression with the new symbols bound.
expr->Accept(*this);
// Restore back to previous symbols.
for (const auto &prev : prev_symbols) {
const auto &prev_symbol = prev.first;
const auto &identifier = prev.second;
if (prev_symbol) {
scope.symbols[identifier->name_] = *prev_symbol;
} else {
scope.symbols.erase(identifier->name_);
}
}
}
SymbolTable *symbol_table_;
// Identifiers which are injected from outside the query. Each identifier
// is mapped by its name.
std::unordered_map<std::string, Identifier *> predefined_identifiers_;
std::vector<Scope> scopes_;
std::unordered_set<std::string> prev_return_names_;
std::unordered_set<std::string> curr_return_names_;
};
inline SymbolTable MakeSymbolTable(CypherQuery *query, const std::vector<Identifier *> &predefined_identifiers = {}) {
SymbolTable symbol_table;
SymbolGenerator symbol_generator(&symbol_table, predefined_identifiers);
query->single_query_->Accept(symbol_generator);
for (auto *cypher_union : query->cypher_unions_) {
cypher_union->Accept(symbol_generator);
}
return symbol_table;
}
} // namespace memgraph::expr

View File

@@ -0,0 +1,64 @@
// Copyright 2022 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.
#pragma once
#include <map>
#include <string>
#include "expr/ast.hpp"
#include "expr/semantic/symbol.hpp"
#include "utils/logging.hpp"
namespace memgraph::expr {
class SymbolTable final {
public:
SymbolTable() {}
const Symbol &CreateSymbol(const std::string &name, bool user_declared, Symbol::Type type = Symbol::Type::ANY,
int32_t token_position = -1) {
MG_ASSERT(table_.size() <= std::numeric_limits<int32_t>::max(),
"SymbolTable size doesn't fit into 32-bit integer!");
auto got = table_.emplace(position_, Symbol(name, position_, user_declared, type, token_position));
MG_ASSERT(got.second, "Duplicate symbol ID!");
position_++;
return got.first->second;
}
// TODO(buda): This is the same logic as in the cypher_main_visitor. During
// parsing phase symbol table doesn't exist. Figure out a better solution.
const Symbol &CreateAnonymousSymbol(Symbol::Type type = Symbol::Type::ANY) {
int id = 1;
while (true) {
static const std::string &kAnonPrefix = "anon";
std::string name_candidate = kAnonPrefix + std::to_string(id++);
if (std::find_if(std::begin(table_), std::end(table_), [&name_candidate](const auto &item) -> bool {
return item.second.name_ == name_candidate;
}) == std::end(table_)) {
return CreateSymbol(name_candidate, false, type);
}
}
}
const Symbol &at(const Identifier &ident) const { return table_.at(ident.symbol_pos_); }
const Symbol &at(const NamedExpression &nexpr) const { return table_.at(nexpr.symbol_pos_); }
const Symbol &at(const Aggregation &aggr) const { return table_.at(aggr.symbol_pos_); }
// TODO: Remove these since members are public
int32_t max_position() const { return static_cast<int32_t>(table_.size()); }
const auto &table() const { return table_; }
int32_t position_{0};
std::map<int32_t, Symbol> table_;
};
} // namespace memgraph::expr

1512
src/expr/typed_value.hpp Normal file

File diff suppressed because it is too large Load Diff

64
src/glue/v2/auth.cpp Normal file
View File

@@ -0,0 +1,64 @@
// Copyright 2022 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 "glue/v2/auth.hpp"
namespace memgraph::glue::v2 {
auth::Permission PrivilegeToPermission(query::v2::AuthQuery::Privilege privilege) {
switch (privilege) {
case query::v2::AuthQuery::Privilege::MATCH:
return auth::Permission::MATCH;
case query::v2::AuthQuery::Privilege::CREATE:
return auth::Permission::CREATE;
case query::v2::AuthQuery::Privilege::MERGE:
return auth::Permission::MERGE;
case query::v2::AuthQuery::Privilege::DELETE:
return auth::Permission::DELETE;
case query::v2::AuthQuery::Privilege::SET:
return auth::Permission::SET;
case query::v2::AuthQuery::Privilege::REMOVE:
return auth::Permission::REMOVE;
case query::v2::AuthQuery::Privilege::INDEX:
return auth::Permission::INDEX;
case query::v2::AuthQuery::Privilege::STATS:
return auth::Permission::STATS;
case query::v2::AuthQuery::Privilege::CONSTRAINT:
return auth::Permission::CONSTRAINT;
case query::v2::AuthQuery::Privilege::DUMP:
return auth::Permission::DUMP;
case query::v2::AuthQuery::Privilege::REPLICATION:
return auth::Permission::REPLICATION;
case query::v2::AuthQuery::Privilege::DURABILITY:
return auth::Permission::DURABILITY;
case query::v2::AuthQuery::Privilege::READ_FILE:
return auth::Permission::READ_FILE;
case query::v2::AuthQuery::Privilege::FREE_MEMORY:
return auth::Permission::FREE_MEMORY;
case query::v2::AuthQuery::Privilege::TRIGGER:
return auth::Permission::TRIGGER;
case query::v2::AuthQuery::Privilege::CONFIG:
return auth::Permission::CONFIG;
case query::v2::AuthQuery::Privilege::AUTH:
return auth::Permission::AUTH;
case query::v2::AuthQuery::Privilege::STREAM:
return auth::Permission::STREAM;
case query::v2::AuthQuery::Privilege::MODULE_READ:
return auth::Permission::MODULE_READ;
case query::v2::AuthQuery::Privilege::MODULE_WRITE:
return auth::Permission::MODULE_WRITE;
case query::v2::AuthQuery::Privilege::WEBSOCKET:
return auth::Permission::WEBSOCKET;
case query::v2::AuthQuery::Privilege::SCHEMA:
return auth::Permission::SCHEMA;
}
}
} // namespace memgraph::glue::v2

23
src/glue/v2/auth.hpp Normal file
View File

@@ -0,0 +1,23 @@
// Copyright 2022 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 "auth/models.hpp"
#include "query/v2/frontend/ast/ast.hpp"
namespace memgraph::glue::v2 {
/**
* This function converts query::AuthQuery::Privilege to its corresponding
* auth::Permission.
*/
auth::Permission PrivilegeToPermission(query::v2::AuthQuery::Privilege privilege);
} // namespace memgraph::glue::v2

View File

@@ -0,0 +1,277 @@
// Copyright 2022 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 "glue/v2/communication.hpp"
#include <map>
#include <string>
#include <vector>
#include "storage/v3/edge_accessor.hpp"
#include "storage/v3/shard.hpp"
#include "storage/v3/vertex_accessor.hpp"
#include "utils/temporal.hpp"
using memgraph::communication::bolt::Value;
namespace memgraph::glue::v2 {
query::v2::TypedValue ToTypedValue(const Value &value) {
switch (value.type()) {
case Value::Type::Null:
return {};
case Value::Type::Bool:
return query::v2::TypedValue(value.ValueBool());
case Value::Type::Int:
return query::v2::TypedValue(value.ValueInt());
case Value::Type::Double:
return query::v2::TypedValue(value.ValueDouble());
case Value::Type::String:
return query::v2::TypedValue(value.ValueString());
case Value::Type::List: {
std::vector<query::v2::TypedValue> list;
list.reserve(value.ValueList().size());
for (const auto &v : value.ValueList()) list.push_back(ToTypedValue(v));
return query::v2::TypedValue(std::move(list));
}
case Value::Type::Map: {
std::map<std::string, query::v2::TypedValue> map;
for (const auto &kv : value.ValueMap()) map.emplace(kv.first, ToTypedValue(kv.second));
return query::v2::TypedValue(std::move(map));
}
case Value::Type::Vertex:
case Value::Type::Edge:
case Value::Type::UnboundedEdge:
case Value::Type::Path:
throw communication::bolt::ValueException("Unsupported conversion from Value to TypedValue");
case Value::Type::Date:
return query::v2::TypedValue(value.ValueDate());
case Value::Type::LocalTime:
return query::v2::TypedValue(value.ValueLocalTime());
case Value::Type::LocalDateTime:
return query::v2::TypedValue(value.ValueLocalDateTime());
case Value::Type::Duration:
return query::v2::TypedValue(value.ValueDuration());
}
}
storage::v3::Result<communication::bolt::Vertex> ToBoltVertex(const query::v2::VertexAccessor &vertex,
const storage::v3::Shard &db, storage::v3::View view) {
return ToBoltVertex(vertex.impl_, db, view);
}
storage::v3::Result<communication::bolt::Edge> ToBoltEdge(const query::v2::EdgeAccessor &edge,
const storage::v3::Shard &db, storage::v3::View view) {
return ToBoltEdge(edge.impl_, db, view);
}
storage::v3::Result<Value> ToBoltValue(const query::v2::TypedValue &value, const storage::v3::Shard &db,
storage::v3::View view) {
switch (value.type()) {
case query::v2::TypedValue::Type::Null:
return Value();
case query::v2::TypedValue::Type::Bool:
return Value(value.ValueBool());
case query::v2::TypedValue::Type::Int:
return Value(value.ValueInt());
case query::v2::TypedValue::Type::Double:
return Value(value.ValueDouble());
case query::v2::TypedValue::Type::String:
return Value(std::string(value.ValueString()));
case query::v2::TypedValue::Type::List: {
std::vector<Value> values;
values.reserve(value.ValueList().size());
for (const auto &v : value.ValueList()) {
auto maybe_value = ToBoltValue(v, db, view);
if (maybe_value.HasError()) return maybe_value.GetError();
values.emplace_back(std::move(*maybe_value));
}
return Value(std::move(values));
}
case query::v2::TypedValue::Type::Map: {
std::map<std::string, Value> map;
for (const auto &kv : value.ValueMap()) {
auto maybe_value = ToBoltValue(kv.second, db, view);
if (maybe_value.HasError()) return maybe_value.GetError();
map.emplace(kv.first, std::move(*maybe_value));
}
return Value(std::move(map));
}
case query::v2::TypedValue::Type::Vertex: {
auto maybe_vertex = ToBoltVertex(value.ValueVertex(), db, view);
if (maybe_vertex.HasError()) return maybe_vertex.GetError();
return Value(std::move(*maybe_vertex));
}
case query::v2::TypedValue::Type::Edge: {
auto maybe_edge = ToBoltEdge(value.ValueEdge(), db, view);
if (maybe_edge.HasError()) return maybe_edge.GetError();
return Value(std::move(*maybe_edge));
}
case query::v2::TypedValue::Type::Path: {
auto maybe_path = ToBoltPath(value.ValuePath(), db, view);
if (maybe_path.HasError()) return maybe_path.GetError();
return Value(std::move(*maybe_path));
}
case query::v2::TypedValue::Type::Date:
return Value(value.ValueDate());
case query::v2::TypedValue::Type::LocalTime:
return Value(value.ValueLocalTime());
case query::v2::TypedValue::Type::LocalDateTime:
return Value(value.ValueLocalDateTime());
case query::v2::TypedValue::Type::Duration:
return Value(value.ValueDuration());
}
}
storage::v3::Result<communication::bolt::Vertex> ToBoltVertex(const storage::v3::VertexAccessor &vertex,
const storage::v3::Shard &db, storage::v3::View view) {
// TODO(jbajic) Fix bolt communication
auto id = communication::bolt::Id::FromUint(0);
auto maybe_labels = vertex.Labels(view);
if (maybe_labels.HasError()) return maybe_labels.GetError();
std::vector<std::string> labels;
labels.reserve(maybe_labels->size());
for (const auto &label : *maybe_labels) {
labels.push_back(db.LabelToName(label));
}
auto maybe_properties = vertex.Properties(view);
if (maybe_properties.HasError()) return maybe_properties.GetError();
std::map<std::string, Value> properties;
for (const auto &prop : *maybe_properties) {
properties[db.PropertyToName(prop.first)] = ToBoltValue(prop.second);
}
return communication::bolt::Vertex{id, labels, properties};
}
storage::v3::Result<communication::bolt::Edge> ToBoltEdge(const storage::v3::EdgeAccessor &edge,
const storage::v3::Shard &db, storage::v3::View view) {
// TODO(jbajic) Fix bolt communication
auto id = communication::bolt::Id::FromUint(0);
auto from = communication::bolt::Id::FromUint(0);
auto to = communication::bolt::Id::FromUint(0);
const auto &type = db.EdgeTypeToName(edge.EdgeType());
auto maybe_properties = edge.Properties(view);
if (maybe_properties.HasError()) return maybe_properties.GetError();
std::map<std::string, Value> properties;
for (const auto &prop : *maybe_properties) {
properties[db.PropertyToName(prop.first)] = ToBoltValue(prop.second);
}
return communication::bolt::Edge{id, from, to, type, properties};
}
storage::v3::Result<communication::bolt::Path> ToBoltPath(const query::v2::Path &path, const storage::v3::Shard &db,
storage::v3::View view) {
std::vector<communication::bolt::Vertex> vertices;
vertices.reserve(path.vertices().size());
for (const auto &v : path.vertices()) {
auto maybe_vertex = ToBoltVertex(v, db, view);
if (maybe_vertex.HasError()) return maybe_vertex.GetError();
vertices.emplace_back(std::move(*maybe_vertex));
}
std::vector<communication::bolt::Edge> edges;
edges.reserve(path.edges().size());
for (const auto &e : path.edges()) {
auto maybe_edge = ToBoltEdge(e, db, view);
if (maybe_edge.HasError()) return maybe_edge.GetError();
edges.emplace_back(std::move(*maybe_edge));
}
return communication::bolt::Path(vertices, edges);
}
storage::v3::PropertyValue ToPropertyValue(const Value &value) {
switch (value.type()) {
case Value::Type::Null:
return {};
case Value::Type::Bool:
return storage::v3::PropertyValue(value.ValueBool());
case Value::Type::Int:
return storage::v3::PropertyValue(value.ValueInt());
case Value::Type::Double:
return storage::v3::PropertyValue(value.ValueDouble());
case Value::Type::String:
return storage::v3::PropertyValue(value.ValueString());
case Value::Type::List: {
std::vector<storage::v3::PropertyValue> vec;
vec.reserve(value.ValueList().size());
for (const auto &value : value.ValueList()) vec.emplace_back(ToPropertyValue(value));
return storage::v3::PropertyValue(std::move(vec));
}
case Value::Type::Map: {
std::map<std::string, storage::v3::PropertyValue> map;
for (const auto &kv : value.ValueMap()) map.emplace(kv.first, ToPropertyValue(kv.second));
return storage::v3::PropertyValue(std::move(map));
}
case Value::Type::Vertex:
case Value::Type::Edge:
case Value::Type::UnboundedEdge:
case Value::Type::Path:
throw communication::bolt::ValueException("Unsupported conversion from Value to PropertyValue");
case Value::Type::Date:
return storage::v3::PropertyValue(
storage::v3::TemporalData(storage::v3::TemporalType::Date, value.ValueDate().MicrosecondsSinceEpoch()));
case Value::Type::LocalTime:
return storage::v3::PropertyValue(storage::v3::TemporalData(storage::v3::TemporalType::LocalTime,
value.ValueLocalTime().MicrosecondsSinceEpoch()));
case Value::Type::LocalDateTime:
return storage::v3::PropertyValue(storage::v3::TemporalData(storage::v3::TemporalType::LocalDateTime,
value.ValueLocalDateTime().MicrosecondsSinceEpoch()));
case Value::Type::Duration:
return storage::v3::PropertyValue(
storage::v3::TemporalData(storage::v3::TemporalType::Duration, value.ValueDuration().microseconds));
}
}
Value ToBoltValue(const storage::v3::PropertyValue &value) {
switch (value.type()) {
case storage::v3::PropertyValue::Type::Null:
return {};
case storage::v3::PropertyValue::Type::Bool:
return {value.ValueBool()};
case storage::v3::PropertyValue::Type::Int:
return {value.ValueInt()};
break;
case storage::v3::PropertyValue::Type::Double:
return {value.ValueDouble()};
case storage::v3::PropertyValue::Type::String:
return {value.ValueString()};
case storage::v3::PropertyValue::Type::List: {
const auto &values = value.ValueList();
std::vector<Value> vec;
vec.reserve(values.size());
for (const auto &v : values) {
vec.push_back(ToBoltValue(v));
}
return {std::move(vec)};
}
case storage::v3::PropertyValue::Type::Map: {
const auto &map = value.ValueMap();
std::map<std::string, Value> dv_map;
for (const auto &kv : map) {
dv_map.emplace(kv.first, ToBoltValue(kv.second));
}
return {std::move(dv_map)};
}
case storage::v3::PropertyValue::Type::TemporalData:
const auto &type = value.ValueTemporalData();
switch (type.type) {
case storage::v3::TemporalType::Date:
return {utils::Date(type.microseconds)};
case storage::v3::TemporalType::LocalTime:
return {utils::LocalTime(type.microseconds)};
case storage::v3::TemporalType::LocalDateTime:
return {utils::LocalDateTime(type.microseconds)};
case storage::v3::TemporalType::Duration:
return {utils::Duration(type.microseconds)};
}
}
}
} // namespace memgraph::glue::v2

View File

@@ -0,0 +1,68 @@
// Copyright 2022 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.
/// @file Conversion functions between Value and other memgraph types.
#pragma once
#include "communication/bolt/v1/value.hpp"
#include "query/v2/bindings/typed_value.hpp"
#include "storage/v3/property_value.hpp"
#include "storage/v3/result.hpp"
#include "storage/v3/view.hpp"
namespace memgraph::storage::v3 {
class EdgeAccessor;
class Storage;
class VertexAccessor;
} // namespace memgraph::storage::v3
namespace memgraph::glue::v2 {
/// @param storage::v3::VertexAccessor for converting to
/// communication::bolt::Vertex.
/// @param storage::v3::Shard for getting label and property names.
/// @param storage::v3::View for deciding which vertex attributes are visible.
///
/// @throw std::bad_alloc
storage::v3::Result<communication::bolt::Vertex> ToBoltVertex(const storage::v3::VertexAccessor &vertex,
const storage::v3::Shard &db, storage::v3::View view);
/// @param storage::v3::EdgeAccessor for converting to communication::bolt::Edge.
/// @param storage::v3::Shard for getting edge type and property names.
/// @param storage::v3::View for deciding which edge attributes are visible.
///
/// @throw std::bad_alloc
storage::v3::Result<communication::bolt::Edge> ToBoltEdge(const storage::v3::EdgeAccessor &edge,
const storage::v3::Shard &db, storage::v3::View view);
/// @param query::v2::Path for converting to communication::bolt::Path.
/// @param storage::v3::Shard for ToBoltVertex and ToBoltEdge.
/// @param storage::v3::View for ToBoltVertex and ToBoltEdge.
///
/// @throw std::bad_alloc
storage::v3::Result<communication::bolt::Path> ToBoltPath(const query::v2::Path &path, const storage::v3::Shard &db,
storage::v3::View view);
/// @param query::v2::TypedValue for converting to communication::bolt::Value.
/// @param storage::v3::Shard for ToBoltVertex and ToBoltEdge.
/// @param storage::v3::View for ToBoltVertex and ToBoltEdge.
///
/// @throw std::bad_alloc
storage::v3::Result<communication::bolt::Value> ToBoltValue(const query::v2::TypedValue &value,
const storage::v3::Shard &db, storage::v3::View view);
query::v2::TypedValue ToTypedValue(const communication::bolt::Value &value);
communication::bolt::Value ToBoltValue(const storage::v3::PropertyValue &value);
storage::v3::PropertyValue ToPropertyValue(const communication::bolt::Value &value);
} // namespace memgraph::glue::v2

77
src/io/address.hpp Normal file
View File

@@ -0,0 +1,77 @@
// Copyright 2022 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.
#pragma once
#include <compare>
#include <fmt/format.h>
#include <boost/asio/ip/tcp.hpp>
#include <boost/uuid/uuid.hpp>
#include <boost/uuid/uuid_generators.hpp>
#include <boost/uuid/uuid_io.hpp>
namespace memgraph::io {
struct Address {
// It's important for all participants to have a
// unique identifier - IP and port alone are not
// enough, and may change over the lifecycle of
// the nodes. Particularly storage nodes may change
// their IP addresses over time, and the system
// should gracefully update its information
// about them.
boost::uuids::uuid unique_id;
boost::asio::ip::address last_known_ip;
uint16_t last_known_port;
static Address TestAddress(uint16_t port) {
return Address{
.unique_id = boost::uuids::uuid{boost::uuids::random_generator()()},
.last_known_port = port,
};
}
static Address UniqueLocalAddress() {
return Address{
.unique_id = boost::uuids::uuid{boost::uuids::random_generator()()},
};
}
/// Returns a new ID with the same IP and port but a unique UUID.
Address ForkUniqueAddress() {
return Address{
.unique_id = boost::uuids::uuid{boost::uuids::random_generator()()},
.last_known_ip = last_known_ip,
.last_known_port = last_known_port,
};
}
friend bool operator==(const Address &lhs, const Address &rhs) = default;
/// unique_id is most dominant for ordering, then last_known_ip, then last_known_port
friend bool operator<(const Address &lhs, const Address &rhs) {
if (lhs.unique_id != rhs.unique_id) {
return lhs.unique_id < rhs.unique_id;
}
if (lhs.last_known_ip != rhs.last_known_ip) {
return lhs.last_known_ip < rhs.last_known_ip;
}
return lhs.last_known_port < rhs.last_known_port;
}
std::string ToString() const {
return fmt::format("Address {{ unique_id: {}, last_known_ip: {}, last_known_port: {} }}",
boost::uuids::to_string(unique_id), last_known_ip.to_string(), last_known_port);
}
};
}; // namespace memgraph::io

26
src/io/errors.hpp Normal file
View File

@@ -0,0 +1,26 @@
// Copyright 2022 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.
#pragma once
namespace memgraph::io {
// Signifies that a retriable operation was unable to
// complete after a configured number of retries.
struct RetriesExhausted {};
// Signifies that a request was unable to receive a response
// within some configured timeout duration. It is important
// to remember that in distributed systems, a timeout does
// not signify that a request was not received or processed.
// It may be the case that the request was fully processed
// but that the response was not received.
struct TimedOut {};
}; // namespace memgraph::io

262
src/io/future.hpp Normal file
View File

@@ -0,0 +1,262 @@
// Copyright 2022 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.
#pragma once
#include <condition_variable>
#include <memory>
#include <mutex>
#include <optional>
#include <thread>
#include <utility>
#include "io/errors.hpp"
#include "utils/logging.hpp"
namespace memgraph::io {
// Shared is in an anonymous namespace, and the only way to
// construct a Promise or Future is to pass a Shared in. This
// ensures that Promises and Futures can only be constructed
// in this translation unit.
namespace details {
template <typename T>
class Shared {
mutable std::condition_variable cv_;
mutable std::mutex mu_;
std::optional<T> item_;
bool consumed_ = false;
bool waiting_ = false;
std::function<bool()> simulator_notifier_ = nullptr;
public:
explicit Shared(std::function<bool()> simulator_notifier) : simulator_notifier_(simulator_notifier) {}
Shared() = default;
Shared(Shared &&) = delete;
Shared &operator=(Shared &&) = delete;
Shared(const Shared &) = delete;
Shared &operator=(const Shared &) = delete;
~Shared() = default;
/// Takes the item out of our optional item_ and returns it.
T Take() {
MG_ASSERT(item_, "Take called without item_ being present");
MG_ASSERT(!consumed_, "Take called on already-consumed Future");
T ret = std::move(item_).value();
item_.reset();
consumed_ = true;
return ret;
}
T Wait() {
std::unique_lock<std::mutex> lock(mu_);
waiting_ = true;
while (!item_) {
bool simulator_progressed = false;
if (simulator_notifier_) [[unlikely]] {
// We can't hold our own lock while notifying
// the simulator because notifying the simulator
// involves acquiring the simulator's mutex
// to guarantee that our notification linearizes
// with the simulator's condition variable.
// However, the simulator may acquire our
// mutex to check if we are being awaited,
// while determining system quiescence,
// so we have to get out of its way to avoid
// a cyclical deadlock.
lock.unlock();
simulator_progressed = std::invoke(simulator_notifier_);
lock.lock();
if (item_) {
// item may have been filled while we
// had dropped our mutex while notifying
// the simulator of our waiting_ status.
break;
}
}
if (!simulator_progressed) [[likely]] {
cv_.wait(lock);
}
MG_ASSERT(!consumed_, "Future consumed twice!");
}
waiting_ = false;
return Take();
}
bool IsReady() const {
std::unique_lock<std::mutex> lock(mu_);
return item_;
}
std::optional<T> TryGet() {
std::unique_lock<std::mutex> lock(mu_);
if (item_) {
return Take();
}
return std::nullopt;
}
void Fill(T item) {
{
std::unique_lock<std::mutex> lock(mu_);
MG_ASSERT(!consumed_, "Promise filled after it was already consumed!");
MG_ASSERT(!item_, "Promise filled twice!");
item_ = item;
} // lock released before condition variable notification
cv_.notify_all();
}
bool IsAwaited() const {
std::unique_lock<std::mutex> lock(mu_);
return waiting_;
}
};
} // namespace details
template <typename T>
class Future {
bool consumed_or_moved_ = false;
std::shared_ptr<details::Shared<T>> shared_;
public:
explicit Future(std::shared_ptr<details::Shared<T>> shared) : shared_(shared) {}
Future() = delete;
Future(Future &&old) noexcept {
MG_ASSERT(!old.consumed_or_moved_, "Future moved from after already being moved from or consumed.");
shared_ = std::move(old.shared_);
consumed_or_moved_ = old.consumed_or_moved_;
old.consumed_or_moved_ = true;
}
Future &operator=(Future &&old) noexcept {
MG_ASSERT(!old.consumed_or_moved_, "Future moved from after already being moved from or consumed.");
shared_ = std::move(old.shared_);
old.consumed_or_moved_ = true;
}
Future(const Future &) = delete;
Future &operator=(const Future &) = delete;
~Future() = default;
/// Returns true if the Future is ready to
/// be consumed using TryGet or Wait (prefer Wait
/// if you know it's ready, because it doesn't
/// return an optional.
bool IsReady() {
MG_ASSERT(!consumed_or_moved_, "Called IsReady after Future already consumed!");
return shared_->IsReady();
}
/// Non-blocking method that returns the inner
/// item if it's already ready, or std::nullopt
/// if it is not ready yet.
std::optional<T> TryGet() {
MG_ASSERT(!consumed_or_moved_, "Called TryGet after Future already consumed!");
std::optional<T> ret = shared_->TryGet();
if (ret) {
consumed_or_moved_ = true;
}
return ret;
}
/// Block on the corresponding promise to be filled,
/// returning the inner item when ready.
T Wait() && {
MG_ASSERT(!consumed_or_moved_, "Future should only be consumed with Wait once!");
T ret = shared_->Wait();
consumed_or_moved_ = true;
return ret;
}
/// Marks this Future as canceled.
void Cancel() {
MG_ASSERT(!consumed_or_moved_, "Future::Cancel called on a future that was already moved or consumed!");
consumed_or_moved_ = true;
}
};
template <typename T>
class Promise {
std::shared_ptr<details::Shared<T>> shared_;
bool filled_or_moved_ = false;
public:
explicit Promise(std::shared_ptr<details::Shared<T>> shared) : shared_(shared) {}
Promise() = delete;
Promise(Promise &&old) noexcept {
MG_ASSERT(!old.filled_or_moved_, "Promise moved from after already being moved from or filled.");
shared_ = std::move(old.shared_);
old.filled_or_moved_ = true;
}
Promise &operator=(Promise &&old) noexcept {
MG_ASSERT(!old.filled_or_moved_, "Promise moved from after already being moved from or filled.");
shared_ = std::move(old.shared_);
old.filled_or_moved_ = true;
}
Promise(const Promise &) = delete;
Promise &operator=(const Promise &) = delete;
~Promise() { MG_ASSERT(filled_or_moved_, "Promise destroyed before its associated Future was filled!"); }
// Fill the expected item into the Future.
void Fill(T item) {
MG_ASSERT(!filled_or_moved_, "Promise::Fill called on a promise that is already filled or moved!");
shared_->Fill(item);
filled_or_moved_ = true;
}
bool IsAwaited() { return shared_->IsAwaited(); }
/// Moves this Promise into a unique_ptr.
std::unique_ptr<Promise<T>> ToUnique() && {
std::unique_ptr<Promise<T>> up = std::make_unique<Promise<T>>(std::move(shared_));
filled_or_moved_ = true;
return up;
}
};
template <typename T>
std::pair<Future<T>, Promise<T>> FuturePromisePair() {
std::shared_ptr<details::Shared<T>> shared = std::make_shared<details::Shared<T>>();
Future<T> future = Future<T>(shared);
Promise<T> promise = Promise<T>(shared);
return std::make_pair(std::move(future), std::move(promise));
}
template <typename T>
std::pair<Future<T>, Promise<T>> FuturePromisePairWithNotifier(std::function<bool()> simulator_notifier) {
std::shared_ptr<details::Shared<T>> shared = std::make_shared<details::Shared<T>>(simulator_notifier);
Future<T> future = Future<T>(shared);
Promise<T> promise = Promise<T>(shared);
return std::make_pair(std::move(future), std::move(promise));
}
}; // namespace memgraph::io

View File

@@ -0,0 +1,35 @@
// Copyright 2022 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.
#pragma once
#include <random>
#include "io/address.hpp"
#include "io/local_transport/local_transport.hpp"
#include "io/local_transport/local_transport_handle.hpp"
#include "io/transport.hpp"
namespace memgraph::io::local_transport {
class LocalSystem {
std::shared_ptr<LocalTransportHandle> local_transport_handle_ = std::make_shared<LocalTransportHandle>();
public:
Io<LocalTransport> Register(Address address) {
LocalTransport local_transport(local_transport_handle_, address);
return Io{local_transport, address};
}
void ShutDown() { local_transport_handle_->ShutDown(); }
};
} // namespace memgraph::io::local_transport

View File

@@ -0,0 +1,67 @@
// Copyright 2022 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.
#pragma once
#include <chrono>
#include <memory>
#include <random>
#include <utility>
#include "io/address.hpp"
#include "io/local_transport/local_transport_handle.hpp"
#include "io/time.hpp"
#include "io/transport.hpp"
namespace memgraph::io::local_transport {
class LocalTransport {
std::shared_ptr<LocalTransportHandle> local_transport_handle_;
const Address address_;
public:
LocalTransport(std::shared_ptr<LocalTransportHandle> local_transport_handle, Address address)
: local_transport_handle_(std::move(local_transport_handle)), address_(address) {}
template <Message RequestT, Message ResponseT>
ResponseFuture<ResponseT> Request(Address to_address, RequestId request_id, RequestT request, Duration timeout) {
auto [future, promise] = memgraph::io::FuturePromisePair<ResponseResult<ResponseT>>();
Address from_address = address_;
local_transport_handle_->SubmitRequest(to_address, from_address, request_id, std::move(request), timeout,
std::move(promise));
return std::move(future);
}
template <Message... Ms>
requires(sizeof...(Ms) > 0) RequestResult<Ms...> Receive(Duration timeout) {
Address from_address = address_;
return local_transport_handle_->template Receive<Ms...>(timeout);
}
template <Message M>
void Send(Address to_address, Address from_address, RequestId request_id, M &&message) {
return local_transport_handle_->template Send<M>(to_address, from_address, request_id, std::forward<M>(message));
}
Time Now() const { return local_transport_handle_->Now(); }
bool ShouldShutDown() const { return local_transport_handle_->ShouldShutDown(); }
template <class D = std::poisson_distribution<>, class Return = uint64_t>
Return Rand(D distrib) {
std::random_device rng;
return distrib(rng);
}
};
}; // namespace memgraph::io::local_transport

View File

@@ -0,0 +1,139 @@
// Copyright 2022 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.
#pragma once
#include <chrono>
#include <condition_variable>
#include <iostream>
#include <map>
#include <mutex>
#include "io/errors.hpp"
#include "io/message_conversion.hpp"
#include "io/time.hpp"
#include "io/transport.hpp"
namespace memgraph::io::local_transport {
class LocalTransportHandle {
mutable std::mutex mu_{};
mutable std::condition_variable cv_;
bool should_shut_down_ = false;
// the responses to requests that are being waited on
std::map<PromiseKey, DeadlineAndOpaquePromise> promises_;
// messages that are sent to servers that may later receive them
std::vector<OpaqueMessage> can_receive_;
public:
void ShutDown() {
std::unique_lock<std::mutex> lock(mu_);
should_shut_down_ = true;
cv_.notify_all();
}
bool ShouldShutDown() const {
std::unique_lock<std::mutex> lock(mu_);
return should_shut_down_;
}
static Time Now() {
auto nano_time = std::chrono::system_clock::now();
return std::chrono::time_point_cast<std::chrono::microseconds>(nano_time);
}
template <Message... Ms>
requires(sizeof...(Ms) > 0) RequestResult<Ms...> Receive(Duration timeout) {
std::unique_lock lock(mu_);
Time before = Now();
while (can_receive_.empty()) {
Time now = Now();
// protection against non-monotonic timesources
auto maxed_now = std::max(now, before);
auto elapsed = maxed_now - before;
if (timeout < elapsed) {
return TimedOut{};
}
Duration relative_timeout = timeout - elapsed;
std::cv_status cv_status_value = cv_.wait_for(lock, relative_timeout);
if (cv_status_value == std::cv_status::timeout) {
return TimedOut{};
}
}
auto current_message = std::move(can_receive_.back());
can_receive_.pop_back();
auto m_opt = std::move(current_message).Take<Ms...>();
return std::move(m_opt).value();
}
template <Message M>
void Send(Address to_address, Address from_address, RequestId request_id, M &&message) {
std::any message_any(std::forward<M>(message));
OpaqueMessage opaque_message{
.from_address = from_address, .request_id = request_id, .message = std::move(message_any)};
PromiseKey promise_key{.requester_address = to_address,
.request_id = opaque_message.request_id,
.replier_address = opaque_message.from_address};
{
std::unique_lock<std::mutex> lock(mu_);
if (promises_.contains(promise_key)) {
// complete waiting promise if it's there
DeadlineAndOpaquePromise dop = std::move(promises_.at(promise_key));
promises_.erase(promise_key);
dop.promise.Fill(std::move(opaque_message));
} else {
can_receive_.emplace_back(std::move(opaque_message));
}
} // lock dropped
cv_.notify_all();
}
template <Message RequestT, Message ResponseT>
void SubmitRequest(Address to_address, Address from_address, RequestId request_id, RequestT &&request,
Duration timeout, ResponsePromise<ResponseT> promise) {
const bool port_matches = to_address.last_known_port == from_address.last_known_port;
const bool ip_matches = to_address.last_known_ip == from_address.last_known_ip;
MG_ASSERT(port_matches && ip_matches);
const Time deadline = Now() + timeout;
{
std::unique_lock<std::mutex> lock(mu_);
PromiseKey promise_key{
.requester_address = from_address, .request_id = request_id, .replier_address = to_address};
OpaquePromise opaque_promise(std::move(promise).ToUnique());
DeadlineAndOpaquePromise dop{.deadline = deadline, .promise = std::move(opaque_promise)};
promises_.emplace(std::move(promise_key), std::move(dop));
} // lock dropped
Send(to_address, from_address, request_id, std::forward<RequestT>(request));
}
};
} // namespace memgraph::io::local_transport

View File

@@ -0,0 +1,206 @@
// Copyright 2022 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.
#pragma once
#include "io/transport.hpp"
namespace memgraph::io {
using memgraph::io::Duration;
using memgraph::io::Message;
using memgraph::io::Time;
struct PromiseKey {
Address requester_address;
uint64_t request_id;
// TODO(tyler) possibly remove replier_address from promise key
// once we want to support DSR.
Address replier_address;
public:
friend bool operator<(const PromiseKey &lhs, const PromiseKey &rhs) {
if (lhs.requester_address != rhs.requester_address) {
return lhs.requester_address < rhs.requester_address;
}
if (lhs.request_id != rhs.request_id) {
return lhs.request_id < rhs.request_id;
}
return lhs.replier_address < rhs.replier_address;
}
};
struct OpaqueMessage {
Address to_address;
Address from_address;
uint64_t request_id;
std::any message;
/// Recursively tries to match a specific type from the outer
/// variant's parameter pack against the type of the std::any,
/// and if it matches, make it concrete and return it. Otherwise,
/// move on and compare the any with the next type from the
/// parameter pack.
///
/// Return is the full std::variant<Ts...> type that holds the
/// full parameter pack without interfering with recursive
/// narrowing expansion.
template <typename Return, Message Head, Message... Rest>
std::optional<Return> Unpack(std::any &&a) {
if (typeid(Head) == a.type()) {
Head concrete = std::any_cast<Head>(std::move(a));
return concrete;
}
if constexpr (sizeof...(Rest) > 0) {
return Unpack<Return, Rest...>(std::move(a));
} else {
return std::nullopt;
}
}
/// High level "user-facing" conversion function that lets
/// people interested in conversion only supply a single
/// parameter pack for the types that they want to compare
/// with the any and potentially include in the returned
/// variant.
template <Message... Ms>
requires(sizeof...(Ms) > 0) std::optional<std::variant<Ms...>> VariantFromAny(std::any &&a) {
return Unpack<std::variant<Ms...>, Ms...>(std::move(a));
}
template <Message... Ms>
requires(sizeof...(Ms) > 0) std::optional<RequestEnvelope<Ms...>> Take() && {
std::optional<std::variant<Ms...>> m_opt = VariantFromAny<Ms...>(std::move(message));
if (m_opt) {
return RequestEnvelope<Ms...>{
.message = std::move(*m_opt),
.request_id = request_id,
.to_address = to_address,
.from_address = from_address,
};
}
return std::nullopt;
}
};
class OpaquePromiseTraitBase {
public:
virtual const std::type_info *TypeInfo() const = 0;
virtual bool IsAwaited(void *ptr) const = 0;
virtual void Fill(void *ptr, OpaqueMessage &&) const = 0;
virtual void TimeOut(void *ptr) const = 0;
virtual ~OpaquePromiseTraitBase() = default;
OpaquePromiseTraitBase() = default;
OpaquePromiseTraitBase(const OpaquePromiseTraitBase &) = delete;
OpaquePromiseTraitBase &operator=(const OpaquePromiseTraitBase &) = delete;
OpaquePromiseTraitBase(OpaquePromiseTraitBase &&old) = delete;
OpaquePromiseTraitBase &operator=(OpaquePromiseTraitBase &&) = delete;
};
template <typename T>
class OpaquePromiseTrait : public OpaquePromiseTraitBase {
public:
const std::type_info *TypeInfo() const override { return &typeid(T); };
bool IsAwaited(void *ptr) const override { return static_cast<ResponsePromise<T> *>(ptr)->IsAwaited(); };
void Fill(void *ptr, OpaqueMessage &&opaque_message) const override {
T message = std::any_cast<T>(std::move(opaque_message.message));
auto response_envelope = ResponseEnvelope<T>{.message = std::move(message),
.request_id = opaque_message.request_id,
.to_address = opaque_message.to_address,
.from_address = opaque_message.from_address};
auto promise = static_cast<ResponsePromise<T> *>(ptr);
auto unique_promise = std::unique_ptr<ResponsePromise<T>>(promise);
unique_promise->Fill(std::move(response_envelope));
};
void TimeOut(void *ptr) const override {
auto promise = static_cast<ResponsePromise<T> *>(ptr);
auto unique_promise = std::unique_ptr<ResponsePromise<T>>(promise);
ResponseResult<T> result = TimedOut{};
unique_promise->Fill(std::move(result));
}
};
class OpaquePromise {
void *ptr_;
std::unique_ptr<OpaquePromiseTraitBase> trait_;
public:
OpaquePromise(OpaquePromise &&old) noexcept : ptr_(old.ptr_), trait_(std::move(old.trait_)) {
MG_ASSERT(old.ptr_ != nullptr);
old.ptr_ = nullptr;
}
OpaquePromise &operator=(OpaquePromise &&old) noexcept {
MG_ASSERT(ptr_ == nullptr);
MG_ASSERT(old.ptr_ != nullptr);
MG_ASSERT(this != &old);
ptr_ = old.ptr_;
trait_ = std::move(old.trait_);
old.ptr_ = nullptr;
return *this;
}
OpaquePromise(const OpaquePromise &) = delete;
OpaquePromise &operator=(const OpaquePromise &) = delete;
template <typename T>
std::unique_ptr<ResponsePromise<T>> Take() && {
MG_ASSERT(typeid(T) == *trait_->TypeInfo());
MG_ASSERT(ptr_ != nullptr);
auto ptr = static_cast<ResponsePromise<T> *>(ptr_);
ptr_ = nullptr;
return std::unique_ptr<T>(ptr);
}
template <typename T>
explicit OpaquePromise(std::unique_ptr<ResponsePromise<T>> promise)
: ptr_(static_cast<void *>(promise.release())), trait_(std::make_unique<OpaquePromiseTrait<T>>()) {}
bool IsAwaited() {
MG_ASSERT(ptr_ != nullptr);
return trait_->IsAwaited(ptr_);
}
void TimeOut() {
MG_ASSERT(ptr_ != nullptr);
trait_->TimeOut(ptr_);
ptr_ = nullptr;
}
void Fill(OpaqueMessage &&opaque_message) {
MG_ASSERT(ptr_ != nullptr);
trait_->Fill(ptr_, std::move(opaque_message));
ptr_ = nullptr;
}
~OpaquePromise() {
MG_ASSERT(ptr_ == nullptr, "OpaquePromise destroyed without being explicitly timed out or filled");
}
};
struct DeadlineAndOpaquePromise {
Time deadline;
OpaquePromise promise;
};
} // namespace memgraph::io

915
src/io/rsm/raft.hpp Normal file
View File

@@ -0,0 +1,915 @@
// Copyright 2022 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.
// TODO(tyler) buffer out-of-order Append buffers on the Followers to reassemble more quickly
// TODO(tyler) handle granular batch sizes based on simple flow control
#pragma once
#include <deque>
#include <iostream>
#include <map>
#include <set>
#include <thread>
#include <unordered_map>
#include <vector>
#include "io/simulator/simulator.hpp"
#include "io/transport.hpp"
#include "utils/concepts.hpp"
namespace memgraph::io::rsm {
/// Timeout and replication tunables
using namespace std::chrono_literals;
static constexpr auto kMinimumElectionTimeout = 100ms;
static constexpr auto kMaximumElectionTimeout = 200ms;
static constexpr auto kMinimumBroadcastTimeout = 40ms;
static constexpr auto kMaximumBroadcastTimeout = 60ms;
static constexpr auto kMinimumCronInterval = 1ms;
static constexpr auto kMaximumCronInterval = 2ms;
static constexpr auto kMinimumReceiveTimeout = 40ms;
static constexpr auto kMaximumReceiveTimeout = 60ms;
static_assert(kMinimumElectionTimeout > kMaximumBroadcastTimeout,
"The broadcast timeout has to be smaller than the election timeout!");
static_assert(kMinimumElectionTimeout < kMaximumElectionTimeout,
"The minimum election timeout has to be smaller than the maximum election timeout!");
static_assert(kMinimumBroadcastTimeout < kMaximumBroadcastTimeout,
"The minimum broadcast timeout has to be smaller than the maximum broadcast timeout!");
static_assert(kMinimumCronInterval < kMaximumCronInterval,
"The minimum cron interval has to be smaller than the maximum cron interval!");
static_assert(kMinimumReceiveTimeout < kMaximumReceiveTimeout,
"The minimum receive timeout has to be smaller than the maximum receive timeout!");
static constexpr size_t kMaximumAppendBatchSize = 1024;
using Term = uint64_t;
using LogIndex = uint64_t;
using LogSize = uint64_t;
using RequestId = uint64_t;
template <typename WriteOperation>
struct WriteRequest {
WriteOperation operation;
};
/// WriteResponse is returned to a client after
/// their WriteRequest was entered in to the raft
/// log and it reached consensus.
///
/// WriteReturn is the result of applying the WriteRequest to
/// ReplicatedState, and if the ReplicatedState::write
/// method is deterministic, all replicas will
/// have the same ReplicatedState after applying
/// the submitted WriteRequest.
template <typename WriteReturn>
struct WriteResponse {
bool success;
WriteReturn write_return;
std::optional<Address> retry_leader;
LogIndex raft_index;
};
template <typename ReadOperation>
struct ReadRequest {
ReadOperation operation;
};
template <typename ReadReturn>
struct ReadResponse {
bool success;
ReadReturn read_return;
std::optional<Address> retry_leader;
};
/// AppendRequest is a raft-level message that the Leader
/// periodically broadcasts to all Follower peers. This
/// serves three main roles:
/// 1. acts as a heartbeat from the Leader to the Follower
/// 2. replicates new data that the Leader has received to the Follower
/// 3. informs Follower peers when the commit index has increased,
/// signalling that it is now safe to apply log items to the
/// replicated state machine
template <typename WriteRequest>
struct AppendRequest {
Term term = 0;
LogIndex batch_start_log_index;
Term last_log_term;
std::vector<std::pair<Term, WriteRequest>> entries;
LogSize leader_commit;
};
struct AppendResponse {
bool success;
Term term;
Term last_log_term;
// a small optimization over the raft paper, tells
// the leader the offset that we are interested in
// to send log offsets from for us. This will only
// be useful at the beginning of a leader's term.
LogSize log_size;
};
struct VoteRequest {
Term term = 0;
LogSize log_size;
Term last_log_term;
};
struct VoteResponse {
Term term = 0;
LogSize committed_log_size;
bool vote_granted = false;
};
template <typename WriteRequest>
struct CommonState {
Term term = 0;
std::vector<std::pair<Term, WriteRequest>> log;
LogSize committed_log_size = 0;
LogSize applied_size = 0;
};
struct FollowerTracker {
LogIndex next_index = 0;
LogSize confirmed_log_size = 0;
};
struct PendingClientRequest {
RequestId request_id;
Address address;
Time received_at;
};
struct Leader {
std::map<Address, FollowerTracker> followers;
std::unordered_map<LogIndex, PendingClientRequest> pending_client_requests;
Time last_broadcast = Time::min();
std::string static ToString() { return "\tLeader \t"; }
};
struct Candidate {
std::map<Address, LogSize> successful_votes;
Time election_began = Time::min();
std::set<Address> outstanding_votes;
std::string static ToString() { return "\tCandidate\t"; }
};
struct Follower {
Time last_received_append_entries_timestamp;
Address leader_address;
std::string static ToString() { return "\tFollower \t"; }
};
using Role = std::variant<Candidate, Leader, Follower>;
template <typename Role>
concept AllRoles = memgraph::utils::SameAsAnyOf<Role, Leader, Follower, Candidate>;
template <typename Role>
concept LeaderOrFollower = memgraph::utils::SameAsAnyOf<Role, Leader, Follower>;
template <typename Role>
concept FollowerOrCandidate = memgraph::utils::SameAsAnyOf<Role, Follower, Candidate>;
/*
all ReplicatedState classes should have an Apply method
that returns our WriteResponseValue after consensus, and
a Read method that returns our ReadResponseValue without
requiring consensus.
ReadResponse Read(ReadOperation);
WriteResponseValue ReplicatedState::Apply(WriteRequest);
For example:
If the state is uint64_t, and WriteRequest is `struct PlusOne {};`,
and WriteResponseValue is also uint64_t (the new value), then
each call to state.Apply(PlusOne{}) will return the new value
after incrementing it. 0, 1, 2, 3... and this will be sent back
to the client that requested the mutation.
In practice, these mutations will usually be predicated on some
previous value, so that they are idempotent, functioning similarly
to a CAS operation.
*/
template <typename WriteOperation, typename ReadOperation, typename ReplicatedState, typename WriteResponseValue,
typename ReadResponseValue>
concept Rsm = requires(ReplicatedState state, WriteOperation w, ReadOperation r) {
{ state.Read(r) } -> std::same_as<ReadResponseValue>;
{ state.Apply(w) } -> std::same_as<WriteResponseValue>;
};
/// Parameter Purpose
/// --------------------------
/// IoImpl the concrete Io provider - SimulatorTransport, ThriftTransport, etc...
/// ReplicatedState the high-level data structure that is managed by the raft-backed replicated state machine
/// WriteOperation the individual operation type that is applied to the ReplicatedState in identical order
/// across each replica
/// WriteResponseValue the return value of calling ReplicatedState::Apply(WriteOperation), which is executed in
/// identical order across all replicas after an WriteOperation reaches consensus.
/// ReadOperation the type of operations that do not require consensus before executing directly
/// on a const ReplicatedState &
/// ReadResponseValue the return value of calling ReplicatedState::Read(ReadOperation), which is executed directly
/// without going through consensus first
template <typename IoImpl, typename ReplicatedState, typename WriteOperation, typename WriteResponseValue,
typename ReadOperation, typename ReadResponseValue>
requires Rsm<WriteOperation, ReadOperation, ReplicatedState, WriteResponseValue, ReadResponseValue>
class Raft {
CommonState<WriteOperation> state_;
Role role_ = Candidate{};
Io<IoImpl> io_;
std::vector<Address> peers_;
ReplicatedState replicated_state_;
public:
Raft(Io<IoImpl> &&io, std::vector<Address> peers, ReplicatedState &&replicated_state)
: io_(std::forward<Io<IoImpl>>(io)),
peers_(peers),
replicated_state_(std::forward<ReplicatedState>(replicated_state)) {}
void Run() {
Time last_cron = io_.Now();
while (!io_.ShouldShutDown()) {
const auto now = io_.Now();
const Duration random_cron_interval = RandomTimeout(kMinimumCronInterval, kMaximumCronInterval);
if (now - last_cron > random_cron_interval) {
Cron();
last_cron = now;
}
const Duration receive_timeout = RandomTimeout(kMinimumReceiveTimeout, kMaximumReceiveTimeout);
auto request_result =
io_.template ReceiveWithTimeout<ReadRequest<ReadOperation>, AppendRequest<WriteOperation>, AppendResponse,
WriteRequest<WriteOperation>, VoteRequest, VoteResponse>(receive_timeout);
if (request_result.HasError()) {
continue;
}
auto request = std::move(request_result.GetValue());
Handle(std::move(request.message), request.request_id, request.from_address);
}
}
private:
// Raft paper - 5.3
// When the entry has been safely replicated, the leader applies the
// entry to its state machine and returns the result of that
// execution to the client.
//
// "Safely replicated" is defined as being known to be present
// on at least a majority of all peers (inclusive of the Leader).
void BumpCommitIndexAndReplyToClients(Leader &leader) {
auto confirmed_log_sizes = std::vector<LogSize>{};
// We include our own log size in the calculation of the log
// confirmed log size that is present on at least a majority of all peers.
confirmed_log_sizes.push_back(state_.log.size());
for (const auto &[addr, f] : leader.followers) {
confirmed_log_sizes.push_back(f.confirmed_log_size);
Log("Follower at port ", addr.last_known_port, " has confirmed log size of: ", f.confirmed_log_size);
}
// reverse sort from highest to lowest (using std::ranges::greater)
std::ranges::sort(confirmed_log_sizes, std::ranges::greater());
// This is a particularly correctness-critical calculation because it
// determines the committed log size that will be broadcast in
// the next AppendRequest.
//
// If the following sizes are recorded for clusters of different numbers of peers,
// these are the expected sizes that are considered to have reached consensus:
//
// state | expected value | (confirmed_log_sizes.size() / 2)
// [1] 1 (1 / 2) => 0
// [2, 1] 1 (2 / 2) => 1
// [3, 2, 1] 2 (3 / 2) => 1
// [4, 3, 2, 1] 2 (4 / 2) => 2
// [5, 4, 3, 2, 1] 3 (5 / 2) => 2
const size_t majority_index = confirmed_log_sizes.size() / 2;
const LogSize new_committed_log_size = confirmed_log_sizes[majority_index];
// We never go backwards in history.
MG_ASSERT(state_.committed_log_size <= new_committed_log_size,
"as a Leader, we have previously set our committed_log_size to {}, but our Followers have a majority "
"committed_log_size of {}",
state_.committed_log_size, new_committed_log_size);
state_.committed_log_size = new_committed_log_size;
// For each size between the old size and the new one (inclusive),
// Apply that log's WriteOperation to our replicated_state_,
// and use the specific return value of the ReplicatedState::Apply
// method (WriteResponseValue) to respond to the requester.
for (; state_.applied_size < state_.committed_log_size; state_.applied_size++) {
const LogIndex apply_index = state_.applied_size;
const auto &write_request = state_.log[apply_index].second;
const WriteResponseValue write_return = replicated_state_.Apply(write_request);
if (leader.pending_client_requests.contains(apply_index)) {
const PendingClientRequest client_request = std::move(leader.pending_client_requests.at(apply_index));
leader.pending_client_requests.erase(apply_index);
const WriteResponse<WriteResponseValue> resp{
.success = true,
.write_return = std::move(write_return),
.raft_index = apply_index,
};
io_.Send(client_request.address, client_request.request_id, std::move(resp));
}
}
Log("committed_log_size is now ", state_.committed_log_size);
}
// Raft paper - 5.1
// AppendEntries RPCs are initiated by leaders to replicate log entries and to provide a form of heartbeat
void BroadcastAppendEntries(std::map<Address, FollowerTracker> &followers) {
for (auto &[address, follower] : followers) {
const LogIndex next_index = follower.next_index;
const auto missing = state_.log.size() - next_index;
const auto batch_size = std::min(missing, kMaximumAppendBatchSize);
const auto start_index = next_index;
const auto end_index = start_index + batch_size;
// advance follower's next index
follower.next_index += batch_size;
std::vector<std::pair<Term, WriteOperation>> entries;
entries.insert(entries.begin(), state_.log.begin() + start_index, state_.log.begin() + end_index);
const Term previous_term_from_index = PreviousTermFromIndex(start_index);
Log("sending ", entries.size(), " entries to Follower ", address.last_known_port,
" which are above its next_index of ", next_index);
AppendRequest<WriteOperation> ar{
.term = state_.term,
.batch_start_log_index = start_index,
.last_log_term = previous_term_from_index,
.entries = std::move(entries),
.leader_commit = state_.committed_log_size,
};
// request_id not necessary to set because it's not a Future-backed Request.
static constexpr RequestId request_id = 0;
io_.Send(address, request_id, std::move(ar));
}
}
// Raft paper - 5.2
// Raft uses randomized election timeouts to ensure that split votes are rare and that they are resolved quickly
Duration RandomTimeout(Duration min, Duration max) {
std::uniform_int_distribution time_distrib(min.count(), max.count());
const auto rand_micros = io_.Rand(time_distrib);
return Duration{rand_micros};
}
Duration RandomTimeout(int min_micros, int max_micros) {
std::uniform_int_distribution time_distrib(min_micros, max_micros);
const int rand_micros = io_.Rand(time_distrib);
return std::chrono::microseconds{rand_micros};
}
Term PreviousTermFromIndex(LogIndex index) const {
if (index == 0 || state_.log.size() + 1 <= index) {
return 0;
}
const auto &[term, data] = state_.log.at(index - 1);
return term;
}
Term CommittedLogTerm() {
MG_ASSERT(state_.log.size() >= state_.committed_log_size);
if (state_.log.empty() || state_.committed_log_size == 0) {
return 0;
}
const auto &[term, data] = state_.log.at(state_.committed_log_size - 1);
return term;
}
Term LastLogTerm() const {
if (state_.log.empty()) {
return 0;
}
const auto &[term, data] = state_.log.back();
return term;
}
template <typename... Ts>
void Log(Ts &&...args) {
const Time now = io_.Now();
const auto micros = std::chrono::duration_cast<std::chrono::milliseconds>(now.time_since_epoch()).count();
const Term term = state_.term;
const std::string role_string = std::visit([&](const auto &role) { return role.ToString(); }, role_);
std::ostringstream out;
out << '\t' << static_cast<int>(micros) << "\t" << term << "\t" << io_.GetAddress().last_known_port;
out << role_string;
(out << ... << args);
spdlog::info(out.str());
}
/////////////////////////////////////////////////////////////
/// Raft-related Cron methods
///
/// Cron + std::visit is how events are dispatched
/// to certain code based on Raft role.
///
/// Cron(role) takes as the first argument a reference to its
/// role, and as the second argument, the message that has
/// been received.
/////////////////////////////////////////////////////////////
/// Periodic protocol maintenance.
void Cron() {
// dispatch periodic logic based on our role to a specific Cron method.
std::optional<Role> new_role = std::visit([&](auto &role) { return Cron(role); }, role_);
if (new_role) {
role_ = std::move(new_role).value();
}
}
// Raft paper - 5.2
// Candidates keep sending Vote to peers until:
// 1. receiving Append with a higher term (become Follower)
// 2. receiving Vote with a higher term (become a Follower)
// 3. receiving a quorum of responses to our last batch of Vote (become a Leader)
std::optional<Role> Cron(Candidate &candidate) {
const auto now = io_.Now();
const Duration election_timeout = RandomTimeout(kMinimumElectionTimeout, kMaximumElectionTimeout);
const auto election_timeout_us = std::chrono::duration_cast<std::chrono::milliseconds>(election_timeout).count();
if (now - candidate.election_began > election_timeout) {
state_.term++;
Log("becoming Candidate for term ", state_.term, " after leader timeout of ", election_timeout_us,
"ms elapsed since last election attempt");
const VoteRequest request{
.term = state_.term,
.log_size = state_.log.size(),
.last_log_term = LastLogTerm(),
};
auto outstanding_votes = std::set<Address>();
for (const auto &peer : peers_) {
// request_id not necessary to set because it's not a Future-backed Request.
static constexpr auto request_id = 0;
io_.template Send<VoteRequest>(peer, request_id, request);
outstanding_votes.insert(peer);
}
return Candidate{
.successful_votes = std::map<Address, LogIndex>(),
.election_began = now,
.outstanding_votes = outstanding_votes,
};
}
return std::nullopt;
}
// Raft paper - 5.2
// Followers become candidates if we haven't heard from the leader
// after a randomized timeout.
std::optional<Role> Cron(Follower &follower) {
const auto now = io_.Now();
const auto time_since_last_append_entries = now - follower.last_received_append_entries_timestamp;
// randomized follower timeout
const Duration election_timeout = RandomTimeout(kMinimumElectionTimeout, kMaximumElectionTimeout);
if (time_since_last_append_entries > election_timeout) {
// become a Candidate if we haven't heard from the Leader after this timeout
return Candidate{};
}
return std::nullopt;
}
// Leaders (re)send AppendRequest to followers.
std::optional<Role> Cron(Leader &leader) {
const Time now = io_.Now();
const Duration broadcast_timeout = RandomTimeout(kMinimumBroadcastTimeout, kMaximumBroadcastTimeout);
if (now - leader.last_broadcast > broadcast_timeout) {
BroadcastAppendEntries(leader.followers);
leader.last_broadcast = now;
}
return std::nullopt;
}
/////////////////////////////////////////////////////////////
/// Raft-related Handle methods
///
/// Handle + std::visit is how events are dispatched
/// to certain code based on Raft role.
///
/// Handle(role, message, ...)
/// takes as the first argument a reference
/// to its role, and as the second argument, the
/// message that has been received.
/////////////////////////////////////////////////////////////
using ReceiveVariant = std::variant<ReadRequest<ReadOperation>, AppendRequest<WriteOperation>, AppendResponse,
WriteRequest<WriteOperation>, VoteRequest, VoteResponse>;
void Handle(ReceiveVariant &&message_variant, RequestId request_id, Address from_address) {
// dispatch the message to a handler based on our role,
// which can be specified in the Handle first argument,
// or it can be `auto` if it's a handler for several roles
// or messages.
std::optional<Role> new_role = std::visit(
[&](auto &&msg, auto &role) mutable {
return Handle(role, std::forward<decltype(msg)>(msg), request_id, from_address);
},
std::forward<ReceiveVariant>(message_variant), role_);
// TODO(tyler) (M3) maybe replace std::visit with get_if for explicit prioritized matching, [[likely]] etc...
if (new_role) {
role_ = std::move(new_role).value();
}
}
// all roles can receive Vote and possibly become a follower
template <AllRoles ALL>
std::optional<Role> Handle(ALL & /* variable */, VoteRequest &&req, RequestId request_id, Address from_address) {
Log("received VoteRequest from ", from_address.last_known_port, " with term ", req.term);
const bool last_log_term_dominates = req.last_log_term >= LastLogTerm();
const bool term_dominates = req.term > state_.term;
const bool log_size_dominates = req.log_size >= state_.log.size();
const bool new_leader = last_log_term_dominates && term_dominates && log_size_dominates;
if (new_leader) {
MG_ASSERT(req.term > state_.term);
MG_ASSERT(std::max(req.term, state_.term) == req.term);
}
const VoteResponse res{
.term = std::max(req.term, state_.term),
.committed_log_size = state_.committed_log_size,
.vote_granted = new_leader,
};
io_.Send(from_address, request_id, res);
if (new_leader) {
// become a follower
state_.term = req.term;
return Follower{
.last_received_append_entries_timestamp = io_.Now(),
.leader_address = from_address,
};
}
if (term_dominates) {
Log("received a vote from an inferior candidate. Becoming Candidate");
state_.term = std::max(state_.term, req.term) + 1;
return Candidate{};
}
return std::nullopt;
}
std::optional<Role> Handle(Candidate &candidate, VoteResponse &&res, RequestId /* variable */, Address from_address) {
Log("received VoteResponse");
if (!res.vote_granted || res.term != state_.term) {
Log("received unsuccessful VoteResponse from term ", res.term, " when our candidacy term is ", state_.term);
// we received a delayed VoteResponse from the past, which has to do with an election that is
// no longer valid. We can simply drop this.
return std::nullopt;
}
MG_ASSERT(candidate.outstanding_votes.contains(from_address),
"Received unexpected VoteResponse from server not present in Candidate's outstanding_votes!");
candidate.outstanding_votes.erase(from_address);
MG_ASSERT(!candidate.successful_votes.contains(from_address),
"Received unexpected VoteResponse from server already in Candidate's successful_votes!");
candidate.successful_votes.insert({from_address, res.committed_log_size});
if (candidate.successful_votes.size() >= candidate.outstanding_votes.size()) {
std::map<Address, FollowerTracker> followers{};
for (const auto &[address, committed_log_size] : candidate.successful_votes) {
FollowerTracker follower{
.next_index = committed_log_size,
.confirmed_log_size = committed_log_size,
};
followers.insert({address, follower});
}
for (const auto &address : candidate.outstanding_votes) {
FollowerTracker follower{
.next_index = state_.log.size(),
.confirmed_log_size = 0,
};
followers.insert({address, follower});
}
Log("becoming Leader at term ", state_.term);
BroadcastAppendEntries(followers);
return Leader{
.followers = std::move(followers),
.pending_client_requests = std::unordered_map<LogIndex, PendingClientRequest>(),
};
}
return std::nullopt;
}
template <LeaderOrFollower LOF>
std::optional<Role> Handle(LOF & /* variable */, VoteResponse && /* variable */, RequestId /* variable */,
Address /* variable */) {
Log("non-Candidate received VoteResponse");
return std::nullopt;
}
template <AllRoles ALL>
std::optional<Role> Handle(ALL &role, AppendRequest<WriteOperation> &&req, RequestId request_id,
Address from_address) {
// log size starts out as state_.committed_log_size and only if everything is successful do we
// switch it to the log length.
AppendResponse res{
.success = false,
.term = state_.term,
.last_log_term = CommittedLogTerm(),
.log_size = state_.log.size(),
};
if constexpr (std::is_same<ALL, Leader>()) {
MG_ASSERT(req.term != state_.term, "Multiple leaders are acting under the term ", req.term);
}
const bool is_candidate = std::is_same<ALL, Candidate>();
const bool is_failed_competitor = is_candidate && req.term == state_.term;
const Time now = io_.Now();
// Raft paper - 5.2
// While waiting for votes, a candidate may receive an
// AppendEntries RPC from another server claiming to be leader. If
// the leaders term (included in its RPC) is at least as large as
// the candidates current term, then the candidate recognizes the
// leader as legitimate and returns to follower state.
if (req.term > state_.term || is_failed_competitor) {
// become follower of this leader, reply with our log status
state_.term = req.term;
io_.Send(from_address, request_id, res);
Log("becoming Follower of Leader ", from_address.last_known_port, " at term ", req.term);
return Follower{
.last_received_append_entries_timestamp = now,
.leader_address = from_address,
};
}
if (req.term < state_.term) {
// nack this request from an old leader
io_.Send(from_address, request_id, res);
return std::nullopt;
}
// at this point, we're dealing with our own leader
if constexpr (std::is_same<ALL, Follower>()) {
// small specialization for when we're already a Follower
MG_ASSERT(role.leader_address == from_address, "Multiple Leaders are acting under the same term number!");
role.last_received_append_entries_timestamp = now;
} else {
Log("Somehow entered Follower-specific logic as a non-Follower");
MG_ASSERT(false, "Somehow entered Follower-specific logic as a non-Follower");
}
// Handle steady-state conditions.
if (req.batch_start_log_index != state_.log.size()) {
Log("req.batch_start_log_index of ", req.batch_start_log_index, " does not match our log size of ",
state_.log.size());
} else if (req.last_log_term != LastLogTerm()) {
Log("req.last_log_term differs from our leader term at that slot, expected: ", LastLogTerm(), " but got ",
req.last_log_term);
} else {
// happy path - Apply log
Log("applying batch of ", req.entries.size(), " entries to our log starting at index ",
req.batch_start_log_index);
const auto resize_length = req.batch_start_log_index;
MG_ASSERT(resize_length >= state_.committed_log_size,
"Applied history from Leader which goes back in time from our commit_index");
// possibly chop-off stuff that was replaced by
// things with different terms (we got data that
// hasn't reached consensus yet, which is normal)
state_.log.resize(resize_length);
if (req.entries.size() > 0) {
auto &[first_term, op] = req.entries.at(0);
MG_ASSERT(LastLogTerm() <= first_term);
}
state_.log.insert(state_.log.end(), std::make_move_iterator(req.entries.begin()),
std::make_move_iterator(req.entries.end()));
MG_ASSERT(req.leader_commit >= state_.committed_log_size);
state_.committed_log_size = std::min(req.leader_commit, state_.log.size());
for (; state_.applied_size < state_.committed_log_size; state_.applied_size++) {
const auto &write_request = state_.log[state_.applied_size].second;
replicated_state_.Apply(write_request);
}
res.success = true;
}
res.last_log_term = LastLogTerm();
res.log_size = state_.log.size();
Log("returning log_size of ", res.log_size);
io_.Send(from_address, request_id, res);
return std::nullopt;
}
std::optional<Role> Handle(Leader &leader, AppendResponse &&res, RequestId /* variable */, Address from_address) {
if (res.term != state_.term) {
Log("received AppendResponse related to a previous term when we (presumably) were the leader");
return std::nullopt;
}
// TODO(tyler) when we have dynamic membership, this assert will become incorrect, but we should
// keep it in-place until then because it has bug finding value.
MG_ASSERT(leader.followers.contains(from_address), "received AppendResponse from unknown Follower");
// at this point, we know the term matches and we know this Follower
FollowerTracker &follower = leader.followers.at(from_address);
if (res.success) {
Log("got successful AppendResponse from ", from_address.last_known_port, " with log_size of ", res.log_size);
follower.next_index = std::max(follower.next_index, res.log_size);
} else {
Log("got unsuccessful AppendResponse from ", from_address.last_known_port, " with log_size of ", res.log_size);
follower.next_index = res.log_size;
}
follower.confirmed_log_size = std::max(follower.confirmed_log_size, res.log_size);
BumpCommitIndexAndReplyToClients(leader);
return std::nullopt;
}
template <FollowerOrCandidate FOC>
std::optional<Role> Handle(FOC & /* variable */, AppendResponse && /* variable */, RequestId /* variable */,
Address /* variable */) {
// we used to be the leader, and are getting old delayed responses
return std::nullopt;
}
/////////////////////////////////////////////////////////////
/// RSM-related handle methods
/////////////////////////////////////////////////////////////
// Leaders are able to immediately respond to the requester (with a ReadResponseValue) applied to the ReplicatedState
std::optional<Role> Handle(Leader & /* variable */, ReadRequest<ReadOperation> &&req, RequestId request_id,
Address from_address) {
Log("handling ReadOperation");
ReadOperation read_operation = req.operation;
ReadResponseValue read_return = replicated_state_.Read(read_operation);
const ReadResponse<ReadResponseValue> resp{
.success = true,
.read_return = std::move(read_return),
.retry_leader = std::nullopt,
};
io_.Send(from_address, request_id, resp);
return std::nullopt;
}
// Candidates should respond with a failure, similar to the Candidate + WriteRequest failure below
std::optional<Role> Handle(Candidate & /* variable */, ReadRequest<ReadOperation> && /* variable */,
RequestId request_id, Address from_address) {
Log("received ReadOperation - not redirecting because no Leader is known");
const ReadResponse<ReadResponseValue> res{
.success = false,
};
io_.Send(from_address, request_id, res);
Cron();
return std::nullopt;
}
// Followers should respond with a redirection, similar to the Follower + WriteRequest response below
std::optional<Role> Handle(Follower &follower, ReadRequest<ReadOperation> && /* variable */, RequestId request_id,
Address from_address) {
Log("redirecting client to known Leader with port ", follower.leader_address.last_known_port);
const ReadResponse<ReadResponseValue> res{
.success = false,
.retry_leader = follower.leader_address,
};
io_.Send(from_address, request_id, res);
return std::nullopt;
}
// Raft paper - 8
// When a client first starts up, it connects to a randomly chosen
// server. If the clients first choice is not the leader, that
// server will reject the clients request and supply information
// about the most recent leader it has heard from.
std::optional<Role> Handle(Follower &follower, WriteRequest<WriteOperation> && /* variable */, RequestId request_id,
Address from_address) {
Log("redirecting client to known Leader with port ", follower.leader_address.last_known_port);
const WriteResponse<WriteResponseValue> res{
.success = false,
.retry_leader = follower.leader_address,
};
io_.Send(from_address, request_id, res);
return std::nullopt;
}
std::optional<Role> Handle(Candidate & /* variable */, WriteRequest<WriteOperation> && /* variable */,
RequestId request_id, Address from_address) {
Log("received WriteRequest - not redirecting because no Leader is known");
const WriteResponse<WriteResponseValue> res{
.success = false,
};
io_.Send(from_address, request_id, res);
Cron();
return std::nullopt;
}
// only leaders actually handle replication requests from clients
std::optional<Role> Handle(Leader &leader, WriteRequest<WriteOperation> &&req, RequestId request_id,
Address from_address) {
Log("handling WriteRequest");
// we are the leader. add item to log and send Append to peers
MG_ASSERT(state_.term >= LastLogTerm());
state_.log.emplace_back(std::pair(state_.term, std::move(req.operation)));
LogIndex log_index = state_.log.size() - 1;
PendingClientRequest pcr{
.request_id = request_id,
.address = from_address,
.received_at = io_.Now(),
};
leader.pending_client_requests.emplace(log_index, pcr);
BroadcastAppendEntries(leader.followers);
return std::nullopt;
}
};
}; // namespace memgraph::io::rsm

136
src/io/rsm/rsm_client.hpp Normal file
View File

@@ -0,0 +1,136 @@
// Copyright 2022 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.
#pragma once
#include <iostream>
#include <optional>
#include <vector>
#include "io/address.hpp"
#include "io/rsm/raft.hpp"
#include "utils/result.hpp"
namespace memgraph::io::rsm {
using memgraph::io::Address;
using memgraph::io::Duration;
using memgraph::io::ResponseEnvelope;
using memgraph::io::ResponseFuture;
using memgraph::io::ResponseResult;
using memgraph::io::Time;
using memgraph::io::TimedOut;
using memgraph::io::rsm::ReadRequest;
using memgraph::io::rsm::ReadResponse;
using memgraph::io::rsm::WriteRequest;
using memgraph::io::rsm::WriteResponse;
using memgraph::utils::BasicResult;
template <typename IoImpl, typename WriteRequestT, typename WriteResponseT, typename ReadRequestT,
typename ReadResponseT>
class RsmClient {
using ServerPool = std::vector<Address>;
Io<IoImpl> io_;
Address leader_;
ServerPool server_addrs_;
template <typename ResponseT>
void PossiblyRedirectLeader(const ResponseT &response) {
if (response.retry_leader) {
MG_ASSERT(!response.success, "retry_leader should never be set for successful responses");
leader_ = response.retry_leader.value();
spdlog::debug("client redirected to leader server {}", leader_.ToString());
} else if (!response.success) {
std::uniform_int_distribution<size_t> addr_distrib(0, (server_addrs_.size() - 1));
size_t addr_index = io_.Rand(addr_distrib);
leader_ = server_addrs_[addr_index];
spdlog::debug(
"client NOT redirected to leader server despite our success failing to be processed (it probably was sent to "
"a RSM Candidate) trying a random one at index {} with address {}",
addr_index, leader_.ToString());
}
}
public:
RsmClient(Io<IoImpl> io, Address leader, ServerPool server_addrs)
: io_{io}, leader_{leader}, server_addrs_{server_addrs} {}
RsmClient() = delete;
BasicResult<TimedOut, WriteResponseT> SendWriteRequest(WriteRequestT req) {
WriteRequest<WriteRequestT> client_req;
client_req.operation = req;
const Duration overall_timeout = io_.GetDefaultTimeout();
const Time before = io_.Now();
do {
spdlog::debug("client sending WriteRequest to Leader {}", leader_.ToString());
ResponseFuture<WriteResponse<WriteResponseT>> response_future =
io_.template Request<WriteRequest<WriteRequestT>, WriteResponse<WriteResponseT>>(leader_, client_req);
ResponseResult<WriteResponse<WriteResponseT>> response_result = std::move(response_future).Wait();
if (response_result.HasError()) {
spdlog::debug("client timed out while trying to communicate with leader server {}", leader_.ToString());
return response_result.GetError();
}
ResponseEnvelope<WriteResponse<WriteResponseT>> &&response_envelope = std::move(response_result.GetValue());
WriteResponse<WriteResponseT> &&write_response = std::move(response_envelope.message);
if (write_response.success) {
return std::move(write_response.write_return);
}
PossiblyRedirectLeader(write_response);
} while (io_.Now() < before + overall_timeout);
return TimedOut{};
}
BasicResult<TimedOut, ReadResponseT> SendReadRequest(ReadRequestT req) {
ReadRequest<ReadRequestT> read_req;
read_req.operation = req;
const Duration overall_timeout = io_.GetDefaultTimeout();
const Time before = io_.Now();
do {
spdlog::debug("client sending ReadRequest to Leader {}", leader_.ToString());
ResponseFuture<ReadResponse<ReadResponseT>> get_response_future =
io_.template Request<ReadRequest<ReadRequestT>, ReadResponse<ReadResponseT>>(leader_, read_req);
// receive response
ResponseResult<ReadResponse<ReadResponseT>> get_response_result = std::move(get_response_future).Wait();
if (get_response_result.HasError()) {
spdlog::debug("client timed out while trying to communicate with leader server {}", leader_.ToString());
return get_response_result.GetError();
}
ResponseEnvelope<ReadResponse<ReadResponseT>> &&get_response_envelope = std::move(get_response_result.GetValue());
ReadResponse<ReadResponseT> &&read_get_response = std::move(get_response_envelope.message);
if (read_get_response.success) {
return std::move(read_get_response.read_return);
}
PossiblyRedirectLeader(read_get_response);
} while (io_.Now() < before + overall_timeout);
return TimedOut{};
}
};
} // namespace memgraph::io::rsm

155
src/io/rsm/shard_rsm.hpp Normal file
View File

@@ -0,0 +1,155 @@
// Copyright 2022 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.
#pragma once
/// The ShardRsm is a simple in-memory raft-backed kv store that can be used for simple testing
/// and implementation of some query engine logic before storage engines are fully implemented.
///
/// To implement multiple read and write commands, change the StorageRead* and StorageWrite* requests
/// and responses to a std::variant of the different options, and route them to specific handlers in
/// the ShardRsm's Read and Apply methods. Remember that Read is called immediately when the Raft
/// leader receives the request, and does not replicate anything over Raft. Apply is called only
/// AFTER the StorageWriteRequest is replicated to a majority of Raft peers, and the result of calling
/// ShardRsm::Apply(StorageWriteRequest) is returned to the client that submitted the request.
#include <deque>
#include <iostream>
#include <map>
#include <optional>
#include <set>
#include <thread>
#include <vector>
#include "coordinator/hybrid_logical_clock.hpp"
#include "io/address.hpp"
#include "io/rsm/raft.hpp"
#include "storage/v3/id_types.hpp"
#include "storage/v3/property_value.hpp"
#include "utils/logging.hpp"
namespace memgraph::io::rsm {
using memgraph::coordinator::Hlc;
using memgraph::storage::v3::LabelId;
using memgraph::storage::v3::PropertyValue;
using ShardRsmKey = std::vector<PropertyValue>;
struct StorageWriteRequest {
LabelId label_id;
Hlc transaction_id;
ShardRsmKey key;
std::optional<int> value;
};
struct StorageWriteResponse {
bool shard_rsm_success;
std::optional<int> last_value;
// Only has a value if the given shard does not contain the requested key
std::optional<Hlc> latest_known_shard_map_version{std::nullopt};
};
struct StorageReadRequest {
LabelId label_id;
Hlc transaction_id;
ShardRsmKey key;
};
struct StorageReadResponse {
bool shard_rsm_success;
std::optional<int> value;
// Only has a value if the given shard does not contain the requested key
std::optional<Hlc> latest_known_shard_map_version{std::nullopt};
};
class ShardRsm {
std::map<ShardRsmKey, int> state_;
ShardRsmKey minimum_key_;
std::optional<ShardRsmKey> maximum_key_{std::nullopt};
Hlc shard_map_version_;
// The key is not located in this shard
bool IsKeyInRange(const ShardRsmKey &key) const {
if (maximum_key_) [[likely]] {
return (key >= minimum_key_ && key <= maximum_key_);
}
return key >= minimum_key_;
}
public:
StorageReadResponse Read(StorageReadRequest request) const {
StorageReadResponse ret;
if (!IsKeyInRange(request.key)) {
ret.latest_known_shard_map_version = shard_map_version_;
ret.shard_rsm_success = false;
} else if (state_.contains(request.key)) {
ret.value = state_.at(request.key);
ret.shard_rsm_success = true;
} else {
ret.shard_rsm_success = false;
ret.value = std::nullopt;
}
return ret;
}
StorageWriteResponse Apply(StorageWriteRequest request) {
StorageWriteResponse ret;
// Key is outside the prohibited range
if (!IsKeyInRange(request.key)) {
ret.latest_known_shard_map_version = shard_map_version_;
ret.shard_rsm_success = false;
}
// Key exist
else if (state_.contains(request.key)) {
auto &val = state_[request.key];
/*
* Delete
*/
if (!request.value) {
ret.shard_rsm_success = true;
ret.last_value = val;
state_.erase(state_.find(request.key));
}
/*
* Update
*/
// Does old_value match?
if (request.value == val) {
ret.last_value = val;
ret.shard_rsm_success = true;
val = request.value.value();
} else {
ret.last_value = val;
ret.shard_rsm_success = false;
}
}
/*
* Create
*/
else {
ret.last_value = std::nullopt;
ret.shard_rsm_success = true;
state_.emplace(request.key, request.value.value());
}
return ret;
}
};
} // namespace memgraph::io::rsm

View File

@@ -0,0 +1,8 @@
set(io_simulator_sources
simulator_handle.cpp)
find_package(fmt REQUIRED)
find_package(Threads REQUIRED)
add_library(mg-io-simulator STATIC ${io_simulator_sources})
target_link_libraries(mg-io-simulator stdc++fs Threads::Threads fmt::fmt mg-utils)

View File

@@ -0,0 +1,51 @@
// Copyright 2022 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.
#pragma once
#include <memory>
#include <random>
#include "io/address.hpp"
#include "io/simulator/simulator_config.hpp"
#include "io/simulator/simulator_handle.hpp"
#include "io/simulator/simulator_transport.hpp"
namespace memgraph::io::simulator {
class Simulator {
std::mt19937 rng_;
std::shared_ptr<SimulatorHandle> simulator_handle_;
uint16_t auto_port_ = 0;
public:
explicit Simulator(SimulatorConfig config)
: rng_(std::mt19937{config.rng_seed}), simulator_handle_{std::make_shared<SimulatorHandle>(config)} {}
void ShutDown() { simulator_handle_->ShutDown(); }
Io<SimulatorTransport> RegisterNew() {
Address address = Address::TestAddress(auto_port_++);
return Register(address);
}
Io<SimulatorTransport> Register(Address address) {
std::uniform_int_distribution<uint64_t> seed_distrib;
uint64_t seed = seed_distrib(rng_);
return Io{SimulatorTransport{simulator_handle_, address, seed}, address};
}
void IncrementServerCountAndWaitForQuiescentState(Address address) {
simulator_handle_->IncrementServerCountAndWaitForQuiescentState(address);
}
SimulatorStats Stats() { return simulator_handle_->Stats(); }
};
}; // namespace memgraph::io::simulator

View File

@@ -0,0 +1,30 @@
// Copyright 2022 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.
#pragma once
#include <chrono>
#include "io/time.hpp"
namespace memgraph::io::simulator {
using memgraph::io::Time;
struct SimulatorConfig {
uint8_t drop_percent = 0;
bool perform_timeouts = false;
bool scramble_messages = true;
uint64_t rng_seed = 0;
Time start_time = Time::min();
Time abort_time = Time::max();
};
}; // namespace memgraph::io::simulator

View File

@@ -0,0 +1,142 @@
// Copyright 2022 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 "io/simulator/simulator_handle.hpp"
#include "io/address.hpp"
#include "io/errors.hpp"
#include "io/simulator/simulator_config.hpp"
#include "io/simulator/simulator_stats.hpp"
#include "io/time.hpp"
#include "io/transport.hpp"
namespace memgraph::io::simulator {
using memgraph::io::Duration;
using memgraph::io::Time;
void SimulatorHandle::ShutDown() {
std::unique_lock<std::mutex> lock(mu_);
should_shut_down_ = true;
cv_.notify_all();
}
bool SimulatorHandle::ShouldShutDown() const {
std::unique_lock<std::mutex> lock(mu_);
return should_shut_down_;
}
void SimulatorHandle::IncrementServerCountAndWaitForQuiescentState(Address address) {
std::unique_lock<std::mutex> lock(mu_);
server_addresses_.insert(address);
while (true) {
const size_t blocked_servers = blocked_on_receive_;
const bool all_servers_blocked = blocked_servers == server_addresses_.size();
if (all_servers_blocked) {
return;
}
cv_.wait(lock);
}
}
bool SimulatorHandle::MaybeTickSimulator() {
std::unique_lock<std::mutex> lock(mu_);
const size_t blocked_servers = blocked_on_receive_;
if (blocked_servers < server_addresses_.size()) {
// we only need to advance the simulator when all
// servers have reached a quiescent state, blocked
// on their own futures or receive methods.
return false;
}
stats_.simulator_ticks++;
cv_.notify_all();
TimeoutPromisesPastDeadline();
if (in_flight_.empty()) {
// return early here because there are no messages to schedule
// We tick the clock forward when all servers are blocked but
// there are no in-flight messages to schedule delivery of.
std::poisson_distribution<> time_distrib(50);
Duration clock_advance = std::chrono::microseconds{time_distrib(rng_)};
cluster_wide_time_microseconds_ += clock_advance;
MG_ASSERT(cluster_wide_time_microseconds_ < config_.abort_time,
"Cluster has executed beyond its configured abort_time, and something may be failing to make progress "
"in an expected amount of time.");
return true;
}
if (config_.scramble_messages) {
// scramble messages
std::uniform_int_distribution<size_t> swap_distrib(0, in_flight_.size() - 1);
const size_t swap_index = swap_distrib(rng_);
std::swap(in_flight_[swap_index], in_flight_.back());
}
auto [to_address, opaque_message] = std::move(in_flight_.back());
in_flight_.pop_back();
std::uniform_int_distribution<int> drop_distrib(0, 99);
const int drop_threshold = drop_distrib(rng_);
const bool should_drop = drop_threshold < config_.drop_percent;
if (should_drop) {
stats_.dropped_messages++;
}
PromiseKey promise_key{.requester_address = to_address,
.request_id = opaque_message.request_id,
.replier_address = opaque_message.from_address};
if (promises_.contains(promise_key)) {
// complete waiting promise if it's there
DeadlineAndOpaquePromise dop = std::move(promises_.at(promise_key));
promises_.erase(promise_key);
const bool normal_timeout = config_.perform_timeouts && (dop.deadline < cluster_wide_time_microseconds_);
if (should_drop || normal_timeout) {
stats_.timed_out_requests++;
dop.promise.TimeOut();
} else {
stats_.total_responses++;
dop.promise.Fill(std::move(opaque_message));
}
} else if (should_drop) {
// don't add it anywhere, let it drop
} else {
// add to can_receive_ if not
const auto &[om_vec, inserted] = can_receive_.try_emplace(to_address, std::vector<OpaqueMessage>());
om_vec->second.emplace_back(std::move(opaque_message));
}
return true;
}
Time SimulatorHandle::Now() const {
std::unique_lock<std::mutex> lock(mu_);
return cluster_wide_time_microseconds_;
}
SimulatorStats SimulatorHandle::Stats() {
std::unique_lock<std::mutex> lock(mu_);
return stats_;
}
} // namespace memgraph::io::simulator

View File

@@ -0,0 +1,177 @@
// Copyright 2022 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.
#pragma once
#include <any>
#include <compare>
#include <iostream>
#include <map>
#include <memory>
#include <optional>
#include <set>
#include <utility>
#include <variant>
#include <vector>
#include "io/address.hpp"
#include "io/errors.hpp"
#include "io/message_conversion.hpp"
#include "io/simulator/simulator_config.hpp"
#include "io/simulator/simulator_stats.hpp"
#include "io/time.hpp"
#include "io/transport.hpp"
namespace memgraph::io::simulator {
class SimulatorHandle {
mutable std::mutex mu_{};
mutable std::condition_variable cv_;
// messages that have not yet been scheduled or dropped
std::vector<std::pair<Address, OpaqueMessage>> in_flight_;
// the responses to requests that are being waited on
std::map<PromiseKey, DeadlineAndOpaquePromise> promises_;
// messages that are sent to servers that may later receive them
std::map<Address, std::vector<OpaqueMessage>> can_receive_;
Time cluster_wide_time_microseconds_;
bool should_shut_down_ = false;
SimulatorStats stats_;
size_t blocked_on_receive_ = 0;
std::set<Address> server_addresses_;
std::mt19937 rng_;
SimulatorConfig config_;
void TimeoutPromisesPastDeadline() {
const Time now = cluster_wide_time_microseconds_;
for (auto it = promises_.begin(); it != promises_.end();) {
auto &[promise_key, dop] = *it;
if (dop.deadline < now) {
spdlog::info("timing out request from requester {} to replier {}.", promise_key.requester_address.ToString(),
promise_key.replier_address.ToString());
std::move(dop).promise.TimeOut();
it = promises_.erase(it);
stats_.timed_out_requests++;
} else {
++it;
}
}
}
public:
explicit SimulatorHandle(SimulatorConfig config)
: cluster_wide_time_microseconds_(config.start_time), rng_(config.rng_seed), config_(config) {}
void IncrementServerCountAndWaitForQuiescentState(Address address);
/// This method causes most of the interesting simulation logic to happen, wrt network behavior.
/// It checks to see if all background "server" threads are blocked on new messages, and if so,
/// it will decide whether to drop, reorder, or deliver in-flight messages based on the SimulatorConfig
/// that was used to create the Simulator.
bool MaybeTickSimulator();
void ShutDown();
bool ShouldShutDown() const;
template <Message Request, Message Response>
void SubmitRequest(Address to_address, Address from_address, RequestId request_id, Request &&request,
Duration timeout, ResponsePromise<Response> &&promise) {
std::unique_lock<std::mutex> lock(mu_);
const Time deadline = cluster_wide_time_microseconds_ + timeout;
std::any message(request);
OpaqueMessage om{.to_address = to_address,
.from_address = from_address,
.request_id = request_id,
.message = std::move(message)};
in_flight_.emplace_back(std::make_pair(to_address, std::move(om)));
PromiseKey promise_key{.requester_address = from_address, .request_id = request_id, .replier_address = to_address};
OpaquePromise opaque_promise(std::move(promise).ToUnique());
DeadlineAndOpaquePromise dop{.deadline = deadline, .promise = std::move(opaque_promise)};
promises_.emplace(std::move(promise_key), std::move(dop));
stats_.total_messages++;
stats_.total_requests++;
cv_.notify_all();
}
template <Message... Ms>
requires(sizeof...(Ms) > 0) RequestResult<Ms...> Receive(const Address &receiver, Duration timeout) {
std::unique_lock<std::mutex> lock(mu_);
blocked_on_receive_ += 1;
const Time deadline = cluster_wide_time_microseconds_ + timeout;
while (!should_shut_down_ && (cluster_wide_time_microseconds_ < deadline)) {
if (can_receive_.contains(receiver)) {
std::vector<OpaqueMessage> &can_rx = can_receive_.at(receiver);
if (!can_rx.empty()) {
OpaqueMessage message = std::move(can_rx.back());
can_rx.pop_back();
// TODO(tyler) search for item in can_receive_ that matches the desired types, rather
// than asserting that the last item in can_rx matches.
auto m_opt = std::move(message).Take<Ms...>();
blocked_on_receive_ -= 1;
return std::move(m_opt).value();
}
}
lock.unlock();
bool made_progress = MaybeTickSimulator();
lock.lock();
if (!should_shut_down_ && !made_progress) {
cv_.wait(lock);
}
}
blocked_on_receive_ -= 1;
return TimedOut{};
}
template <Message M>
void Send(Address to_address, Address from_address, RequestId request_id, M message) {
std::unique_lock<std::mutex> lock(mu_);
std::any message_any(std::move(message));
OpaqueMessage om{.to_address = to_address,
.from_address = from_address,
.request_id = request_id,
.message = std::move(message_any)};
in_flight_.emplace_back(std::make_pair(std::move(to_address), std::move(om)));
stats_.total_messages++;
cv_.notify_all();
}
Time Now() const;
template <class D = std::poisson_distribution<>, class Return = uint64_t>
Return Rand(D distrib) {
std::unique_lock<std::mutex> lock(mu_);
return distrib(rng_);
}
SimulatorStats Stats();
};
}; // namespace memgraph::io::simulator

View File

@@ -0,0 +1,25 @@
// Copyright 2022 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.
#pragma once
#include <cstdint>
namespace memgraph::io::simulator {
struct SimulatorStats {
uint64_t total_messages = 0;
uint64_t dropped_messages = 0;
uint64_t timed_out_requests = 0;
uint64_t total_requests = 0;
uint64_t total_responses = 0;
uint64_t simulator_ticks = 0;
};
}; // namespace memgraph::io::simulator

View File

@@ -0,0 +1,65 @@
// Copyright 2022 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.
#pragma once
#include <memory>
#include <utility>
#include "io/address.hpp"
#include "io/simulator/simulator_handle.hpp"
#include "io/time.hpp"
namespace memgraph::io::simulator {
using memgraph::io::Duration;
using memgraph::io::Time;
class SimulatorTransport {
std::shared_ptr<SimulatorHandle> simulator_handle_;
const Address address_;
std::mt19937 rng_;
public:
SimulatorTransport(std::shared_ptr<SimulatorHandle> simulator_handle, Address address, uint64_t seed)
: simulator_handle_(simulator_handle), address_(address), rng_(std::mt19937{seed}) {}
template <Message RequestT, Message ResponseT>
ResponseFuture<ResponseT> Request(Address address, uint64_t request_id, RequestT request, Duration timeout) {
std::function<bool()> maybe_tick_simulator = [this] { return simulator_handle_->MaybeTickSimulator(); };
auto [future, promise] =
memgraph::io::FuturePromisePairWithNotifier<ResponseResult<ResponseT>>(maybe_tick_simulator);
simulator_handle_->SubmitRequest(address, address_, request_id, std::move(request), timeout, std::move(promise));
return std::move(future);
}
template <Message... Ms>
requires(sizeof...(Ms) > 0) RequestResult<Ms...> Receive(Duration timeout) {
return simulator_handle_->template Receive<Ms...>(address_, timeout);
}
template <Message M>
void Send(Address to_address, Address from_address, uint64_t request_id, M message) {
return simulator_handle_->template Send<M>(to_address, from_address, request_id, message);
}
Time Now() const { return simulator_handle_->Now(); }
bool ShouldShutDown() const { return simulator_handle_->ShouldShutDown(); }
template <class D = std::poisson_distribution<>, class Return = uint64_t>
Return Rand(D distrib) {
return distrib(rng_);
}
};
}; // namespace memgraph::io::simulator

21
src/io/time.hpp Normal file
View File

@@ -0,0 +1,21 @@
// Copyright 2022 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.
#pragma once
#include <chrono>
namespace memgraph::io {
using Duration = std::chrono::microseconds;
using Time = std::chrono::time_point<std::chrono::system_clock, Duration>;
} // namespace memgraph::io

138
src/io/transport.hpp Normal file
View File

@@ -0,0 +1,138 @@
// Copyright 2022 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.
#pragma once
#include <chrono>
#include <concepts>
#include <random>
#include <variant>
#include "io/address.hpp"
#include "io/errors.hpp"
#include "io/future.hpp"
#include "io/time.hpp"
#include "utils/result.hpp"
namespace memgraph::io {
using memgraph::utils::BasicResult;
// TODO(tyler) ensure that Message continues to represent
// reasonable constraints around message types over time,
// as we adapt things to use Thrift-generated message types.
template <typename T>
concept Message = std::same_as<T, std::decay_t<T>>;
using RequestId = uint64_t;
template <Message M>
struct ResponseEnvelope {
M message;
RequestId request_id;
Address to_address;
Address from_address;
};
template <Message M>
using ResponseResult = BasicResult<TimedOut, ResponseEnvelope<M>>;
template <Message M>
using ResponseFuture = memgraph::io::Future<ResponseResult<M>>;
template <Message M>
using ResponsePromise = memgraph::io::Promise<ResponseResult<M>>;
template <Message... Ms>
struct RequestEnvelope {
std::variant<Ms...> message;
RequestId request_id;
Address to_address;
Address from_address;
};
template <Message... Ms>
using RequestResult = BasicResult<TimedOut, RequestEnvelope<Ms...>>;
template <typename I>
class Io {
I implementation_;
Address address_;
RequestId request_id_counter_ = 0;
Duration default_timeout_ = std::chrono::microseconds{100000};
public:
Io(I io, Address address) : implementation_(io), address_(address) {}
/// Set the default timeout for all requests that are issued
/// without an explicit timeout set.
void SetDefaultTimeout(Duration timeout) { default_timeout_ = timeout; }
/// Returns the current default timeout for this Io instance.
Duration GetDefaultTimeout() { return default_timeout_; }
/// Issue a request with an explicit timeout in microseconds provided. This tends to be used by clients.
template <Message RequestT, Message ResponseT>
ResponseFuture<ResponseT> RequestWithTimeout(Address address, RequestT request, Duration timeout) {
const RequestId request_id = ++request_id_counter_;
return implementation_.template Request<RequestT, ResponseT>(address, request_id, request, timeout);
}
/// Issue a request that times out after the default timeout. This tends
/// to be used by clients.
template <Message RequestT, Message ResponseT>
ResponseFuture<ResponseT> Request(Address address, RequestT request) {
const RequestId request_id = ++request_id_counter_;
const Duration timeout = default_timeout_;
return implementation_.template Request<RequestT, ResponseT>(address, request_id, std::move(request), timeout);
}
/// Wait for an explicit number of microseconds for a request of one of the
/// provided types to arrive. This tends to be used by servers.
template <Message... Ms>
RequestResult<Ms...> ReceiveWithTimeout(Duration timeout) {
return implementation_.template Receive<Ms...>(timeout);
}
/// Wait the default number of microseconds for a request of one of the
/// provided types to arrive. This tends to be used by servers.
template <Message... Ms>
requires(sizeof...(Ms) > 0) RequestResult<Ms...> Receive() {
const Duration timeout = default_timeout_;
return implementation_.template Receive<Ms...>(timeout);
}
/// Send a message in a best-effort fashion. This is used for messaging where
/// responses are not necessarily expected, and for servers to respond to requests.
/// If you need reliable delivery, this must be built on-top. TCP is not enough for most use cases.
template <Message M>
void Send(Address to_address, RequestId request_id, M message) {
Address from_address = address_;
return implementation_.template Send<M>(to_address, from_address, request_id, std::move(message));
}
/// The current system time. This time source should be preferred over any other,
/// because it lets us deterministically control clocks from tests for making
/// things like timeouts deterministic.
Time Now() const { return implementation_.Now(); }
/// Returns true if the system should shut-down.
bool ShouldShutDown() const { return implementation_.ShouldShutDown(); }
/// Returns a random number within the specified distribution.
template <class D = std::poisson_distribution<>, class Return = uint64_t>
Return Rand(D distrib) {
return implementation_.template Rand<D, Return>(distrib);
}
Address GetAddress() { return address_; }
};
}; // namespace memgraph::io

41
src/parser/CMakeLists.txt Normal file
View File

@@ -0,0 +1,41 @@
## Generate Antlr openCypher parser
set(opencypher_frontend ${CMAKE_CURRENT_SOURCE_DIR}/opencypher)
set(opencypher_generated ${opencypher_frontend}/generated)
set(opencypher_lexer_grammar ${opencypher_frontend}/grammar/MemgraphCypherLexer.g4)
set(opencypher_parser_grammar ${opencypher_frontend}/grammar/MemgraphCypher.g4)
set(antlr_opencypher_generated_src
${opencypher_generated}/MemgraphCypherLexer.cpp
${opencypher_generated}/MemgraphCypher.cpp
${opencypher_generated}/MemgraphCypherBaseVisitor.cpp
${opencypher_generated}/MemgraphCypherVisitor.cpp
)
set(antlr_opencypher_generated_include
${opencypher_generated}/MemgraphCypherLexer.h
${opencypher_generated}/MemgraphCypher.h
${opencypher_generated}/MemgraphCypherBaseVisitor.h
${opencypher_generated}/MemgraphCypherVisitor.h
)
add_custom_command(
OUTPUT ${antlr_opencypher_generated_src} ${antlr_opencypher_generated_include}
COMMAND ${CMAKE_COMMAND} -E make_directory ${opencypher_generated}
COMMAND
java -jar ${CMAKE_SOURCE_DIR}/libs/antlr-4.10.1-complete.jar
-Dlanguage=Cpp -visitor -package antlropencypher
-o ${opencypher_generated}
${opencypher_lexer_grammar} ${opencypher_parser_grammar}
WORKING_DIRECTORY "${CMAKE_BINARY_DIR}"
DEPENDS
${opencypher_lexer_grammar} ${opencypher_parser_grammar}
${opencypher_frontend}/grammar/CypherLexer.g4
${opencypher_frontend}/grammar/Cypher.g4)
add_custom_target(generated_opencypher_parser
DEPENDS ${antlr_opencypher_generated_src} ${antlr_opencypher_generated_include})
add_library(mg-parser STATIC ${antlr_opencypher_generated_src})
add_dependencies(mg-parser generated_opencypher_parser)
target_include_directories(mg-parser PUBLIC ${CMAKE_CURRENT_SOURCE_DIR})
target_link_libraries(mg-parser antlr4)

View File

@@ -0,0 +1,391 @@
/*
* Copyright (c) 2015-2016 "Neo Technology,"
* Network Engine for Objects in Lund AB [http://neotechnology.com]
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
parser grammar Cypher;
options { tokenVocab=CypherLexer; }
cypher : statement ';'? EOF ;
statement : query ;
query : cypherQuery
| indexQuery
| explainQuery
| profileQuery
| infoQuery
| constraintQuery
;
constraintQuery : ( CREATE | DROP ) CONSTRAINT ON constraint ;
constraint : '(' nodeName=variable ':' labelName ')' ASSERT EXISTS '(' constraintPropertyList ')'
| '(' nodeName=variable ':' labelName ')' ASSERT constraintPropertyList IS UNIQUE
| '(' nodeName=variable ':' labelName ')' ASSERT '(' constraintPropertyList ')' IS NODE KEY
;
constraintPropertyList : variable propertyLookup ( ',' variable propertyLookup )* ;
storageInfo : STORAGE INFO ;
indexInfo : INDEX INFO ;
constraintInfo : CONSTRAINT INFO ;
infoQuery : SHOW ( storageInfo | indexInfo | constraintInfo ) ;
explainQuery : EXPLAIN cypherQuery ;
profileQuery : PROFILE cypherQuery ;
cypherQuery : singleQuery ( cypherUnion )* ( queryMemoryLimit )? ;
indexQuery : createIndex | dropIndex;
singleQuery : clause ( clause )* ;
cypherUnion : ( UNION ALL singleQuery )
| ( UNION singleQuery )
;
clause : cypherMatch
| unwind
| merge
| create
| set
| cypherDelete
| remove
| with
| cypherReturn
| callProcedure
;
cypherMatch : OPTIONAL? MATCH pattern where? ;
unwind : UNWIND expression AS variable ;
merge : MERGE patternPart ( mergeAction )* ;
mergeAction : ( ON MATCH set )
| ( ON CREATE set )
;
create : CREATE pattern ;
set : SET setItem ( ',' setItem )* ;
setItem : ( propertyExpression '=' expression )
| ( variable '=' expression )
| ( variable '+=' expression )
| ( variable nodeLabels )
;
cypherDelete : DETACH? DELETE expression ( ',' expression )* ;
remove : REMOVE removeItem ( ',' removeItem )* ;
removeItem : ( variable nodeLabels )
| propertyExpression
;
with : WITH ( DISTINCT )? returnBody ( where )? ;
cypherReturn : RETURN ( DISTINCT )? returnBody ;
callProcedure : CALL procedureName '(' ( expression ( ',' expression )* )? ')' ( procedureMemoryLimit )? ( yieldProcedureResults )? ;
procedureName : symbolicName ( '.' symbolicName )* ;
yieldProcedureResults : YIELD ( '*' | ( procedureResult ( ',' procedureResult )* ) ) ;
memoryLimit : MEMORY ( UNLIMITED | LIMIT literal ( MB | KB ) ) ;
queryMemoryLimit : QUERY memoryLimit ;
procedureMemoryLimit : PROCEDURE memoryLimit ;
procedureResult : ( variable AS variable ) | variable ;
returnBody : returnItems ( order )? ( skip )? ( limit )? ;
returnItems : ( '*' ( ',' returnItem )* )
| ( returnItem ( ',' returnItem )* )
;
returnItem : ( expression AS variable )
| expression
;
order : ORDER BY sortItem ( ',' sortItem )* ;
skip : L_SKIP expression ;
limit : LIMIT expression ;
sortItem : expression ( ASCENDING | ASC | DESCENDING | DESC )? ;
where : WHERE expression ;
pattern : patternPart ( ',' patternPart )* ;
patternPart : ( variable '=' anonymousPatternPart )
| anonymousPatternPart
;
anonymousPatternPart : patternElement ;
patternElement : ( nodePattern ( patternElementChain )* )
| ( '(' patternElement ')' )
;
nodePattern : '(' ( variable )? ( nodeLabels )? ( properties )? ')' ;
patternElementChain : relationshipPattern nodePattern ;
relationshipPattern : ( leftArrowHead dash ( relationshipDetail )? dash rightArrowHead )
| ( leftArrowHead dash ( relationshipDetail )? dash )
| ( dash ( relationshipDetail )? dash rightArrowHead )
| ( dash ( relationshipDetail )? dash )
;
leftArrowHead : '<' | LeftArrowHeadPart ;
rightArrowHead : '>' | RightArrowHeadPart ;
dash : '-' | DashPart ;
relationshipDetail : '[' ( name=variable )? ( relationshipTypes )? ( variableExpansion )? properties ']'
| '[' ( name=variable )? ( relationshipTypes )? ( variableExpansion )? relationshipLambda ( total_weight=variable )? (relationshipLambda )? ']'
| '[' ( name=variable )? ( relationshipTypes )? ( variableExpansion )? (properties )* ( relationshipLambda total_weight=variable )? (relationshipLambda )? ']';
relationshipLambda: '(' traversed_edge=variable ',' traversed_node=variable '|' expression ')';
variableExpansion : '*' (BFS | WSHORTEST)? ( expression )? ( '..' ( expression )? )? ;
properties : mapLiteral
| parameter
;
relationshipTypes : ':' relTypeName ( '|' ':'? relTypeName )* ;
nodeLabels : nodeLabel ( nodeLabel )* ;
nodeLabel : ':' labelName ;
labelName : symbolicName ;
relTypeName : symbolicName ;
expression : expression12 ;
expression12 : expression11 ( OR expression11 )* ;
expression11 : expression10 ( XOR expression10 )* ;
expression10 : expression9 ( AND expression9 )* ;
expression9 : ( NOT )* expression8 ;
expression8 : expression7 ( partialComparisonExpression )* ;
expression7 : expression6 ( ( '+' expression6 ) | ( '-' expression6 ) )* ;
expression6 : expression5 ( ( '*' expression5 ) | ( '/' expression5 ) | ( '%' expression5 ) )* ;
expression5 : expression4 ( '^' expression4 )* ;
expression4 : ( ( '+' | '-' ) )* expression3a ;
expression3a : expression3b ( stringAndNullOperators )* ;
stringAndNullOperators : ( ( ( ( '=~' ) | ( IN ) | ( STARTS WITH ) | ( ENDS WITH ) | ( CONTAINS ) ) expression3b) | ( IS CYPHERNULL ) | ( IS NOT CYPHERNULL ) ) ;
expression3b : expression2a ( listIndexingOrSlicing )* ;
listIndexingOrSlicing : ( '[' expression ']' )
| ( '[' lower_bound=expression? '..' upper_bound=expression? ']' )
;
expression2a : expression2b ( nodeLabels )? ;
expression2b : atom ( propertyLookup )* ;
atom : literal
| parameter
| caseExpression
| ( COUNT '(' '*' ')' )
| listComprehension
| patternComprehension
| ( FILTER '(' filterExpression ')' )
| ( EXTRACT '(' extractExpression ')' )
| ( REDUCE '(' reduceExpression ')' )
| ( COALESCE '(' expression ( ',' expression )* ')' )
| ( ALL '(' filterExpression ')' )
| ( ANY '(' filterExpression ')' )
| ( NONE '(' filterExpression ')' )
| ( SINGLE '(' filterExpression ')' )
| relationshipsPattern
| parenthesizedExpression
| functionInvocation
| variable
;
literal : numberLiteral
| StringLiteral
| booleanLiteral
| CYPHERNULL
| mapLiteral
| listLiteral
;
booleanLiteral : TRUE
| FALSE
;
listLiteral : '[' ( expression ( ',' expression )* )? ']' ;
partialComparisonExpression : ( '=' expression7 )
| ( '<>' expression7 )
| ( '!=' expression7 )
| ( '<' expression7 )
| ( '>' expression7 )
| ( '<=' expression7 )
| ( '>=' expression7 )
;
parenthesizedExpression : '(' expression ')' ;
relationshipsPattern : nodePattern ( patternElementChain )+ ;
filterExpression : idInColl ( where )? ;
reduceExpression : accumulator=variable '=' initial=expression ',' idInColl '|' expression ;
extractExpression : idInColl '|' expression ;
idInColl : variable IN expression ;
functionInvocation : functionName '(' ( DISTINCT )? ( expression ( ',' expression )* )? ')' ;
functionName : symbolicName ( '.' symbolicName )* ;
listComprehension : '[' filterExpression ( '|' expression )? ']' ;
patternComprehension : '[' ( variable '=' )? relationshipsPattern ( WHERE expression )? '|' expression ']' ;
propertyLookup : '.' ( propertyKeyName ) ;
caseExpression : ( ( CASE ( caseAlternatives )+ ) | ( CASE test=expression ( caseAlternatives )+ ) ) ( ELSE else_expression=expression )? END ;
caseAlternatives : WHEN when_expression=expression THEN then_expression=expression ;
variable : symbolicName ;
numberLiteral : doubleLiteral
| integerLiteral
;
mapLiteral : '{' ( propertyKeyName ':' expression ( ',' propertyKeyName ':' expression )* )? '}' ;
parameter : '$' ( symbolicName | DecimalLiteral ) ;
propertyExpression : atom ( propertyLookup )+ ;
propertyKeyName : symbolicName ;
integerLiteral : DecimalLiteral
| OctalLiteral
| HexadecimalLiteral
;
createIndex : CREATE INDEX ON ':' labelName ( '(' propertyKeyName ')' )? ;
dropIndex : DROP INDEX ON ':' labelName ( '(' propertyKeyName ')' )? ;
doubleLiteral : FloatingLiteral ;
cypherKeyword : ALL
| AND
| ANY
| AS
| ASC
| ASCENDING
| ASSERT
| BFS
| BY
| CALL
| CASE
| CONSTRAINT
| CONTAINS
| COUNT
| CREATE
| CYPHERNULL
| DELETE
| DESC
| DESCENDING
| DETACH
| DISTINCT
| ELSE
| END
| ENDS
| EXISTS
| EXPLAIN
| EXTRACT
| FALSE
| FILTER
| IN
| INDEX
| INFO
| IS
| KEY
| LIMIT
| L_SKIP
| MATCH
| MERGE
| NODE
| NONE
| NOT
| ON
| OPTIONAL
| OR
| ORDER
| PROCEDURE
| PROFILE
| QUERY
| REDUCE
| REMOVE
| RETURN
| SET
| SHOW
| SINGLE
| STARTS
| STORAGE
| THEN
| TRUE
| UNION
| UNIQUE
| UNWIND
| WHEN
| WHERE
| WITH
| WSHORTEST
| XOR
| YIELD
;
symbolicName : UnescapedSymbolicName
| EscapedSymbolicName
| cypherKeyword
;

View File

@@ -0,0 +1,208 @@
/*
* When changing this grammar make sure to update constants in
* src/parser/stripped_lexer_constants.hpp (kKeywords, kSpecialTokens
* and bitsets) if needed.
*/
lexer grammar CypherLexer ;
import UnicodeCategories ;
/* Skip whitespace and comments. */
Skipped : ( Whitespace | Comment ) -> skip ;
fragment Whitespace : '\u0020'
| [\u0009-\u000D]
| [\u001C-\u001F]
| '\u1680' | '\u180E'
| [\u2000-\u200A]
| '\u2028' | '\u2029'
| '\u205F'
| '\u3000'
| '\u00A0'
| '\u202F'
;
fragment Comment : '/*' .*? '*/'
| '//' ~[\r\n]*
;
/* Special symbols. */
LPAREN : '(' ;
RPAREN : ')' ;
LBRACK : '[' ;
RBRACK : ']' ;
LBRACE : '{' ;
RBRACE : '}' ;
COMMA : ',' ;
DOT : '.' ;
DOTS : '..' ;
COLON : ':' ;
SEMICOLON : ';' ;
DOLLAR : '$' ;
PIPE : '|' ;
EQ : '=' ;
LT : '<' ;
GT : '>' ;
LTE : '<=' ;
GTE : '>=' ;
NEQ1 : '<>' ;
NEQ2 : '!=' ;
SIM : '=~' ;
PLUS : '+' ;
MINUS : '-' ;
ASTERISK : '*' ;
SLASH : '/' ;
PERCENT : '%' ;
CARET : '^' ;
PLUS_EQ : '+=' ;
/* Some random unicode characters that can be used to draw arrows. */
LeftArrowHeadPart : '⟨' | '〈' | '﹤' | '' ;
RightArrowHeadPart : '⟩' | '〉' | '﹥' | '' ;
DashPart : '­' | '' | '' | '' | '' | '—' | '―'
| '' | '' | '﹣' | ''
;
/* Cypher reserved words. */
ALL : A L L ;
AND : A N D ;
ANY : A N Y ;
AS : A S ;
ASC : A S C ;
ASCENDING : A S C E N D I N G ;
ASSERT : A S S E R T ;
BFS : B F S ;
BY : B Y ;
CALL : C A L L ;
CASE : C A S E ;
COALESCE : C O A L E S C E ;
CONSTRAINT : C O N S T R A I N T ;
CONTAINS : C O N T A I N S ;
COUNT : C O U N T ;
CREATE : C R E A T E ;
CYPHERNULL : N U L L ;
DELETE : D E L E T E ;
DESC : D E S C ;
DESCENDING : D E S C E N D I N G ;
DETACH : D E T A C H ;
DISTINCT : D I S T I N C T ;
DROP : D R O P ;
ELSE : E L S E ;
END : E N D ;
ENDS : E N D S ;
EXISTS : E X I S T S ;
EXPLAIN : E X P L A I N ;
EXTRACT : E X T R A C T ;
FALSE : F A L S E ;
FILTER : F I L T E R ;
IN : I N ;
INDEX : I N D E X ;
INFO : I N F O ;
IS : I S ;
KB : K B ;
KEY : K E Y ;
LIMIT : L I M I T ;
L_SKIP : S K I P ;
MATCH : M A T C H ;
MB : M B ;
MEMORY : M E M O R Y ;
MERGE : M E R G E ;
NODE : N O D E ;
NONE : N O N E ;
NOT : N O T ;
ON : O N ;
OPTIONAL : O P T I O N A L ;
OR : O R ;
ORDER : O R D E R ;
PROCEDURE : P R O C E D U R E ;
PROFILE : P R O F I L E ;
QUERY : Q U E R Y ;
REDUCE : R E D U C E ;
REMOVE : R E M O V E ;
RETURN : R E T U R N ;
SET : S E T ;
SHOW : S H O W ;
SINGLE : S I N G L E ;
STARTS : S T A R T S ;
STORAGE : S T O R A G E ;
THEN : T H E N ;
TRUE : T R U E ;
UNION : U N I O N ;
UNIQUE : U N I Q U E ;
UNLIMITED : U N L I M I T E D ;
UNWIND : U N W I N D ;
WHEN : W H E N ;
WHERE : W H E R E ;
WITH : W I T H ;
WSHORTEST : W S H O R T E S T ;
XOR : X O R ;
YIELD : Y I E L D ;
/* Double and single quoted string literals. */
StringLiteral : '"' ( ~[\\"] | EscapeSequence )* '"'
| '\'' ( ~[\\'] | EscapeSequence )* '\''
;
fragment EscapeSequence : '\\' ( B | F | N | R | T | '\\' | '\'' | '"' )
| '\\u' HexDigit HexDigit HexDigit HexDigit
| '\\U' HexDigit HexDigit HexDigit HexDigit
HexDigit HexDigit HexDigit HexDigit
;
/* Number literals. */
DecimalLiteral : '0' | NonZeroDigit ( DecDigit )* ;
OctalLiteral : '0' ( OctDigit )+ ;
HexadecimalLiteral : '0x' ( HexDigit )+ ;
FloatingLiteral : DecDigit* '.' DecDigit+ ( E '-'? DecDigit+ )?
| DecDigit+ ( '.' DecDigit* )? ( E '-'? DecDigit+ )
| DecDigit+ ( E '-'? DecDigit+ )
;
fragment NonZeroDigit : [1-9] ;
fragment DecDigit : [0-9] ;
fragment OctDigit : [0-7] ;
fragment HexDigit : [0-9] | [a-f] | [A-F] ;
/* Symbolic names. */
UnescapedSymbolicName : IdentifierStart ( IdentifierPart )* ;
EscapedSymbolicName : ( '`' ~[`]* '`' )+ ;
/**
* Based on the unicode identifier and pattern syntax
* (http://www.unicode.org/reports/tr31/)
* and extended with a few characters.
*/
IdentifierStart : ID_Start | Pc ;
IdentifierPart : ID_Continue | Sc ;
/* Hack for case-insensitive reserved words */
fragment A : 'A' | 'a' ;
fragment B : 'B' | 'b' ;
fragment C : 'C' | 'c' ;
fragment D : 'D' | 'd' ;
fragment E : 'E' | 'e' ;
fragment F : 'F' | 'f' ;
fragment G : 'G' | 'g' ;
fragment H : 'H' | 'h' ;
fragment I : 'I' | 'i' ;
fragment J : 'J' | 'j' ;
fragment K : 'K' | 'k' ;
fragment L : 'L' | 'l' ;
fragment M : 'M' | 'm' ;
fragment N : 'N' | 'n' ;
fragment O : 'O' | 'o' ;
fragment P : 'P' | 'p' ;
fragment Q : 'Q' | 'q' ;
fragment R : 'R' | 'r' ;
fragment S : 'S' | 's' ;
fragment T : 'T' | 't' ;
fragment U : 'U' | 'u' ;
fragment V : 'V' | 'v' ;
fragment W : 'W' | 'w' ;
fragment X : 'X' | 'x' ;
fragment Y : 'Y' | 'y' ;
fragment Z : 'Z' | 'z' ;

View File

@@ -0,0 +1,399 @@
/*
* Copyright 2021 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.
*/
/* Memgraph specific part of Cypher grammar with enterprise features. */
parser grammar MemgraphCypher ;
options { tokenVocab=MemgraphCypherLexer; }
import Cypher ;
memgraphCypherKeyword : cypherKeyword
| AFTER
| ALTER
| ASYNC
| AUTH
| BAD
| BATCH_INTERVAL
| BATCH_LIMIT
| BATCH_SIZE
| BEFORE
| BOOTSTRAP_SERVERS
| CHECK
| CLEAR
| COMMIT
| COMMITTED
| CONFIG
| CONFIGS
| CONSUMER_GROUP
| CREDENTIALS
| CSV
| DATA
| DELIMITER
| DATABASE
| DENY
| DROP
| DUMP
| EXECUTE
| FREE
| FROM
| FOR
| FOREACH
| GLOBAL
| GRANT
| HEADER
| IDENTIFIED
| ISOLATION
| KAFKA
| LEVEL
| LOAD
| LOCK
| MAIN
| MODE
| NEXT
| NO
| PASSWORD
| PULSAR
| PORT
| PRIVILEGES
| READ
| REGISTER
| REPLICA
| REPLICAS
| REPLICATION
| REVOKE
| ROLE
| ROLES
| QUOTE
| SCHEMA
| SCHEMAS
| SESSION
| SETTING
| SETTINGS
| SNAPSHOT
| START
| STATS
| STREAM
| STREAMS
| SYNC
| TIMEOUT
| TO
| TOPICS
| TRANSACTION
| TRANSFORM
| TRIGGER
| TRIGGERS
| UNCOMMITTED
| UNLOCK
| UPDATE
| USER
| USERS
| VERSION
;
symbolicName : UnescapedSymbolicName
| EscapedSymbolicName
| memgraphCypherKeyword
;
query : cypherQuery
| indexQuery
| explainQuery
| profileQuery
| infoQuery
| constraintQuery
| authQuery
| dumpQuery
| replicationQuery
| lockPathQuery
| freeMemoryQuery
| triggerQuery
| isolationLevelQuery
| createSnapshotQuery
| streamQuery
| settingQuery
| versionQuery
| schemaQuery
;
authQuery : createRole
| dropRole
| showRoles
| createUser
| setPassword
| dropUser
| showUsers
| setRole
| clearRole
| grantPrivilege
| denyPrivilege
| revokePrivilege
| showPrivileges
| showRoleForUser
| showUsersForRole
;
replicationQuery : setReplicationRole
| showReplicationRole
| registerReplica
| dropReplica
| showReplicas
;
triggerQuery : createTrigger
| dropTrigger
| showTriggers
;
clause : cypherMatch
| unwind
| merge
| create
| set
| cypherDelete
| remove
| with
| cypherReturn
| callProcedure
| loadCsv
| foreach
;
updateClause : set
| remove
| create
| merge
| cypherDelete
| foreach
;
foreach : FOREACH '(' variable IN expression '|' updateClause+ ')' ;
streamQuery : checkStream
| createStream
| dropStream
| startStream
| startAllStreams
| stopStream
| stopAllStreams
| showStreams
;
settingQuery : setSetting
| showSetting
| showSettings
;
schemaQuery : showSchema
| showSchemas
| createSchema
| dropSchema
;
loadCsv : LOAD CSV FROM csvFile ( WITH | NO ) HEADER
( IGNORE BAD ) ?
( DELIMITER delimiter ) ?
( QUOTE quote ) ?
AS rowVar ;
csvFile : literal ;
delimiter : literal ;
quote : literal ;
rowVar : variable ;
userOrRoleName : symbolicName ;
createRole : CREATE ROLE role=userOrRoleName ;
dropRole : DROP ROLE role=userOrRoleName ;
showRoles : SHOW ROLES ;
createUser : CREATE USER user=userOrRoleName
( IDENTIFIED BY password=literal )? ;
setPassword : SET PASSWORD FOR user=userOrRoleName TO password=literal;
dropUser : DROP USER user=userOrRoleName ;
showUsers : SHOW USERS ;
setRole : SET ROLE FOR user=userOrRoleName TO role=userOrRoleName;
clearRole : CLEAR ROLE FOR user=userOrRoleName ;
grantPrivilege : GRANT ( ALL PRIVILEGES | privileges=privilegeList ) TO userOrRole=userOrRoleName ;
denyPrivilege : DENY ( ALL PRIVILEGES | privileges=privilegeList ) TO userOrRole=userOrRoleName ;
revokePrivilege : REVOKE ( ALL PRIVILEGES | privileges=privilegeList ) FROM userOrRole=userOrRoleName ;
privilege : CREATE
| DELETE
| MATCH
| MERGE
| SET
| REMOVE
| INDEX
| STATS
| AUTH
| CONSTRAINT
| DUMP
| REPLICATION
| READ_FILE
| FREE_MEMORY
| TRIGGER
| CONFIG
| DURABILITY
| STREAM
| MODULE_READ
| MODULE_WRITE
| WEBSOCKET
| SCHEMA
;
privilegeList : privilege ( ',' privilege )* ;
showPrivileges : SHOW PRIVILEGES FOR userOrRole=userOrRoleName ;
showRoleForUser : SHOW ROLE FOR user=userOrRoleName ;
showUsersForRole : SHOW USERS FOR role=userOrRoleName ;
dumpQuery: DUMP DATABASE ;
setReplicationRole : SET REPLICATION ROLE TO ( MAIN | REPLICA )
( WITH PORT port=literal ) ? ;
showReplicationRole : SHOW REPLICATION ROLE ;
replicaName : symbolicName ;
socketAddress : literal ;
registerReplica : REGISTER REPLICA replicaName ( SYNC | ASYNC )
TO socketAddress ;
dropReplica : DROP REPLICA replicaName ;
showReplicas : SHOW REPLICAS ;
lockPathQuery : ( LOCK | UNLOCK ) DATA DIRECTORY ;
freeMemoryQuery : FREE MEMORY ;
triggerName : symbolicName ;
triggerStatement : .*? ;
emptyVertex : '(' ')' ;
emptyEdge : dash dash rightArrowHead ;
createTrigger : CREATE TRIGGER triggerName ( ON ( emptyVertex | emptyEdge ) ? ( CREATE | UPDATE | DELETE ) ) ?
( AFTER | BEFORE ) COMMIT EXECUTE triggerStatement ;
dropTrigger : DROP TRIGGER triggerName ;
showTriggers : SHOW TRIGGERS ;
isolationLevel : SNAPSHOT ISOLATION | READ COMMITTED | READ UNCOMMITTED ;
isolationLevelScope : GLOBAL | SESSION | NEXT ;
isolationLevelQuery : SET isolationLevelScope TRANSACTION ISOLATION LEVEL isolationLevel ;
createSnapshotQuery : CREATE SNAPSHOT ;
streamName : symbolicName ;
symbolicNameWithMinus : symbolicName ( MINUS symbolicName )* ;
symbolicNameWithDotsAndMinus: symbolicNameWithMinus ( DOT symbolicNameWithMinus )* ;
symbolicTopicNames : symbolicNameWithDotsAndMinus ( COMMA symbolicNameWithDotsAndMinus )* ;
topicNames : symbolicTopicNames | literal ;
commonCreateStreamConfig : TRANSFORM transformationName=procedureName
| BATCH_INTERVAL batchInterval=literal
| BATCH_SIZE batchSize=literal
;
createStream : kafkaCreateStream | pulsarCreateStream ;
configKeyValuePair : literal ':' literal ;
configMap : '{' ( configKeyValuePair ( ',' configKeyValuePair )* )? '}' ;
kafkaCreateStreamConfig : TOPICS topicNames
| CONSUMER_GROUP consumerGroup=symbolicNameWithDotsAndMinus
| BOOTSTRAP_SERVERS bootstrapServers=literal
| CONFIGS configsMap=configMap
| CREDENTIALS credentialsMap=configMap
| commonCreateStreamConfig
;
kafkaCreateStream : CREATE KAFKA STREAM streamName ( kafkaCreateStreamConfig ) * ;
pulsarCreateStreamConfig : TOPICS topicNames
| SERVICE_URL serviceUrl=literal
| commonCreateStreamConfig
;
pulsarCreateStream : CREATE PULSAR STREAM streamName ( pulsarCreateStreamConfig ) * ;
dropStream : DROP STREAM streamName ;
startStream : START STREAM streamName ( BATCH_LIMIT batchLimit=literal ) ? ( TIMEOUT timeout=literal ) ? ;
startAllStreams : START ALL STREAMS ;
stopStream : STOP STREAM streamName ;
stopAllStreams : STOP ALL STREAMS ;
showStreams : SHOW STREAMS ;
checkStream : CHECK STREAM streamName ( BATCH_LIMIT batchLimit=literal ) ? ( TIMEOUT timeout=literal ) ? ;
settingName : literal ;
settingValue : literal ;
setSetting : SET DATABASE SETTING settingName TO settingValue ;
showSetting : SHOW DATABASE SETTING settingName ;
showSettings : SHOW DATABASE SETTINGS ;
versionQuery : SHOW VERSION ;
showSchema : SHOW SCHEMA ON ':' labelName ;
showSchemas : SHOW SCHEMAS ;
propertyType : symbolicName ;
propertyKeyTypePair : propertyKeyName propertyType ;
schemaPropertyMap : '(' propertyKeyTypePair ( ',' propertyKeyTypePair )* ')' ;
createSchema : CREATE SCHEMA ON ':' labelName schemaPropertyMap ;
dropSchema : DROP SCHEMA ON ':' labelName ;

View File

@@ -0,0 +1,118 @@
/*
* Copyright 2021 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.
*/
/* Memgraph specific Cypher reserved words used for enterprise features. */
/*
* When changing this grammar make sure to update constants in
* src/parser/stripped_lexer_constants.hpp (kKeywords, kSpecialTokens
* and bitsets) if needed.
*/
lexer grammar MemgraphCypherLexer ;
import CypherLexer ;
UNDERSCORE : '_' ;
AFTER : A F T E R ;
ALTER : A L T E R ;
ASYNC : A S Y N C ;
AUTH : A U T H ;
BAD : B A D ;
BATCH_INTERVAL : B A T C H UNDERSCORE I N T E R V A L ;
BATCH_LIMIT : B A T C H UNDERSCORE L I M I T ;
BATCH_SIZE : B A T C H UNDERSCORE S I Z E ;
BEFORE : B E F O R E ;
BOOTSTRAP_SERVERS : B O O T S T R A P UNDERSCORE S E R V E R S ;
CHECK : C H E C K ;
CLEAR : C L E A R ;
COMMIT : C O M M I T ;
COMMITTED : C O M M I T T E D ;
CONFIG : C O N F I G ;
CONFIGS : C O N F I G S;
CONSUMER_GROUP : C O N S U M E R UNDERSCORE G R O U P ;
CREDENTIALS : C R E D E N T I A L S ;
CSV : C S V ;
DATA : D A T A ;
DELIMITER : D E L I M I T E R ;
DATABASE : D A T A B A S E ;
DENY : D E N Y ;
DIRECTORY : D I R E C T O R Y ;
DROP : D R O P ;
DUMP : D U M P ;
DURABILITY : D U R A B I L I T Y ;
EXECUTE : E X E C U T E ;
FOR : F O R ;
FOREACH : F O R E A C H;
FREE : F R E E ;
FREE_MEMORY : F R E E UNDERSCORE M E M O R Y ;
FROM : F R O M ;
GLOBAL : G L O B A L ;
GRANT : G R A N T ;
GRANTS : G R A N T S ;
HEADER : H E A D E R ;
IDENTIFIED : I D E N T I F I E D ;
IGNORE : I G N O R E ;
ISOLATION : I S O L A T I O N ;
KAFKA : K A F K A ;
LEVEL : L E V E L ;
LOAD : L O A D ;
LOCK : L O C K ;
MAIN : M A I N ;
MODE : M O D E ;
MODULE_READ : M O D U L E UNDERSCORE R E A D ;
MODULE_WRITE : M O D U L E UNDERSCORE W R I T E ;
NEXT : N E X T ;
NO : N O ;
PASSWORD : P A S S W O R D ;
PORT : P O R T ;
PRIVILEGES : P R I V I L E G E S ;
PULSAR : P U L S A R ;
READ : R E A D ;
READ_FILE : R E A D UNDERSCORE F I L E ;
REGISTER : R E G I S T E R ;
REPLICA : R E P L I C A ;
REPLICAS : R E P L I C A S ;
REPLICATION : R E P L I C A T I O N ;
REVOKE : R E V O K E ;
ROLE : R O L E ;
ROLES : R O L E S ;
QUOTE : Q U O T E ;
SCHEMA : S C H E M A ;
SCHEMAS : S C H E M A S ;
SERVICE_URL : S E R V I C E UNDERSCORE U R L ;
SESSION : S E S S I O N ;
SETTING : S E T T I N G ;
SETTINGS : S E T T I N G S ;
SNAPSHOT : S N A P S H O T ;
START : S T A R T ;
STATS : S T A T S ;
STOP : S T O P ;
STREAM : S T R E A M ;
STREAMS : S T R E A M S ;
SYNC : S Y N C ;
TIMEOUT : T I M E O U T ;
TO : T O ;
TOPICS : T O P I C S;
TRANSACTION : T R A N S A C T I O N ;
TRANSFORM : T R A N S F O R M ;
TRIGGER : T R I G G E R ;
TRIGGERS : T R I G G E R S ;
UNCOMMITTED : U N C O M M I T T E D ;
UNLOCK : U N L O C K ;
UPDATE : U P D A T E ;
USER : U S E R ;
USERS : U S E R S ;
VERSION : V E R S I O N ;
WEBSOCKET : W E B S O C K E T ;

File diff suppressed because one or more lines are too long

View File

@@ -0,0 +1,85 @@
// Copyright 2022 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.
#pragma once
#include <string>
#include "antlr4-runtime.h"
#include "utils/exceptions.hpp"
#include "parser/opencypher/generated/MemgraphCypher.h"
#include "parser/opencypher/generated/MemgraphCypherLexer.h"
#include "utils/concepts.hpp"
namespace memgraph::frontend::opencypher {
class SyntaxException : public utils::BasicException {
public:
using utils::BasicException::BasicException;
SyntaxException() : SyntaxException("") {}
};
/**
* Generates openCypher AST
* This thing must me a class since parser.cypher() returns pointer and there is
* no way for us to get ownership over the object.
*/
enum class ParserOpTag : uint8_t {
CYPHER, EXPRESSION
};
template<ParserOpTag Tag = ParserOpTag::CYPHER>
class Parser {
public:
/**
* @param query incoming query that has to be compiled into query plan
* the first step is to generate AST
*/
Parser(const std::string query) : query_(std::move(query)) {
parser_.removeErrorListeners();
parser_.addErrorListener(&error_listener_);
if constexpr(Tag == ParserOpTag::CYPHER) {
tree_ = parser_.cypher();
}
else {
tree_ = parser_.expression();
}
if (parser_.getNumberOfSyntaxErrors()) {
throw SyntaxException(error_listener_.error_);
}
}
auto tree() { return tree_; }
private:
class FirstMessageErrorListener : public antlr4::BaseErrorListener {
void syntaxError(antlr4::Recognizer *, antlr4::Token *, size_t line, size_t position, const std::string &message,
std::exception_ptr) override {
if (error_.empty()) {
error_ = "line " + std::to_string(line) + ":" + std::to_string(position + 1) + " " + message;
}
}
public:
std::string error_;
};
FirstMessageErrorListener error_listener_;
std::string query_;
antlr4::ANTLRInputStream input_{query_};
antlropencypher::MemgraphCypherLexer lexer_{&input_};
antlr4::CommonTokenStream tokens_{&lexer_};
// generate ast
antlropencypher::MemgraphCypher parser_{&tokens_};
antlr4::tree::ParseTree *tree_ = nullptr;
};
} // namespace memgraph::frontend::opencypher

File diff suppressed because it is too large Load Diff

View File

@@ -51,7 +51,6 @@ class EdgeAccessor final {
public:
storage::EdgeAccessor impl_;
public:
explicit EdgeAccessor(storage::EdgeAccessor impl) : impl_(std::move(impl)) {}
bool IsVisible(storage::View view) const { return impl_.IsVisible(view); }
@@ -97,7 +96,6 @@ class VertexAccessor final {
static EdgeAccessor MakeEdgeAccessor(const storage::EdgeAccessor impl) { return EdgeAccessor(impl); }
public:
explicit VertexAccessor(storage::VertexAccessor impl) : impl_(impl) {}
bool IsVisible(storage::View view) const { return impl_.IsVisible(view); }

View File

@@ -204,7 +204,7 @@ const trie::Trie kKeywords = {"union",
"pulsar",
"service_url",
"version",
"websocket"
"websocket",
"foreach"};
// Unicode codepoints that are allowed at the start of the unescaped name.

View File

@@ -114,4 +114,4 @@ std::string ExecutionStatsKeyToString(const ExecutionStats::Key key) {
}
}
} // namespace memgraph::query
} // namespace memgraph::query

View File

@@ -368,13 +368,12 @@ VertexAccessor &CreateExpand::CreateExpandCursor::OtherVertex(Frame &frame, Exec
TypedValue &dest_node_value = frame[self_.node_info_.symbol];
ExpectType(self_.node_info_.symbol, dest_node_value, TypedValue::Type::Vertex);
return dest_node_value.ValueVertex();
} else {
auto &created_vertex = CreateLocalVertex(self_.node_info_, &frame, context);
if (context.trigger_context_collector) {
context.trigger_context_collector->RegisterCreatedObject(created_vertex);
}
return created_vertex;
}
auto &created_vertex = CreateLocalVertex(self_.node_info_, &frame, context);
if (context.trigger_context_collector) {
context.trigger_context_collector->RegisterCreatedObject(created_vertex);
}
return created_vertex;
}
template <class TVerticesFun>

View File

@@ -0,0 +1,58 @@
define_add_lcp(add_lcp_query lcp_query_v2_cpp_files generated_lcp_query_v2_files)
add_lcp_query(frontend/ast/ast.lcp)
add_lcp_query(plan/operator.lcp)
add_custom_target(generate_lcp_query_v2 DEPENDS ${generated_lcp_query_v2_files})
set(mg_query_v2_sources
${lcp_query_v2_cpp_files}
common.cpp
cypher_query_interpreter.cpp
dump.cpp
frontend/semantic/required_privileges.cpp
frontend/stripped.cpp
interpret/awesome_memgraph_functions.cpp
interpreter.cpp
metadata.cpp
plan/operator.cpp
plan/preprocess.cpp
plan/pretty_print.cpp
plan/profile.cpp
plan/read_write_type_checker.cpp
plan/rewrite/index_lookup.cpp
plan/rule_based_planner.cpp
plan/variable_start_planner.cpp
procedure/mg_procedure_impl.cpp
procedure/mg_procedure_helpers.cpp
procedure/module.cpp
procedure/py_module.cpp
serialization/property_value.cpp
stream/streams.cpp
stream/sources.cpp
stream/common.cpp
trigger.cpp
trigger_context.cpp
bindings/typed_value.cpp)
find_package(Boost REQUIRED)
add_library(mg-query-v2 STATIC ${mg_query_v2_sources})
add_dependencies(mg-query-v2 generate_lcp_query_v2)
target_include_directories(mg-query-v2 PUBLIC ${CMAKE_SOURCE_DIR}/include)
target_include_directories(mg-query-v2 PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}/bindings)
target_link_libraries(mg-query-v2 dl cppitertools Boost::headers)
target_link_libraries(mg-query-v2 mg-integrations-pulsar mg-integrations-kafka mg-storage-v3 mg-license mg-utils mg-kvstore mg-memory)
target_link_libraries(mg-query-v2 mg-expr)
if(NOT "${MG_PYTHON_PATH}" STREQUAL "")
set(Python3_ROOT_DIR "${MG_PYTHON_PATH}")
endif()
if("${MG_PYTHON_VERSION}" STREQUAL "")
find_package(Python3 3.5 REQUIRED COMPONENTS Development)
else()
find_package(Python3 "${MG_PYTHON_VERSION}" EXACT REQUIRED COMPONENTS Development)
endif()
target_link_libraries(mg-query-v2 Python3::Python)

View File

@@ -0,0 +1,29 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/frontend/ast/ast.hpp"
namespace memgraph::query::v2 {
class AuthChecker {
public:
virtual bool IsUserAuthorized(const std::optional<std::string> &username,
const std::vector<query::v2::AuthQuery::Privilege> &privileges) const = 0;
};
class AllowEverythingAuthChecker final : public query::v2::AuthChecker {
bool IsUserAuthorized(const std::optional<std::string> & /*username*/,
const std::vector<query::v2::AuthQuery::Privilege> & /*privileges*/) const override {
return true;
}
};
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,16 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/bindings/bindings.hpp"
#include "expr/ast/ast_visitor.hpp"

View File

@@ -0,0 +1,15 @@
// Copyright 2022 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.
#pragma once
#define MG_AST_INCLUDE_PATH "query/v2/frontend/ast/ast.hpp" // NOLINT(cppcoreguidelines-macro-usage)
#define MG_INJECTED_NAMESPACE_NAME memgraph::query::v2 // NOLINT(cppcoreguidelines-macro-usage)

View File

@@ -0,0 +1,20 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/bindings/bindings.hpp"
#include "expr/ast/cypher_main_visitor.hpp"
namespace memgraph::query::v2 {
using CypherMainVisitor = memgraph::expr::CypherMainVisitor;
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,34 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/bindings/bindings.hpp"
#include "expr/interpret/eval.hpp"
#include "query/v2/bindings/typed_value.hpp"
#include "query/v2/context.hpp"
#include "query/v2/db_accessor.hpp"
#include "storage/v3/id_types.hpp"
#include "storage/v3/property_store.hpp"
#include "storage/v3/view.hpp"
#include "storage/v3/conversions.hpp"
namespace memgraph::query::v2 {
inline const auto lam = [](const auto &val) { return memgraph::storage::v3::PropertyToTypedValue<TypedValue>(val); };
using ExpressionEvaluator =
memgraph::expr::ExpressionEvaluator<TypedValue, EvaluationContext, DbAccessor, storage::v3::View,
storage::v3::LabelId, storage::v3::PropertyStore, decltype(lam),
memgraph::storage::v3::Error>;
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,21 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/bindings/bindings.hpp"
#include "query/v2/bindings/typed_value.hpp"
#include "expr/interpret/frame.hpp"
namespace memgraph::query::v2 {
using Frame = memgraph::expr::Frame<TypedValue>;
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,19 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/bindings/bindings.hpp"
#include "expr/ast/pretty_print.hpp"
namespace memgraph::query::v2 {
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,20 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/bindings/bindings.hpp"
#include "expr/semantic/symbol.hpp"
namespace memgraph::query::v2 {
using Symbol = memgraph::expr::Symbol;
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,16 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/bindings/bindings.hpp"
#include "expr/semantic/symbol_generator.hpp"

View File

@@ -0,0 +1,20 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/bindings/bindings.hpp"
#include "expr/semantic/symbol_table.hpp"
namespace memgraph::query::v2 {
using SymbolTable = memgraph::expr::SymbolTable;
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,19 @@
// Copyright 2022 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 "expr/typed_value.hpp"
#include "query/v2/db_accessor.hpp"
#include "query/v2/path.hpp"
namespace memgraph::expr {
namespace v2 = memgraph::query::v2;
template class TypedValueT<v2::VertexAccessor, v2::EdgeAccessor, v2::Path>;
} // namespace memgraph::expr

View File

@@ -0,0 +1,26 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/bindings/bindings.hpp"
#include "expr/typed_value.hpp"
#include "query/v2/db_accessor.hpp"
#include "query/v2/path.hpp"
namespace memgraph::expr {
namespace v2 = memgraph::query::v2;
extern template class memgraph::expr::TypedValueT<v2::VertexAccessor, v2::EdgeAccessor, v2::Path>;
} // namespace memgraph::expr
namespace memgraph::query::v2 {
using TypedValue = memgraph::expr::TypedValueT<VertexAccessor, EdgeAccessor, Path>;
} // namespace memgraph::query::v2

76
src/query/v2/common.cpp Normal file
View File

@@ -0,0 +1,76 @@
// Copyright 2022 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 "query/v2/common.hpp"
namespace memgraph::query::v2 {
namespace impl {
bool TypedValueCompare(const TypedValue &a, const TypedValue &b) {
// in ordering null comes after everything else
// at the same time Null is not less that null
// first deal with Null < Whatever case
if (a.IsNull()) return false;
// now deal with NotNull < Null case
if (b.IsNull()) return true;
// comparisons are from this point legal only between values of
// the same type, or int+float combinations
if ((a.type() != b.type() && !(a.IsNumeric() && b.IsNumeric())))
throw QueryRuntimeException("Can't compare value of type {} to value of type {}.", a.type(), b.type());
switch (a.type()) {
case TypedValue::Type::Bool:
return !a.ValueBool() && b.ValueBool();
case TypedValue::Type::Int:
if (b.type() == TypedValue::Type::Double)
return a.ValueInt() < b.ValueDouble();
else
return a.ValueInt() < b.ValueInt();
case TypedValue::Type::Double:
if (b.type() == TypedValue::Type::Int)
return a.ValueDouble() < b.ValueInt();
else
return a.ValueDouble() < b.ValueDouble();
case TypedValue::Type::String:
// NOLINTNEXTLINE(modernize-use-nullptr)
return a.ValueString() < b.ValueString();
case TypedValue::Type::Date:
// NOLINTNEXTLINE(modernize-use-nullptr)
return a.ValueDate() < b.ValueDate();
case TypedValue::Type::LocalTime:
// NOLINTNEXTLINE(modernize-use-nullptr)
return a.ValueLocalTime() < b.ValueLocalTime();
case TypedValue::Type::LocalDateTime:
// NOLINTNEXTLINE(modernize-use-nullptr)
return a.ValueLocalDateTime() < b.ValueLocalDateTime();
case TypedValue::Type::Duration:
// NOLINTNEXTLINE(modernize-use-nullptr)
return a.ValueDuration() < b.ValueDuration();
case TypedValue::Type::List:
case TypedValue::Type::Map:
case TypedValue::Type::Vertex:
case TypedValue::Type::Edge:
case TypedValue::Type::Path:
throw QueryRuntimeException("Comparison is not defined for values of type {}.", a.type());
case TypedValue::Type::Null:
LOG_FATAL("Invalid type");
}
}
} // namespace impl
int64_t QueryTimestamp() {
return std::chrono::duration_cast<std::chrono::microseconds>(std::chrono::system_clock::now().time_since_epoch())
.count();
}
} // namespace memgraph::query::v2

183
src/query/v2/common.hpp Normal file
View File

@@ -0,0 +1,183 @@
// Copyright 2022 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.
/// @file
#pragma once
#include <concepts>
#include <cstdint>
#include <string>
#include <string_view>
#include <type_traits>
#include "query/v2/bindings/symbol.hpp"
#include "query/v2/bindings/typed_value.hpp"
#include "query/v2/db_accessor.hpp"
#include "query/v2/exceptions.hpp"
#include "query/v2/frontend/ast/ast.hpp"
#include "query/v2/path.hpp"
#include "storage/v3/conversions.hpp"
#include "storage/v3/id_types.hpp"
#include "storage/v3/property_value.hpp"
#include "storage/v3/result.hpp"
#include "storage/v3/schema_validator.hpp"
#include "storage/v3/view.hpp"
#include "utils/exceptions.hpp"
#include "utils/logging.hpp"
#include "utils/variant_helpers.hpp"
namespace memgraph::query::v2 {
namespace impl {
bool TypedValueCompare(const TypedValue &a, const TypedValue &b);
} // namespace impl
/// Custom Comparator type for comparing vectors of TypedValues.
///
/// Does lexicographical ordering of elements based on the above
/// defined TypedValueCompare, and also accepts a vector of Orderings
/// the define how respective elements compare.
class TypedValueVectorCompare final {
public:
TypedValueVectorCompare() {}
explicit TypedValueVectorCompare(const std::vector<Ordering> &ordering) : ordering_(ordering) {}
template <class TAllocator>
bool operator()(const std::vector<TypedValue, TAllocator> &c1, const std::vector<TypedValue, TAllocator> &c2) const {
// ordering is invalid if there are more elements in the collections
// then there are in the ordering_ vector
MG_ASSERT(c1.size() <= ordering_.size() && c2.size() <= ordering_.size(),
"Collections contain more elements then there are orderings");
auto c1_it = c1.begin();
auto c2_it = c2.begin();
auto ordering_it = ordering_.begin();
for (; c1_it != c1.end() && c2_it != c2.end(); c1_it++, c2_it++, ordering_it++) {
if (impl::TypedValueCompare(*c1_it, *c2_it)) return *ordering_it == Ordering::ASC;
if (impl::TypedValueCompare(*c2_it, *c1_it)) return *ordering_it == Ordering::DESC;
}
// at least one collection is exhausted
// c1 is less then c2 iff c1 reached the end but c2 didn't
return (c1_it == c1.end()) && (c2_it != c2.end());
}
// TODO: Remove this, member is public
const auto &ordering() const { return ordering_; }
std::vector<Ordering> ordering_;
};
/// Raise QueryRuntimeException if the value for symbol isn't of expected type.
inline void ExpectType(const Symbol &symbol, const TypedValue &value, TypedValue::Type expected) {
if (value.type() != expected)
throw QueryRuntimeException("Expected a {} for '{}', but got {}.", expected, symbol.name(), value.type());
}
template <typename T>
concept AccessorWithSetProperty = requires(T accessor, const storage::v3::PropertyId key,
const storage::v3::PropertyValue new_value) {
{ accessor.SetProperty(key, new_value) } -> std::same_as<storage::v3::Result<storage::v3::PropertyValue>>;
};
template <typename T>
concept AccessorWithSetPropertyAndValidate = requires(T accessor, const storage::v3::PropertyId key,
const storage::v3::PropertyValue new_value) {
{
accessor.SetPropertyAndValidate(key, new_value)
} -> std::same_as<storage::v3::ResultSchema<storage::v3::PropertyValue>>;
};
template <typename TRecordAccessor>
concept RecordAccessor =
AccessorWithSetProperty<TRecordAccessor> || AccessorWithSetPropertyAndValidate<TRecordAccessor>;
inline void HandleSchemaViolation(const storage::v3::SchemaViolation &schema_violation, const DbAccessor &dba) {
switch (schema_violation.status) {
case storage::v3::SchemaViolation::ValidationStatus::VERTEX_HAS_NO_PRIMARY_PROPERTY: {
throw SchemaViolationException(
fmt::format("Primary key {} not defined on label :{}",
storage::v3::SchemaTypeToString(schema_violation.violated_schema_property->type),
dba.LabelToName(schema_violation.label)));
}
case storage::v3::SchemaViolation::ValidationStatus::NO_SCHEMA_DEFINED_FOR_LABEL: {
throw SchemaViolationException(
fmt::format("Label :{} is not a primary label", dba.LabelToName(schema_violation.label)));
}
case storage::v3::SchemaViolation::ValidationStatus::VERTEX_PROPERTY_WRONG_TYPE: {
throw SchemaViolationException(
fmt::format("Wrong type of property {} in schema :{}, should be of type {}",
*schema_violation.violated_property_value, dba.LabelToName(schema_violation.label),
storage::v3::SchemaTypeToString(schema_violation.violated_schema_property->type)));
}
case storage::v3::SchemaViolation::ValidationStatus::VERTEX_UPDATE_PRIMARY_KEY: {
throw SchemaViolationException(fmt::format("Updating of primary key {} on schema :{} not supported",
*schema_violation.violated_property_value,
dba.LabelToName(schema_violation.label)));
}
case storage::v3::SchemaViolation::ValidationStatus::VERTEX_UPDATE_PRIMARY_LABEL: {
throw SchemaViolationException(fmt::format(
"Adding primary label as secondary or removing primary label:", *schema_violation.violated_property_value,
dba.LabelToName(schema_violation.label)));
}
case storage::v3::SchemaViolation::ValidationStatus::VERTEX_SECONDARY_LABEL_IS_PRIMARY: {
throw SchemaViolationException(fmt::format("Cannot create vertex where primary label is secondary:{}",
dba.LabelToName(schema_violation.label)));
}
}
}
inline void HandleErrorOnPropertyUpdate(const storage::v3::Error error) {
switch (error) {
case storage::v3::Error::SERIALIZATION_ERROR:
throw TransactionSerializationException();
case storage::v3::Error::DELETED_OBJECT:
throw QueryRuntimeException("Trying to set properties on a deleted object.");
case storage::v3::Error::PROPERTIES_DISABLED:
throw QueryRuntimeException("Can't set property because properties on edges are disabled.");
case storage::v3::Error::VERTEX_HAS_EDGES:
case storage::v3::Error::NONEXISTENT_OBJECT:
throw QueryRuntimeException("Unexpected error when setting a property.");
}
}
/// Set a property `value` mapped with given `key` on a `record`.
///
/// @throw QueryRuntimeException if value cannot be set as a property value
template <RecordAccessor T>
storage::v3::PropertyValue PropsSetChecked(T *record, const DbAccessor &dba, const storage::v3::PropertyId &key,
const TypedValue &value) {
try {
if constexpr (std::is_same_v<T, VertexAccessor>) {
const auto maybe_old_value = record->SetPropertyAndValidate(key, storage::v3::TypedToPropertyValue(value));
if (maybe_old_value.HasError()) {
std::visit(utils::Overloaded{[](const storage::v3::Error error) { HandleErrorOnPropertyUpdate(error); },
[&dba](const storage::v3::SchemaViolation &schema_violation) {
HandleSchemaViolation(schema_violation, dba);
}},
maybe_old_value.GetError());
}
return std::move(*maybe_old_value);
} else {
// No validation on edge properties
const auto maybe_old_value = record->SetProperty(key, storage::v3::TypedToPropertyValue(value));
if (maybe_old_value.HasError()) {
HandleErrorOnPropertyUpdate(maybe_old_value.GetError());
}
return std::move(*maybe_old_value);
}
} catch (const expr::TypedValueException &) {
throw QueryRuntimeException("'{}' cannot be used as a property value.", value.type());
}
}
int64_t QueryTimestamp();
} // namespace memgraph::query::v2

32
src/query/v2/config.hpp Normal file
View File

@@ -0,0 +1,32 @@
// Copyright 2022 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.
#pragma once
#include <chrono>
#include <string>
namespace memgraph::query::v2 {
struct InterpreterConfig {
struct Query {
bool allow_load_csv{true};
} query;
// The default execution timeout is 10 minutes.
double execution_timeout_sec{600.0};
// The same as \ref memgraph::storage::v3::replication::ReplicationClientConfig
std::chrono::seconds replication_replica_check_frequency{1};
std::string default_kafka_bootstrap_servers;
std::string default_pulsar_service_url;
uint32_t stream_transaction_conflict_retries;
std::chrono::milliseconds stream_transaction_retry_interval;
};
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,19 @@
// Copyright 2022 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.
#pragma once
#include <cstdint>
#include <string>
namespace memgraph::query::v2 {
inline constexpr uint16_t kDefaultReplicationPort = 10000;
inline constexpr auto *kDefaultReplicationServerIp = "0.0.0.0";
} // namespace memgraph::query::v2

89
src/query/v2/context.hpp Normal file
View File

@@ -0,0 +1,89 @@
// Copyright 2022 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.
#pragma once
#include <type_traits>
#include "query/v2/bindings/symbol_table.hpp"
#include "query/v2/common.hpp"
#include "query/v2/metadata.hpp"
#include "query/v2/parameters.hpp"
#include "query/v2/plan/profile.hpp"
#include "query/v2/trigger.hpp"
#include "utils/async_timer.hpp"
namespace memgraph::query::v2 {
struct EvaluationContext {
/// Memory for allocations during evaluation of a *single* Pull call.
///
/// Although the assigned memory may live longer than the duration of a Pull
/// (e.g. memory is the same as the whole execution memory), you have to treat
/// it as if the lifetime is only valid during the Pull.
utils::MemoryResource *memory{utils::NewDeleteResource()};
int64_t timestamp{-1};
Parameters parameters;
/// All properties indexable via PropertyIx
std::vector<storage::v3::PropertyId> properties;
/// All labels indexable via LabelIx
std::vector<storage::v3::LabelId> labels;
/// All counters generated by `counter` function, mutable because the function
/// modifies the values
mutable std::unordered_map<std::string, int64_t> counters;
};
inline std::vector<storage::v3::PropertyId> NamesToProperties(const std::vector<std::string> &property_names,
DbAccessor *dba) {
std::vector<storage::v3::PropertyId> properties;
properties.reserve(property_names.size());
for (const auto &name : property_names) {
properties.push_back(dba->NameToProperty(name));
}
return properties;
}
inline std::vector<storage::v3::LabelId> NamesToLabels(const std::vector<std::string> &label_names, DbAccessor *dba) {
std::vector<storage::v3::LabelId> labels;
labels.reserve(label_names.size());
for (const auto &name : label_names) {
labels.push_back(dba->NameToLabel(name));
}
return labels;
}
struct ExecutionContext {
DbAccessor *db_accessor{nullptr};
SymbolTable symbol_table;
EvaluationContext evaluation_context;
std::atomic<bool> *is_shutting_down{nullptr};
bool is_profile_query{false};
std::chrono::duration<double> profile_execution_time;
plan::ProfilingStats stats;
plan::ProfilingStats *stats_root{nullptr};
ExecutionStats execution_stats;
TriggerContextCollector *trigger_context_collector{nullptr};
utils::AsyncTimer timer;
};
static_assert(std::is_move_assignable_v<ExecutionContext>, "ExecutionContext must be move assignable!");
static_assert(std::is_move_constructible_v<ExecutionContext>, "ExecutionContext must be move constructible!");
inline bool MustAbort(const ExecutionContext &context) noexcept {
return (context.is_shutting_down != nullptr && context.is_shutting_down->load(std::memory_order_acquire)) ||
context.timer.IsExpired();
}
inline plan::ProfilingStatsWithTotalTime GetStatsWithTotalTime(const ExecutionContext &context) {
return plan::ProfilingStatsWithTotalTime{context.stats, context.profile_execution_time};
}
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,154 @@
// Copyright 2022 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 "query/v2/cypher_query_interpreter.hpp"
#include "query/v2/bindings/symbol_generator.hpp"
// NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
DEFINE_HIDDEN_bool(query_cost_planner, true, "Use the cost-estimating query planner.");
// NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
DEFINE_VALIDATED_int32(query_plan_cache_ttl, 60, "Time to live for cached query plans, in seconds.",
FLAG_IN_RANGE(0, std::numeric_limits<int32_t>::max()));
namespace memgraph::query::v2 {
CachedPlan::CachedPlan(std::unique_ptr<LogicalPlan> plan) : plan_(std::move(plan)) {}
ParsedQuery ParseQuery(const std::string &query_string, const std::map<std::string, storage::v3::PropertyValue> &params,
utils::SkipList<QueryCacheEntry> *cache, const InterpreterConfig::Query &query_config) {
// Strip the query for caching purposes. The process of stripping a query
// "normalizes" it by replacing any literals with new parameters. This
// results in just the *structure* of the query being taken into account for
// caching.
frontend::StrippedQuery stripped_query{query_string};
// Copy over the parameters that were introduced during stripping.
Parameters parameters{stripped_query.literals()};
// Check that all user-specified parameters are provided.
for (const auto &param_pair : stripped_query.parameters()) {
auto it = params.find(param_pair.second);
if (it == params.end()) {
throw query::v2::UnprovidedParameterError("Parameter ${} not provided.", param_pair.second);
}
parameters.Add(param_pair.first, it->second);
}
// Cache the query's AST if it isn't already.
auto hash = stripped_query.hash();
auto accessor = cache->access();
auto it = accessor.find(hash);
std::unique_ptr<memgraph::frontend::opencypher::Parser<>> parser;
// Return a copy of both the AST storage and the query.
CachedQuery result;
bool is_cacheable = true;
auto get_information_from_cache = [&](const auto &cached_query) {
result.ast_storage.properties_ = cached_query.ast_storage.properties_;
result.ast_storage.labels_ = cached_query.ast_storage.labels_;
result.ast_storage.edge_types_ = cached_query.ast_storage.edge_types_;
result.query = cached_query.query->Clone(&result.ast_storage);
result.required_privileges = cached_query.required_privileges;
};
if (it == accessor.end()) {
try {
parser = std::make_unique<memgraph::frontend::opencypher::Parser<>>(stripped_query.query());
} catch (const SyntaxException &e) {
// There is a syntax exception in the stripped query. Re-run the parser
// on the original query to get an appropriate error messsage.
parser = std::make_unique<memgraph::frontend::opencypher::Parser<>>(query_string);
// If an exception was not thrown here, the stripper messed something
// up.
LOG_FATAL("The stripped query can't be parsed, but the original can.");
}
// Convert the ANTLR4 parse tree into an AST.
AstStorage ast_storage;
expr::ParsingContext context{true};
memgraph::expr::CypherMainVisitor visitor(context, &ast_storage);
visitor.visit(parser->tree());
if (visitor.GetQueryInfo().has_load_csv && !query_config.allow_load_csv) {
throw utils::BasicException("Load CSV not allowed on this instance because it was disabled by a config.");
}
if (visitor.GetQueryInfo().is_cacheable) {
CachedQuery cached_query{std::move(ast_storage), visitor.query(),
query::v2::GetRequiredPrivileges(visitor.query())};
it = accessor.insert({hash, std::move(cached_query)}).first;
get_information_from_cache(it->second);
} else {
result.ast_storage.properties_ = ast_storage.properties_;
result.ast_storage.labels_ = ast_storage.labels_;
result.ast_storage.edge_types_ = ast_storage.edge_types_;
result.query = visitor.query()->Clone(&result.ast_storage);
result.required_privileges = query::v2::GetRequiredPrivileges(visitor.query());
is_cacheable = false;
}
} else {
get_information_from_cache(it->second);
}
return ParsedQuery{query_string,
params,
std::move(parameters),
std::move(stripped_query),
std::move(result.ast_storage),
result.query,
std::move(result.required_privileges),
is_cacheable};
}
std::unique_ptr<LogicalPlan> MakeLogicalPlan(AstStorage ast_storage, CypherQuery *query, const Parameters &parameters,
DbAccessor *db_accessor,
const std::vector<Identifier *> &predefined_identifiers) {
auto vertex_counts = plan::MakeVertexCountCache(db_accessor);
auto symbol_table = expr::MakeSymbolTable(query, predefined_identifiers);
auto planning_context = plan::MakePlanningContext(&ast_storage, &symbol_table, query, &vertex_counts);
auto [root, cost] = plan::MakeLogicalPlan(&planning_context, parameters, FLAGS_query_cost_planner);
return std::make_unique<SingleNodeLogicalPlan>(std::move(root), cost, std::move(ast_storage),
std::move(symbol_table));
}
std::shared_ptr<CachedPlan> CypherQueryToPlan(uint64_t hash, AstStorage ast_storage, CypherQuery *query,
const Parameters &parameters, utils::SkipList<PlanCacheEntry> *plan_cache,
DbAccessor *db_accessor,
const std::vector<Identifier *> &predefined_identifiers) {
std::optional<utils::SkipList<PlanCacheEntry>::Accessor> plan_cache_access;
if (plan_cache) {
plan_cache_access.emplace(plan_cache->access());
auto it = plan_cache_access->find(hash);
if (it != plan_cache_access->end()) {
if (it->second->IsExpired()) {
plan_cache_access->remove(hash);
} else {
return it->second;
}
}
}
auto plan = std::make_shared<CachedPlan>(
MakeLogicalPlan(std::move(ast_storage), query, parameters, db_accessor, predefined_identifiers));
if (plan_cache_access) {
plan_cache_access->insert({hash, plan});
}
return plan;
}
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,151 @@
// Copyright 2022 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.
#pragma once
#include "parser/opencypher/parser.hpp"
#include "query/v2/bindings/cypher_main_visitor.hpp"
#include "query/v2/bindings/symbol_table.hpp"
#include "query/v2/config.hpp"
#include "query/v2/frontend/semantic/required_privileges.hpp"
#include "query/v2/frontend/stripped.hpp"
#include "query/v2/plan/planner.hpp"
#include "utils/flag_validation.hpp"
#include "utils/timer.hpp"
// NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
DECLARE_bool(query_cost_planner);
// NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
DECLARE_int32(query_plan_cache_ttl);
namespace memgraph::query::v2 {
// TODO: Maybe this should move to query/plan/planner.
/// Interface for accessing the root operator of a logical plan.
class LogicalPlan {
public:
explicit LogicalPlan() = default;
virtual ~LogicalPlan() = default;
LogicalPlan(const LogicalPlan &) = default;
LogicalPlan &operator=(const LogicalPlan &) = default;
LogicalPlan(LogicalPlan &&) = default;
LogicalPlan &operator=(LogicalPlan &&) = default;
virtual const plan::LogicalOperator &GetRoot() const = 0;
virtual double GetCost() const = 0;
virtual const SymbolTable &GetSymbolTable() const = 0;
virtual const AstStorage &GetAstStorage() const = 0;
};
class CachedPlan {
public:
explicit CachedPlan(std::unique_ptr<LogicalPlan> plan);
const auto &plan() const { return plan_->GetRoot(); }
double cost() const { return plan_->GetCost(); }
const auto &symbol_table() const { return plan_->GetSymbolTable(); }
const auto &ast_storage() const { return plan_->GetAstStorage(); }
bool IsExpired() const {
// NOLINTNEXTLINE (modernize-use-nullptr)
return cache_timer_.Elapsed() > std::chrono::seconds(FLAGS_query_plan_cache_ttl);
};
private:
std::unique_ptr<LogicalPlan> plan_;
utils::Timer cache_timer_;
};
struct CachedQuery {
AstStorage ast_storage;
Query *query;
std::vector<AuthQuery::Privilege> required_privileges;
};
struct QueryCacheEntry {
bool operator==(const QueryCacheEntry &other) const { return first == other.first; }
bool operator<(const QueryCacheEntry &other) const { return first < other.first; }
bool operator==(const uint64_t &other) const { return first == other; }
bool operator<(const uint64_t &other) const { return first < other; }
uint64_t first;
// TODO: Maybe store the query string here and use it as a key with the hash
// so that we eliminate the risk of hash collisions.
CachedQuery second;
};
struct PlanCacheEntry {
bool operator==(const PlanCacheEntry &other) const { return first == other.first; }
bool operator<(const PlanCacheEntry &other) const { return first < other.first; }
bool operator==(const uint64_t &other) const { return first == other; }
bool operator<(const uint64_t &other) const { return first < other; }
uint64_t first;
// TODO: Maybe store the query string here and use it as a key with the hash
// so that we eliminate the risk of hash collisions.
std::shared_ptr<CachedPlan> second;
};
/**
* A container for data related to the parsing of a query.
*/
struct ParsedQuery {
std::string query_string;
std::map<std::string, storage::v3::PropertyValue> user_parameters;
Parameters parameters;
frontend::StrippedQuery stripped_query;
AstStorage ast_storage;
Query *query;
std::vector<AuthQuery::Privilege> required_privileges;
bool is_cacheable{true};
};
ParsedQuery ParseQuery(const std::string &query_string, const std::map<std::string, storage::v3::PropertyValue> &params,
utils::SkipList<QueryCacheEntry> *cache, const InterpreterConfig::Query &query_config);
class SingleNodeLogicalPlan final : public LogicalPlan {
public:
SingleNodeLogicalPlan(std::unique_ptr<plan::LogicalOperator> root, double cost, AstStorage storage,
const SymbolTable &symbol_table)
: root_(std::move(root)), cost_(cost), storage_(std::move(storage)), symbol_table_(symbol_table) {}
const plan::LogicalOperator &GetRoot() const override { return *root_; }
double GetCost() const override { return cost_; }
const SymbolTable &GetSymbolTable() const override { return symbol_table_; }
const AstStorage &GetAstStorage() const override { return storage_; }
private:
std::unique_ptr<plan::LogicalOperator> root_;
double cost_;
AstStorage storage_;
SymbolTable symbol_table_;
};
std::unique_ptr<LogicalPlan> MakeLogicalPlan(AstStorage ast_storage, CypherQuery *query, const Parameters &parameters,
DbAccessor *db_accessor,
const std::vector<Identifier *> &predefined_identifiers);
/**
* Return the parsed *Cypher* query's AST cached logical plan, or create and
* cache a fresh one if it doesn't yet exist.
* @param predefined_identifiers optional identifiers you want to inject into a query.
* If an identifier is not defined in a scope, we check the predefined identifiers.
* If an identifier is contained there, we inject it at that place and remove it,
* because a predefined identifier can be used only in one scope.
*/
std::shared_ptr<CachedPlan> CypherQueryToPlan(uint64_t hash, AstStorage ast_storage, CypherQuery *query,
const Parameters &parameters, utils::SkipList<PlanCacheEntry> *plan_cache,
DbAccessor *db_accessor,
const std::vector<Identifier *> &predefined_identifiers = {});
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,439 @@
// Copyright 2022 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.
#pragma once
#include <cstdint>
#include <optional>
#include <vector>
#include <cppitertools/filter.hpp>
#include <cppitertools/imap.hpp>
#include "query/v2/exceptions.hpp"
#include "storage/v3/id_types.hpp"
#include "storage/v3/key_store.hpp"
#include "storage/v3/property_value.hpp"
#include "storage/v3/result.hpp"
///////////////////////////////////////////////////////////
// Our communication layer and query engine don't mix
// very well on Centos because OpenSSL version available
// on Centos 7 include libkrb5 which has brilliant macros
// called TRUE and FALSE. For more detailed explanation go
// to memgraph.cpp.
//
// Because of the replication storage now uses some form of
// communication so we have some unwanted macros.
// This cannot be avoided by simple include orderings so we
// simply undefine those macros as we're sure that libkrb5
// won't and can't be used anywhere in the query engine.
#include "storage/v3/storage.hpp"
#include "utils/logging.hpp"
#include "utils/result.hpp"
#undef FALSE
#undef TRUE
///////////////////////////////////////////////////////////
#include "storage/v3/view.hpp"
#include "utils/bound.hpp"
#include "utils/exceptions.hpp"
namespace memgraph::query::v2 {
class VertexAccessor;
class EdgeAccessor final {
public:
storage::v3::EdgeAccessor impl_;
explicit EdgeAccessor(storage::v3::EdgeAccessor impl) : impl_(std::move(impl)) {}
bool IsVisible(storage::v3::View view) const { return impl_.IsVisible(view); }
storage::v3::EdgeTypeId EdgeType() const { return impl_.EdgeType(); }
auto Properties(storage::v3::View view) const { return impl_.Properties(view); }
storage::v3::Result<storage::v3::PropertyValue> GetProperty(storage::v3::View view,
storage::v3::PropertyId key) const {
return impl_.GetProperty(key, view);
}
storage::v3::Result<storage::v3::PropertyValue> SetProperty(storage::v3::PropertyId key,
const storage::v3::PropertyValue &value) {
return impl_.SetProperty(key, value);
}
storage::v3::Result<storage::v3::PropertyValue> RemoveProperty(storage::v3::PropertyId key) {
return SetProperty(key, storage::v3::PropertyValue());
}
storage::v3::Result<std::map<storage::v3::PropertyId, storage::v3::PropertyValue>> ClearProperties() {
return impl_.ClearProperties();
}
VertexAccessor To() const;
VertexAccessor From() const;
bool IsCycle() const;
int64_t CypherId() const { return impl_.Gid().AsInt(); }
storage::v3::Gid Gid() const noexcept { return impl_.Gid(); }
bool operator==(const EdgeAccessor &e) const noexcept { return impl_ == e.impl_; }
bool operator!=(const EdgeAccessor &e) const noexcept { return !(*this == e); }
};
class VertexAccessor final {
public:
storage::v3::VertexAccessor impl_;
static EdgeAccessor MakeEdgeAccessor(const storage::v3::EdgeAccessor impl) { return EdgeAccessor(impl); }
explicit VertexAccessor(storage::v3::VertexAccessor impl) : impl_(impl) {}
bool IsVisible(storage::v3::View view) const { return impl_.IsVisible(view); }
auto Labels(storage::v3::View view) const { return impl_.Labels(view); }
auto PrimaryLabel(storage::v3::View view) const { return impl_.PrimaryLabel(view); }
auto PrimaryKey(storage::v3::View view) const { return impl_.PrimaryKey(view); }
storage::v3::ResultSchema<bool> AddLabel(storage::v3::LabelId label) { return impl_.AddLabelAndValidate(label); }
storage::v3::ResultSchema<bool> AddLabelAndValidate(storage::v3::LabelId label) {
return impl_.AddLabelAndValidate(label);
}
storage::v3::ResultSchema<bool> RemoveLabel(storage::v3::LabelId label) {
return impl_.RemoveLabelAndValidate(label);
}
storage::v3::ResultSchema<bool> RemoveLabelAndValidate(storage::v3::LabelId label) {
return impl_.RemoveLabelAndValidate(label);
}
storage::v3::Result<bool> HasLabel(storage::v3::View view, storage::v3::LabelId label) const {
return impl_.HasLabel(label, view);
}
auto Properties(storage::v3::View view) const { return impl_.Properties(view); }
storage::v3::Result<storage::v3::PropertyValue> GetProperty(storage::v3::View view,
storage::v3::PropertyId key) const {
return impl_.GetProperty(key, view);
}
storage::v3::ResultSchema<storage::v3::PropertyValue> SetProperty(storage::v3::PropertyId key,
const storage::v3::PropertyValue &value) {
return impl_.SetPropertyAndValidate(key, value);
}
storage::v3::ResultSchema<storage::v3::PropertyValue> SetPropertyAndValidate(
storage::v3::PropertyId key, const storage::v3::PropertyValue &value) {
return impl_.SetPropertyAndValidate(key, value);
}
storage::v3::ResultSchema<storage::v3::PropertyValue> RemovePropertyAndValidate(storage::v3::PropertyId key) {
return SetPropertyAndValidate(key, storage::v3::PropertyValue{});
}
storage::v3::Result<std::map<storage::v3::PropertyId, storage::v3::PropertyValue>> ClearProperties() {
return impl_.ClearProperties();
}
auto InEdges(storage::v3::View view, const std::vector<storage::v3::EdgeTypeId> &edge_types) const
-> storage::v3::Result<decltype(iter::imap(MakeEdgeAccessor, *impl_.InEdges(view)))> {
auto maybe_edges = impl_.InEdges(view, edge_types);
if (maybe_edges.HasError()) return maybe_edges.GetError();
return iter::imap(MakeEdgeAccessor, std::move(*maybe_edges));
}
auto InEdges(storage::v3::View view) const { return InEdges(view, {}); }
auto InEdges(storage::v3::View view, const std::vector<storage::v3::EdgeTypeId> &edge_types,
const VertexAccessor &dest) const
-> storage::v3::Result<decltype(iter::imap(MakeEdgeAccessor, *impl_.InEdges(view)))> {
const auto dest_id = dest.impl_.Id(view).GetValue();
auto maybe_edges = impl_.InEdges(view, edge_types, &dest_id);
if (maybe_edges.HasError()) return maybe_edges.GetError();
return iter::imap(MakeEdgeAccessor, std::move(*maybe_edges));
}
auto OutEdges(storage::v3::View view, const std::vector<storage::v3::EdgeTypeId> &edge_types) const
-> storage::v3::Result<decltype(iter::imap(MakeEdgeAccessor, *impl_.OutEdges(view)))> {
auto maybe_edges = impl_.OutEdges(view, edge_types);
if (maybe_edges.HasError()) return maybe_edges.GetError();
return iter::imap(MakeEdgeAccessor, std::move(*maybe_edges));
}
auto OutEdges(storage::v3::View view) const { return OutEdges(view, {}); }
auto OutEdges(storage::v3::View view, const std::vector<storage::v3::EdgeTypeId> &edge_types,
const VertexAccessor &dest) const
-> storage::v3::Result<decltype(iter::imap(MakeEdgeAccessor, *impl_.OutEdges(view)))> {
const auto dest_id = dest.impl_.Id(view).GetValue();
auto maybe_edges = impl_.OutEdges(view, edge_types, &dest_id);
if (maybe_edges.HasError()) return maybe_edges.GetError();
return iter::imap(MakeEdgeAccessor, std::move(*maybe_edges));
}
storage::v3::Result<size_t> InDegree(storage::v3::View view) const { return impl_.InDegree(view); }
storage::v3::Result<size_t> OutDegree(storage::v3::View view) const { return impl_.OutDegree(view); }
// TODO(jbajic) Fix Remove Gid
static int64_t CypherId() { return 1; }
bool operator==(const VertexAccessor &v) const noexcept {
static_assert(noexcept(impl_ == v.impl_));
return impl_ == v.impl_;
}
bool operator!=(const VertexAccessor &v) const noexcept { return !(*this == v); }
};
#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wnull-dereference"
// NOLINTNEXTLINE(readability-convert-member-functions-to-static,clang-analyzer-core.NonNullParamChecker)
inline VertexAccessor EdgeAccessor::To() const { return *static_cast<VertexAccessor *>(nullptr); }
// NOLINTNEXTLINE(readability-convert-member-functions-to-static,clang-analyzer-core.NonNullParamChecker)
inline VertexAccessor EdgeAccessor::From() const { return *static_cast<VertexAccessor *>(nullptr); }
#pragma clang diagnostic pop
inline bool EdgeAccessor::IsCycle() const { return To() == From(); }
class DbAccessor final {
storage::v3::Shard::Accessor *accessor_;
class VerticesIterable final {
storage::v3::VerticesIterable iterable_;
public:
class Iterator final {
storage::v3::VerticesIterable::Iterator it_;
public:
explicit Iterator(storage::v3::VerticesIterable::Iterator it) : it_(it) {}
VertexAccessor operator*() const { return VertexAccessor(*it_); }
Iterator &operator++() {
++it_;
return *this;
}
bool operator==(const Iterator &other) const { return it_ == other.it_; }
bool operator!=(const Iterator &other) const { return !(other == *this); }
};
explicit VerticesIterable(storage::v3::VerticesIterable iterable) : iterable_(std::move(iterable)) {}
Iterator begin() { return Iterator(iterable_.begin()); }
Iterator end() { return Iterator(iterable_.end()); }
};
public:
explicit DbAccessor(storage::v3::Shard::Accessor *accessor) : accessor_(accessor) {}
// TODO(jbajic) Fix Remove Gid
// NOLINTNEXTLINE(readability-convert-member-functions-to-static)
std::optional<VertexAccessor> FindVertex(uint64_t /*unused*/) { return std::nullopt; }
std::optional<VertexAccessor> FindVertex(storage::v3::PrimaryKey &primary_key, storage::v3::View view) {
auto maybe_vertex = accessor_->FindVertex(primary_key, view);
if (maybe_vertex) return VertexAccessor(*maybe_vertex);
return std::nullopt;
}
void FinalizeTransaction() { accessor_->FinalizeTransaction(); }
VerticesIterable Vertices(storage::v3::View view) { return VerticesIterable(accessor_->Vertices(view)); }
VerticesIterable Vertices(storage::v3::View view, storage::v3::LabelId label) {
return VerticesIterable(accessor_->Vertices(label, view));
}
VerticesIterable Vertices(storage::v3::View view, storage::v3::LabelId label, storage::v3::PropertyId property) {
return VerticesIterable(accessor_->Vertices(label, property, view));
}
VerticesIterable Vertices(storage::v3::View view, storage::v3::LabelId label, storage::v3::PropertyId property,
const storage::v3::PropertyValue &value) {
return VerticesIterable(accessor_->Vertices(label, property, value, view));
}
VerticesIterable Vertices(storage::v3::View view, storage::v3::LabelId label, storage::v3::PropertyId property,
const std::optional<utils::Bound<storage::v3::PropertyValue>> &lower,
const std::optional<utils::Bound<storage::v3::PropertyValue>> &upper) {
return VerticesIterable(accessor_->Vertices(label, property, lower, upper, view));
}
storage::v3::ResultSchema<VertexAccessor> InsertVertexAndValidate(
const storage::v3::LabelId primary_label, const std::vector<storage::v3::LabelId> &labels,
const std::vector<std::pair<storage::v3::PropertyId, storage::v3::PropertyValue>> &properties) {
auto maybe_vertex_acc = accessor_->CreateVertexAndValidate(primary_label, labels, properties);
if (maybe_vertex_acc.HasError()) {
return {std::move(maybe_vertex_acc.GetError())};
}
return VertexAccessor{maybe_vertex_acc.GetValue()};
}
storage::v3::Result<EdgeAccessor> InsertEdge(VertexAccessor *from, VertexAccessor *to,
const storage::v3::EdgeTypeId &edge_type) {
static constexpr auto kDummyGid = storage::v3::Gid::FromUint(0);
auto maybe_edge = accessor_->CreateEdge(from->impl_.Id(storage::v3::View::NEW).GetValue(),
to->impl_.Id(storage::v3::View::NEW).GetValue(), edge_type, kDummyGid);
if (maybe_edge.HasError()) return storage::v3::Result<EdgeAccessor>(maybe_edge.GetError());
return EdgeAccessor(*maybe_edge);
}
storage::v3::Result<std::optional<EdgeAccessor>> RemoveEdge(EdgeAccessor *edge) {
auto res = accessor_->DeleteEdge(edge->impl_.FromVertex(), edge->impl_.ToVertex(), edge->impl_.Gid());
if (res.HasError()) {
return res.GetError();
}
const auto &value = res.GetValue();
if (!value) {
return std::optional<EdgeAccessor>{};
}
return std::make_optional<EdgeAccessor>(*value);
}
storage::v3::Result<std::optional<std::pair<VertexAccessor, std::vector<EdgeAccessor>>>> DetachRemoveVertex(
VertexAccessor *vertex_accessor) {
using ReturnType = std::pair<VertexAccessor, std::vector<EdgeAccessor>>;
auto res = accessor_->DetachDeleteVertex(&vertex_accessor->impl_);
if (res.HasError()) {
return res.GetError();
}
const auto &value = res.GetValue();
if (!value) {
return std::optional<ReturnType>{};
}
const auto &[vertex, edges] = *value;
std::vector<EdgeAccessor> deleted_edges;
deleted_edges.reserve(edges.size());
std::transform(edges.begin(), edges.end(), std::back_inserter(deleted_edges),
[](const auto &deleted_edge) { return EdgeAccessor{deleted_edge}; });
return std::make_optional<ReturnType>(vertex, std::move(deleted_edges));
}
storage::v3::Result<std::optional<VertexAccessor>> RemoveVertex(VertexAccessor *vertex_accessor) {
auto res = accessor_->DeleteVertex(&vertex_accessor->impl_);
if (res.HasError()) {
return res.GetError();
}
const auto &value = res.GetValue();
if (!value) {
return std::optional<VertexAccessor>{};
}
return {std::make_optional<VertexAccessor>(*value)};
}
// TODO(jbajic) Query engine should have a map of labels, properties and edge
// types
// NOLINTNEXTLINE(readability-convert-member-functions-to-static)
storage::v3::PropertyId NameToProperty(const std::string_view /*name*/) {
return storage::v3::PropertyId::FromUint(0);
}
// NOLINTNEXTLINE(readability-convert-member-functions-to-static)
storage::v3::LabelId NameToLabel(const std::string_view /*name*/) { return storage::v3::LabelId::FromUint(0); }
// NOLINTNEXTLINE(readability-convert-member-functions-to-static)
storage::v3::EdgeTypeId NameToEdgeType(const std::string_view /*name*/) {
return storage::v3::EdgeTypeId::FromUint(0);
}
const std::string &PropertyToName(storage::v3::PropertyId prop) const { return accessor_->PropertyToName(prop); }
const std::string &LabelToName(storage::v3::LabelId label) const { return accessor_->LabelToName(label); }
const std::string &EdgeTypeToName(storage::v3::EdgeTypeId type) const { return accessor_->EdgeTypeToName(type); }
void AdvanceCommand() { accessor_->AdvanceCommand(); }
utils::BasicResult<storage::v3::ConstraintViolation, void> Commit() { return accessor_->Commit(); }
void Abort() { accessor_->Abort(); }
bool LabelIndexExists(storage::v3::LabelId label) const { return accessor_->LabelIndexExists(label); }
bool LabelPropertyIndexExists(storage::v3::LabelId label, storage::v3::PropertyId prop) const {
return accessor_->LabelPropertyIndexExists(label, prop);
}
int64_t VerticesCount() const { return accessor_->ApproximateVertexCount(); }
int64_t VerticesCount(storage::v3::LabelId label) const { return accessor_->ApproximateVertexCount(label); }
int64_t VerticesCount(storage::v3::LabelId label, storage::v3::PropertyId property) const {
return accessor_->ApproximateVertexCount(label, property);
}
int64_t VerticesCount(storage::v3::LabelId label, storage::v3::PropertyId property,
const storage::v3::PropertyValue &value) const {
return accessor_->ApproximateVertexCount(label, property, value);
}
int64_t VerticesCount(storage::v3::LabelId label, storage::v3::PropertyId property,
const std::optional<utils::Bound<storage::v3::PropertyValue>> &lower,
const std::optional<utils::Bound<storage::v3::PropertyValue>> &upper) const {
return accessor_->ApproximateVertexCount(label, property, lower, upper);
}
storage::v3::IndicesInfo ListAllIndices() const { return accessor_->ListAllIndices(); }
storage::v3::ConstraintsInfo ListAllConstraints() const { return accessor_->ListAllConstraints(); }
const storage::v3::SchemaValidator &GetSchemaValidator() const { return accessor_->GetSchemaValidator(); }
storage::v3::SchemasInfo ListAllSchemas() const { return accessor_->ListAllSchemas(); }
};
} // namespace memgraph::query::v2
namespace std {
template <>
struct hash<memgraph::query::v2::VertexAccessor> {
size_t operator()(const memgraph::query::v2::VertexAccessor &v) const {
return std::hash<decltype(v.impl_)>{}(v.impl_);
}
};
template <>
struct hash<memgraph::query::v2::EdgeAccessor> {
size_t operator()(const memgraph::query::v2::EdgeAccessor &e) const {
return std::hash<decltype(e.impl_)>{}(e.impl_);
}
};
} // namespace std

View File

@@ -0,0 +1,24 @@
// Copyright 2022 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.
#pragma once
#include <vector>
#include "query/v2/bindings/typed_value.hpp"
namespace memgraph::query::v2 {
struct DiscardValueResultStream final {
void Result(const std::vector<query::v2::TypedValue> & /*values*/) {
// do nothing
}
};
} // namespace memgraph::query::v2

541
src/query/v2/dump.cpp Normal file
View File

@@ -0,0 +1,541 @@
// Copyright 2022 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 "query/v2/dump.hpp"
#include <iomanip>
#include <limits>
#include <map>
#include <optional>
#include <ostream>
#include <utility>
#include <vector>
#include <fmt/format.h>
#include "query/v2/bindings/typed_value.hpp"
#include "query/v2/db_accessor.hpp"
#include "query/v2/exceptions.hpp"
#include "query/v2/stream.hpp"
#include "storage/v3/property_value.hpp"
#include "storage/v3/storage.hpp"
#include "utils/algorithm.hpp"
#include "utils/logging.hpp"
#include "utils/string.hpp"
#include "utils/temporal.hpp"
namespace memgraph::query::v2 {
namespace {
// Property that is used to make a difference among vertices. It is added to
// property set of vertices to match edges and removed after the entire graph
// is built.
const char *kInternalPropertyId = "__mg_id__";
// Label that is attached to each vertex and is used for easier creation of
// index on internal property id.
const char *kInternalVertexLabel = "__mg_vertex__";
/// A helper function that escapes label, edge type and property names.
std::string EscapeName(const std::string_view value) {
std::string out;
out.reserve(value.size() + 2);
out.append(1, '`');
for (auto c : value) {
if (c == '`') {
out.append("``");
} else {
out.append(1, c);
}
}
out.append(1, '`');
return out;
}
void DumpPreciseDouble(std::ostream *os, double value) {
// A temporary stream is used to keep precision of the original output
// stream unchanged.
std::ostringstream temp_oss;
temp_oss << std::setprecision(std::numeric_limits<double>::max_digits10) << value;
*os << temp_oss.str();
}
namespace {
void DumpDate(std::ostream &os, const storage::v3::TemporalData &value) {
utils::Date date(value.microseconds);
os << "DATE(\"" << date << "\")";
}
void DumpLocalTime(std::ostream &os, const storage::v3::TemporalData &value) {
utils::LocalTime lt(value.microseconds);
os << "LOCALTIME(\"" << lt << "\")";
}
void DumpLocalDateTime(std::ostream &os, const storage::v3::TemporalData &value) {
utils::LocalDateTime ldt(value.microseconds);
os << "LOCALDATETIME(\"" << ldt << "\")";
}
void DumpDuration(std::ostream &os, const storage::v3::TemporalData &value) {
utils::Duration dur(value.microseconds);
os << "DURATION(\"" << dur << "\")";
}
void DumpTemporalData(std::ostream &os, const storage::v3::TemporalData &value) {
switch (value.type) {
case storage::v3::TemporalType::Date: {
DumpDate(os, value);
return;
}
case storage::v3::TemporalType::LocalTime: {
DumpLocalTime(os, value);
return;
}
case storage::v3::TemporalType::LocalDateTime: {
DumpLocalDateTime(os, value);
return;
}
case storage::v3::TemporalType::Duration: {
DumpDuration(os, value);
return;
}
}
}
} // namespace
void DumpPropertyValue(std::ostream *os, const storage::v3::PropertyValue &value) {
switch (value.type()) {
case storage::v3::PropertyValue::Type::Null:
*os << "Null";
return;
case storage::v3::PropertyValue::Type::Bool:
*os << (value.ValueBool() ? "true" : "false");
return;
case storage::v3::PropertyValue::Type::String:
*os << utils::Escape(value.ValueString());
return;
case storage::v3::PropertyValue::Type::Int:
*os << value.ValueInt();
return;
case storage::v3::PropertyValue::Type::Double:
DumpPreciseDouble(os, value.ValueDouble());
return;
case storage::v3::PropertyValue::Type::List: {
*os << "[";
const auto &list = value.ValueList();
utils::PrintIterable(*os, list, ", ", [](auto &os, const auto &item) { DumpPropertyValue(&os, item); });
*os << "]";
return;
}
case storage::v3::PropertyValue::Type::Map: {
*os << "{";
const auto &map = value.ValueMap();
utils::PrintIterable(*os, map, ", ", [](auto &os, const auto &kv) {
os << EscapeName(kv.first) << ": ";
DumpPropertyValue(&os, kv.second);
});
*os << "}";
return;
}
case storage::v3::PropertyValue::Type::TemporalData: {
DumpTemporalData(*os, value.ValueTemporalData());
return;
}
}
}
void DumpProperties(std::ostream *os, query::v2::DbAccessor *dba,
const std::map<storage::v3::PropertyId, storage::v3::PropertyValue> &store,
std::optional<int64_t> property_id = std::nullopt) {
*os << "{";
if (property_id) {
*os << kInternalPropertyId << ": " << *property_id;
if (store.size() > 0) *os << ", ";
}
utils::PrintIterable(*os, store, ", ", [&dba](auto &os, const auto &kv) {
os << EscapeName(dba->PropertyToName(kv.first)) << ": ";
DumpPropertyValue(&os, kv.second);
});
*os << "}";
}
void DumpVertex(std::ostream *os, query::v2::DbAccessor *dba, const query::v2::VertexAccessor &vertex) {
*os << "CREATE (";
*os << ":" << kInternalVertexLabel;
auto maybe_labels = vertex.Labels(storage::v3::View::OLD);
if (maybe_labels.HasError()) {
switch (maybe_labels.GetError()) {
case storage::v3::Error::DELETED_OBJECT:
throw query::v2::QueryRuntimeException("Trying to get labels from a deleted node.");
case storage::v3::Error::NONEXISTENT_OBJECT:
throw query::v2::QueryRuntimeException("Trying to get labels from a node that doesn't exist.");
case storage::v3::Error::SERIALIZATION_ERROR:
case storage::v3::Error::VERTEX_HAS_EDGES:
case storage::v3::Error::PROPERTIES_DISABLED:
throw query::v2::QueryRuntimeException("Unexpected error when getting labels.");
}
}
for (const auto &label : *maybe_labels) {
*os << ":" << EscapeName(dba->LabelToName(label));
}
*os << " ";
auto maybe_props = vertex.Properties(storage::v3::View::OLD);
if (maybe_props.HasError()) {
switch (maybe_props.GetError()) {
case storage::v3::Error::DELETED_OBJECT:
throw query::v2::QueryRuntimeException("Trying to get properties from a deleted object.");
case storage::v3::Error::NONEXISTENT_OBJECT:
throw query::v2::QueryRuntimeException("Trying to get properties from a node that doesn't exist.");
case storage::v3::Error::SERIALIZATION_ERROR:
case storage::v3::Error::VERTEX_HAS_EDGES:
case storage::v3::Error::PROPERTIES_DISABLED:
throw query::v2::QueryRuntimeException("Unexpected error when getting properties.");
}
}
DumpProperties(os, dba, *maybe_props, vertex.CypherId());
*os << ");";
}
void DumpEdge(std::ostream *os, query::v2::DbAccessor *dba, const query::v2::EdgeAccessor &edge) {
*os << "MATCH ";
*os << "(u:" << kInternalVertexLabel << "), ";
*os << "(v:" << kInternalVertexLabel << ")";
*os << " WHERE ";
*os << "u." << kInternalPropertyId << " = " << edge.From().CypherId();
*os << " AND ";
*os << "v." << kInternalPropertyId << " = " << edge.To().CypherId() << " ";
*os << "CREATE (u)-[";
*os << ":" << EscapeName(dba->EdgeTypeToName(edge.EdgeType()));
auto maybe_props = edge.Properties(storage::v3::View::OLD);
if (maybe_props.HasError()) {
switch (maybe_props.GetError()) {
case storage::v3::Error::DELETED_OBJECT:
throw query::v2::QueryRuntimeException("Trying to get properties from a deleted object.");
case storage::v3::Error::NONEXISTENT_OBJECT:
throw query::v2::QueryRuntimeException("Trying to get properties from an edge that doesn't exist.");
case storage::v3::Error::SERIALIZATION_ERROR:
case storage::v3::Error::VERTEX_HAS_EDGES:
case storage::v3::Error::PROPERTIES_DISABLED:
throw query::v2::QueryRuntimeException("Unexpected error when getting properties.");
}
}
if (maybe_props->size() > 0) {
*os << " ";
DumpProperties(os, dba, *maybe_props);
}
*os << "]->(v);";
}
void DumpLabelIndex(std::ostream *os, query::v2::DbAccessor *dba, const storage::v3::LabelId label) {
*os << "CREATE INDEX ON :" << EscapeName(dba->LabelToName(label)) << ";";
}
void DumpLabelPropertyIndex(std::ostream *os, query::v2::DbAccessor *dba, storage::v3::LabelId label,
storage::v3::PropertyId property) {
*os << "CREATE INDEX ON :" << EscapeName(dba->LabelToName(label)) << "(" << EscapeName(dba->PropertyToName(property))
<< ");";
}
void DumpExistenceConstraint(std::ostream *os, query::v2::DbAccessor *dba, storage::v3::LabelId label,
storage::v3::PropertyId property) {
*os << "CREATE CONSTRAINT ON (u:" << EscapeName(dba->LabelToName(label)) << ") ASSERT EXISTS (u."
<< EscapeName(dba->PropertyToName(property)) << ");";
}
void DumpUniqueConstraint(std::ostream *os, query::v2::DbAccessor *dba, storage::v3::LabelId label,
const std::set<storage::v3::PropertyId> &properties) {
*os << "CREATE CONSTRAINT ON (u:" << EscapeName(dba->LabelToName(label)) << ") ASSERT ";
utils::PrintIterable(*os, properties, ", ", [&dba](auto &stream, const auto &property) {
stream << "u." << EscapeName(dba->PropertyToName(property));
});
*os << " IS UNIQUE;";
}
} // namespace
PullPlanDump::PullPlanDump(DbAccessor *dba)
: dba_(dba),
vertices_iterable_(dba->Vertices(storage::v3::View::OLD)),
pull_chunks_{// Dump all label indices
CreateLabelIndicesPullChunk(),
// Dump all label property indices
CreateLabelPropertyIndicesPullChunk(),
// Dump all existence constraints
CreateExistenceConstraintsPullChunk(),
// Dump all unique constraints
CreateUniqueConstraintsPullChunk(),
// Create internal index for faster edge creation
CreateInternalIndexPullChunk(),
// Dump all vertices
CreateVertexPullChunk(),
// Dump all edges
CreateEdgePullChunk(),
// Drop the internal index
CreateDropInternalIndexPullChunk(),
// Internal index cleanup
CreateInternalIndexCleanupPullChunk()} {}
bool PullPlanDump::Pull(AnyStream *stream, std::optional<int> n) {
// Iterate all functions that stream some results.
// Each function should return number of results it streamed after it
// finishes. If the function did not finish streaming all the results,
// std::nullopt should be returned because n results have already been sent.
while (current_chunk_index_ < pull_chunks_.size() && (!n || *n > 0)) {
const auto maybe_streamed_count = pull_chunks_[current_chunk_index_](stream, n);
if (!maybe_streamed_count) {
// n wasn't large enough to stream all the results from the current chunk
break;
}
if (n) {
// chunk finished streaming its results
// subtract number of results streamed in current pull
// so we know how many results we need to stream from future
// chunks.
*n -= *maybe_streamed_count;
}
++current_chunk_index_;
}
return current_chunk_index_ == pull_chunks_.size();
}
PullPlanDump::PullChunk PullPlanDump::CreateLabelIndicesPullChunk() {
// Dump all label indices
return [this, global_index = 0U](AnyStream *stream, std::optional<int> n) mutable -> std::optional<size_t> {
// Delay the construction of indices vectors
if (!indices_info_) {
indices_info_.emplace(dba_->ListAllIndices());
}
const auto &label = indices_info_->label;
size_t local_counter = 0;
while (global_index < label.size() && (!n || local_counter < *n)) {
std::ostringstream os;
DumpLabelIndex(&os, dba_, label[global_index]);
stream->Result({TypedValue(os.str())});
++global_index;
++local_counter;
}
if (global_index == label.size()) {
return local_counter;
}
return std::nullopt;
};
}
PullPlanDump::PullChunk PullPlanDump::CreateLabelPropertyIndicesPullChunk() {
return [this, global_index = 0U](AnyStream *stream, std::optional<int> n) mutable -> std::optional<size_t> {
// Delay the construction of indices vectors
if (!indices_info_) {
indices_info_.emplace(dba_->ListAllIndices());
}
const auto &label_property = indices_info_->label_property;
size_t local_counter = 0;
while (global_index < label_property.size() && (!n || local_counter < *n)) {
std::ostringstream os;
const auto &label_property_index = label_property[global_index];
DumpLabelPropertyIndex(&os, dba_, label_property_index.first, label_property_index.second);
stream->Result({TypedValue(os.str())});
++global_index;
++local_counter;
}
if (global_index == label_property.size()) {
return local_counter;
}
return std::nullopt;
};
}
PullPlanDump::PullChunk PullPlanDump::CreateExistenceConstraintsPullChunk() {
return [this, global_index = 0U](AnyStream *stream, std::optional<int> n) mutable -> std::optional<size_t> {
// Delay the construction of constraint vectors
if (!constraints_info_) {
constraints_info_.emplace(dba_->ListAllConstraints());
}
const auto &existence = constraints_info_->existence;
size_t local_counter = 0;
while (global_index < existence.size() && (!n || local_counter < *n)) {
const auto &constraint = existence[global_index];
std::ostringstream os;
DumpExistenceConstraint(&os, dba_, constraint.first, constraint.second);
stream->Result({TypedValue(os.str())});
++global_index;
++local_counter;
}
if (global_index == existence.size()) {
return local_counter;
}
return std::nullopt;
};
}
PullPlanDump::PullChunk PullPlanDump::CreateUniqueConstraintsPullChunk() {
return [this, global_index = 0U](AnyStream *stream, std::optional<int> n) mutable -> std::optional<size_t> {
// Delay the construction of constraint vectors
if (!constraints_info_) {
constraints_info_.emplace(dba_->ListAllConstraints());
}
const auto &unique = constraints_info_->unique;
size_t local_counter = 0;
while (global_index < unique.size() && (!n || local_counter < *n)) {
const auto &constraint = unique[global_index];
std::ostringstream os;
DumpUniqueConstraint(&os, dba_, constraint.first, constraint.second);
stream->Result({TypedValue(os.str())});
++global_index;
++local_counter;
}
if (global_index == unique.size()) {
return local_counter;
}
return std::nullopt;
};
}
PullPlanDump::PullChunk PullPlanDump::CreateInternalIndexPullChunk() {
return [this](AnyStream *stream, std::optional<int>) mutable -> std::optional<size_t> {
if (vertices_iterable_.begin() != vertices_iterable_.end()) {
std::ostringstream os;
os << "CREATE INDEX ON :" << kInternalVertexLabel << "(" << kInternalPropertyId << ");";
stream->Result({TypedValue(os.str())});
internal_index_created_ = true;
return 1;
}
return 0;
};
}
PullPlanDump::PullChunk PullPlanDump::CreateVertexPullChunk() {
return [this, maybe_current_iter = std::optional<VertexAccessorIterableIterator>{}](
AnyStream *stream, std::optional<int> n) mutable -> std::optional<size_t> {
// Delay the call of begin() function
// If multiple begins are called before an iteration,
// one iteration will make the rest of iterators be in undefined
// states.
if (!maybe_current_iter) {
maybe_current_iter.emplace(vertices_iterable_.begin());
}
auto &current_iter{*maybe_current_iter};
size_t local_counter = 0;
while (current_iter != vertices_iterable_.end() && (!n || local_counter < *n)) {
std::ostringstream os;
DumpVertex(&os, dba_, *current_iter);
stream->Result({TypedValue(os.str())});
++local_counter;
++current_iter;
}
if (current_iter == vertices_iterable_.end()) {
return local_counter;
}
return std::nullopt;
};
}
PullPlanDump::PullChunk PullPlanDump::CreateEdgePullChunk() {
return [this, maybe_current_vertex_iter = std::optional<VertexAccessorIterableIterator>{},
// we need to save the iterable which contains list of accessor so
// our saved iterator is valid in the next run
maybe_edge_iterable = std::shared_ptr<EdgeAccessorIterable>{nullptr},
maybe_current_edge_iter = std::optional<EdgeAccessorIterableIterator>{}](
AnyStream *stream, std::optional<int> n) mutable -> std::optional<size_t> {
// Delay the call of begin() function
// If multiple begins are called before an iteration,
// one iteration will make the rest of iterators be in undefined
// states.
if (!maybe_current_vertex_iter) {
maybe_current_vertex_iter.emplace(vertices_iterable_.begin());
}
auto &current_vertex_iter{*maybe_current_vertex_iter};
size_t local_counter = 0U;
for (; current_vertex_iter != vertices_iterable_.end() && (!n || local_counter < *n); ++current_vertex_iter) {
const auto &vertex = *current_vertex_iter;
// If we have a saved iterable from a previous pull
// we need to use the same iterable
if (!maybe_edge_iterable) {
maybe_edge_iterable = std::make_shared<EdgeAccessorIterable>(vertex.OutEdges(storage::v3::View::OLD));
}
auto &maybe_edges = *maybe_edge_iterable;
MG_ASSERT(maybe_edges.HasValue(), "Invalid database state!");
auto current_edge_iter = maybe_current_edge_iter ? *maybe_current_edge_iter : maybe_edges->begin();
for (; current_edge_iter != maybe_edges->end() && (!n || local_counter < *n); ++current_edge_iter) {
std::ostringstream os;
DumpEdge(&os, dba_, *current_edge_iter);
stream->Result({TypedValue(os.str())});
++local_counter;
}
if (current_edge_iter != maybe_edges->end()) {
maybe_current_edge_iter.emplace(current_edge_iter);
return std::nullopt;
}
maybe_current_edge_iter = std::nullopt;
maybe_edge_iterable = nullptr;
}
if (current_vertex_iter == vertices_iterable_.end()) {
return local_counter;
}
return std::nullopt;
};
}
PullPlanDump::PullChunk PullPlanDump::CreateDropInternalIndexPullChunk() {
return [this](AnyStream *stream, std::optional<int>) {
if (internal_index_created_) {
std::ostringstream os;
os << "DROP INDEX ON :" << kInternalVertexLabel << "(" << kInternalPropertyId << ");";
stream->Result({TypedValue(os.str())});
return 1;
}
return 0;
};
}
PullPlanDump::PullChunk PullPlanDump::CreateInternalIndexCleanupPullChunk() {
return [this](AnyStream *stream, std::optional<int>) {
if (internal_index_created_) {
std::ostringstream os;
os << "MATCH (u) REMOVE u:" << kInternalVertexLabel << ", u." << kInternalPropertyId << ";";
stream->Result({TypedValue(os.str())});
return 1;
}
return 0;
};
}
void DumpDatabaseToCypherQueries(query::v2::DbAccessor *dba, AnyStream *stream) { PullPlanDump(dba).Pull(stream, {}); }
} // namespace memgraph::query::v2

66
src/query/v2/dump.hpp Normal file
View File

@@ -0,0 +1,66 @@
// Copyright 2022 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.
#pragma once
#include <ostream>
#include "query/v2/db_accessor.hpp"
#include "query/v2/stream.hpp"
#include "storage/v3/storage.hpp"
namespace memgraph::query::v2 {
void DumpDatabaseToCypherQueries(query::v2::DbAccessor *dba, AnyStream *stream);
struct PullPlanDump {
explicit PullPlanDump(query::v2::DbAccessor *dba);
/// Pull the dump results lazily
/// @return true if all results were returned, false otherwise
bool Pull(AnyStream *stream, std::optional<int> n);
private:
query::v2::DbAccessor *dba_ = nullptr;
std::optional<storage::v3::IndicesInfo> indices_info_ = std::nullopt;
std::optional<storage::v3::ConstraintsInfo> constraints_info_ = std::nullopt;
using VertexAccessorIterable = decltype(std::declval<query::v2::DbAccessor>().Vertices(storage::v3::View::OLD));
using VertexAccessorIterableIterator = decltype(std::declval<VertexAccessorIterable>().begin());
using EdgeAccessorIterable = decltype(std::declval<VertexAccessor>().OutEdges(storage::v3::View::OLD));
using EdgeAccessorIterableIterator = decltype(std::declval<EdgeAccessorIterable>().GetValue().begin());
VertexAccessorIterable vertices_iterable_;
bool internal_index_created_ = false;
size_t current_chunk_index_ = 0;
using PullChunk = std::function<std::optional<size_t>(AnyStream *stream, std::optional<int> n)>;
// We define every part of the dump query in a self contained function.
// Each functions is responsible of keeping track of its execution status.
// If a function did finish its execution, it should return number of results
// it streamed so we know how many rows should be pulled from the next
// function, otherwise std::nullopt is returned.
std::vector<PullChunk> pull_chunks_;
PullChunk CreateLabelIndicesPullChunk();
PullChunk CreateLabelPropertyIndicesPullChunk();
PullChunk CreateExistenceConstraintsPullChunk();
PullChunk CreateUniqueConstraintsPullChunk();
PullChunk CreateInternalIndexPullChunk();
PullChunk CreateVertexPullChunk();
PullChunk CreateEdgePullChunk();
PullChunk CreateDropInternalIndexPullChunk();
PullChunk CreateInternalIndexCleanupPullChunk();
};
} // namespace memgraph::query::v2

235
src/query/v2/exceptions.hpp Normal file
View File

@@ -0,0 +1,235 @@
// Copyright 2022 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.
#pragma once
#include "utils/exceptions.hpp"
#include <fmt/format.h>
namespace memgraph::query::v2 {
/**
* @brief Base class of all query language related exceptions. All exceptions
* derived from this one will be interpreted as ClientError-s, i. e. if client
* executes same query again without making modifications to the database data,
* query will fail again.
*/
class QueryException : public utils::BasicException {
using utils::BasicException::BasicException;
};
class LexingException : public QueryException {
public:
using QueryException::QueryException;
LexingException() : QueryException("") {}
};
class SyntaxException : public QueryException {
public:
using QueryException::QueryException;
SyntaxException() : QueryException("") {}
};
// TODO: Figure out what information to put in exception.
// Error reporting is tricky since we get stripped query and position of error
// in original query is not same as position of error in stripped query. Most
// correct approach would be to do semantic analysis with original query even
// for already hashed queries, but that has obvious performance issues. Other
// approach would be to report some of the semantic errors in runtime of the
// query and only report line numbers of semantic errors (not position in the
// line) if multiple line strings are not allowed by grammar. We could also
// print whole line that contains error instead of specifying line number.
class SemanticException : public QueryException {
public:
using QueryException::QueryException;
SemanticException() : QueryException("") {}
};
class UnboundVariableError : public SemanticException {
public:
explicit UnboundVariableError(const std::string &name) : SemanticException("Unbound variable: " + name + ".") {}
};
class RedeclareVariableError : public SemanticException {
public:
explicit RedeclareVariableError(const std::string &name) : SemanticException("Redeclaring variable: " + name + ".") {}
};
class TypeMismatchError : public SemanticException {
public:
TypeMismatchError(const std::string &name, const std::string &datum, const std::string &expected)
: SemanticException(fmt::format("Type mismatch: {} already defined as {}, expected {}.", name, datum, expected)) {
}
};
class UnprovidedParameterError : public QueryException {
public:
using QueryException::QueryException;
};
class ProfileInMulticommandTxException : public QueryException {
public:
using QueryException::QueryException;
ProfileInMulticommandTxException() : QueryException("PROFILE not allowed in multicommand transactions.") {}
};
class IndexInMulticommandTxException : public QueryException {
public:
using QueryException::QueryException;
IndexInMulticommandTxException() : QueryException("Index manipulation not allowed in multicommand transactions.") {}
};
class ConstraintInMulticommandTxException : public QueryException {
public:
using QueryException::QueryException;
ConstraintInMulticommandTxException()
: QueryException(
"Constraint manipulation not allowed in multicommand "
"transactions.") {}
};
class InfoInMulticommandTxException : public QueryException {
public:
using QueryException::QueryException;
InfoInMulticommandTxException() : QueryException("Info reporting not allowed in multicommand transactions.") {}
};
/**
* An exception for an illegal operation that can not be detected
* before the query starts executing over data.
*/
class QueryRuntimeException : public QueryException {
public:
using QueryException::QueryException;
};
// This one is inherited from BasicException and will be treated as
// TransientError, i. e. client will be encouraged to retry execution because it
// could succeed if executed again.
class HintedAbortError : public utils::BasicException {
public:
using utils::BasicException::BasicException;
HintedAbortError()
: utils::BasicException(
"Transaction was asked to abort, most likely because it was "
"executing longer than time specified by "
"--query-execution-timeout-sec flag.") {}
};
class ExplicitTransactionUsageException : public QueryRuntimeException {
public:
using QueryRuntimeException::QueryRuntimeException;
};
/**
* An exception for serialization error
*/
class TransactionSerializationException : public QueryException {
public:
using QueryException::QueryException;
TransactionSerializationException()
: QueryException(
"Cannot resolve conflicting transactions. You can retry this transaction when the conflicting transaction "
"is finished") {}
};
class ReconstructionException : public QueryException {
public:
ReconstructionException()
: QueryException(
"Record invalid after WITH clause. Most likely deleted by a "
"preceeding DELETE.") {}
};
class RemoveAttachedVertexException : public QueryRuntimeException {
public:
RemoveAttachedVertexException()
: QueryRuntimeException(
"Failed to remove node because of it's existing "
"connections. Consider using DETACH DELETE.") {}
};
class UserModificationInMulticommandTxException : public QueryException {
public:
UserModificationInMulticommandTxException()
: QueryException("Authentication clause not allowed in multicommand transactions.") {}
};
class InvalidArgumentsException : public QueryException {
public:
InvalidArgumentsException(const std::string &argument_name, const std::string &message)
: QueryException(fmt::format("Invalid arguments sent: {} - {}", argument_name, message)) {}
};
class ReplicationModificationInMulticommandTxException : public QueryException {
public:
ReplicationModificationInMulticommandTxException()
: QueryException("Replication clause not allowed in multicommand transactions.") {}
};
class LockPathModificationInMulticommandTxException : public QueryException {
public:
LockPathModificationInMulticommandTxException()
: QueryException("Lock path query not allowed in multicommand transactions.") {}
};
class FreeMemoryModificationInMulticommandTxException : public QueryException {
public:
FreeMemoryModificationInMulticommandTxException()
: QueryException("Free memory query not allowed in multicommand transactions.") {}
};
class TriggerModificationInMulticommandTxException : public QueryException {
public:
TriggerModificationInMulticommandTxException()
: QueryException("Trigger queries not allowed in multicommand transactions.") {}
};
class StreamQueryInMulticommandTxException : public QueryException {
public:
StreamQueryInMulticommandTxException()
: QueryException("Stream queries are not allowed in multicommand transactions.") {}
};
class IsolationLevelModificationInMulticommandTxException : public QueryException {
public:
IsolationLevelModificationInMulticommandTxException()
: QueryException("Isolation level cannot be modified in multicommand transactions.") {}
};
class CreateSnapshotInMulticommandTxException final : public QueryException {
public:
CreateSnapshotInMulticommandTxException()
: QueryException("Snapshot cannot be created in multicommand transactions.") {}
};
class SettingConfigInMulticommandTxException final : public QueryException {
public:
SettingConfigInMulticommandTxException()
: QueryException("Settings cannot be changed or fetched in multicommand transactions.") {}
};
class VersionInfoInMulticommandTxException : public QueryException {
public:
VersionInfoInMulticommandTxException()
: QueryException("Version info query not allowed in multicommand transactions.") {}
};
/**
* An exception for an illegal operation that violates schema
*/
class SchemaViolationException : public QueryRuntimeException {
public:
using QueryRuntimeException::QueryRuntimeException;
};
} // namespace memgraph::query::v2

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,154 @@
// Copyright 2022 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 "query/v2/bindings/ast_visitor.hpp"
#include "query/v2/frontend/ast/ast.hpp"
#include "query/v2/procedure/module.hpp"
#include "utils/memory.hpp"
namespace memgraph::query::v2 {
class PrivilegeExtractor : public QueryVisitor<void>, public HierarchicalTreeVisitor {
public:
using HierarchicalTreeVisitor::PostVisit;
using HierarchicalTreeVisitor::PreVisit;
using HierarchicalTreeVisitor::Visit;
using QueryVisitor<void>::Visit;
std::vector<AuthQuery::Privilege> privileges() { return privileges_; }
void Visit(IndexQuery &) override { AddPrivilege(AuthQuery::Privilege::INDEX); }
void Visit(AuthQuery &) override { AddPrivilege(AuthQuery::Privilege::AUTH); }
void Visit(ExplainQuery &query) override { query.cypher_query_->Accept(*this); }
void Visit(ProfileQuery &query) override { query.cypher_query_->Accept(*this); }
void Visit(InfoQuery &info_query) override {
switch (info_query.info_type_) {
case InfoQuery::InfoType::INDEX:
// TODO: This should be INDEX | STATS, but we don't have support for
// *or* with privileges.
AddPrivilege(AuthQuery::Privilege::INDEX);
break;
case InfoQuery::InfoType::STORAGE:
AddPrivilege(AuthQuery::Privilege::STATS);
break;
case InfoQuery::InfoType::CONSTRAINT:
// TODO: This should be CONSTRAINT | STATS, but we don't have support
// for *or* with privileges.
AddPrivilege(AuthQuery::Privilege::CONSTRAINT);
break;
}
}
void Visit(ConstraintQuery &constraint_query) override { AddPrivilege(AuthQuery::Privilege::CONSTRAINT); }
void Visit(CypherQuery &query) override {
query.single_query_->Accept(*this);
for (auto *cypher_union : query.cypher_unions_) {
cypher_union->Accept(*this);
}
}
void Visit(DumpQuery &dump_query) override { AddPrivilege(AuthQuery::Privilege::DUMP); }
void Visit(LockPathQuery &lock_path_query) override { AddPrivilege(AuthQuery::Privilege::DURABILITY); }
void Visit(FreeMemoryQuery &free_memory_query) override { AddPrivilege(AuthQuery::Privilege::FREE_MEMORY); }
void Visit(TriggerQuery &trigger_query) override { AddPrivilege(AuthQuery::Privilege::TRIGGER); }
void Visit(StreamQuery &stream_query) override { AddPrivilege(AuthQuery::Privilege::STREAM); }
void Visit(ReplicationQuery &replication_query) override { AddPrivilege(AuthQuery::Privilege::REPLICATION); }
void Visit(IsolationLevelQuery &isolation_level_query) override { AddPrivilege(AuthQuery::Privilege::CONFIG); }
void Visit(CreateSnapshotQuery &create_snapshot_query) override { AddPrivilege(AuthQuery::Privilege::DURABILITY); }
void Visit(SettingQuery & /*setting_query*/) override { AddPrivilege(AuthQuery::Privilege::CONFIG); }
void Visit(VersionQuery & /*version_query*/) override { AddPrivilege(AuthQuery::Privilege::STATS); }
void Visit(SchemaQuery & /*schema_query*/) override { AddPrivilege(AuthQuery::Privilege::SCHEMA); }
bool PreVisit(Create & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::CREATE);
return false;
}
bool PreVisit(CallProcedure &procedure) override {
const auto maybe_proc =
procedure::FindProcedure(procedure::gModuleRegistry, procedure.procedure_name_, utils::NewDeleteResource());
if (maybe_proc && maybe_proc->second->info.required_privilege) {
AddPrivilege(*maybe_proc->second->info.required_privilege);
}
return false;
}
bool PreVisit(Delete & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::DELETE);
return false;
}
bool PreVisit(Match & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::MATCH);
return false;
}
bool PreVisit(Merge & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::MERGE);
return false;
}
bool PreVisit(SetProperty & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::SET);
return false;
}
bool PreVisit(SetProperties & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::SET);
return false;
}
bool PreVisit(SetLabels & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::SET);
return false;
}
bool PreVisit(RemoveProperty & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::REMOVE);
return false;
}
bool PreVisit(RemoveLabels & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::REMOVE);
return false;
}
bool PreVisit(LoadCsv & /*unused*/) override {
AddPrivilege(AuthQuery::Privilege::READ_FILE);
return false;
}
bool Visit(Identifier & /*unused*/) override { return true; }
bool Visit(PrimitiveLiteral & /*unused*/) override { return true; }
bool Visit(ParameterLookup & /*unused*/) override { return true; }
private:
void AddPrivilege(AuthQuery::Privilege privilege) {
if (!utils::Contains(privileges_, privilege)) {
privileges_.push_back(privilege);
}
}
std::vector<AuthQuery::Privilege> privileges_;
};
std::vector<AuthQuery::Privilege> GetRequiredPrivileges(Query *query) {
PrivilegeExtractor extractor;
query->Accept(extractor);
return extractor.privileges();
}
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,18 @@
// Copyright 2022 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.
#pragma once
#include "query/v2/frontend/ast/ast.hpp"
namespace memgraph::query::v2 {
std::vector<AuthQuery::Privilege> GetRequiredPrivileges(Query *query);
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,535 @@
// Copyright 2022 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 "query/v2/frontend/stripped.hpp"
#include <cctype>
#include <cstdint>
#include <iostream>
#include <span>
#include <string>
#include <vector>
#include "expr/parsing.hpp"
#include "parser/opencypher/generated/MemgraphCypher.h"
#include "parser/opencypher/generated/MemgraphCypherBaseVisitor.h"
#include "parser/opencypher/generated/MemgraphCypherLexer.h"
#include "parser/stripped_lexer_constants.hpp"
#include "query/v2/exceptions.hpp"
#include "utils/fnv.hpp"
#include "utils/logging.hpp"
#include "utils/string.hpp"
namespace memgraph::query::v2::frontend {
using namespace parser::lexer_constants; // NOLINT(google-build-using-namespace)
StrippedQuery::StrippedQuery(const std::string &query) : original_(query) {
enum class Token {
UNMATCHED,
KEYWORD, // Including true, false and null.
SPECIAL, // +, .., +=, (, { and so on.
STRING,
INT, // Decimal, octal and hexadecimal.
REAL,
PARAMETER,
ESCAPED_NAME,
UNESCAPED_NAME,
SPACE
};
std::vector<std::pair<Token, std::string>> tokens;
std::string unstripped_chunk;
for (int i = 0; i < static_cast<int>(original_.size());) {
Token token = Token::UNMATCHED;
int len = 0;
auto update = [&](int new_len, Token new_token) {
if (new_len > len) {
len = new_len;
token = new_token;
}
};
update(MatchKeyword(i), Token::KEYWORD);
update(MatchSpecial(i), Token::SPECIAL);
update(MatchString(i), Token::STRING);
update(MatchDecimalInt(i), Token::INT);
update(MatchOctalInt(i), Token::INT);
update(MatchHexadecimalInt(i), Token::INT);
update(MatchReal(i), Token::REAL);
update(MatchParameter(i), Token::PARAMETER);
update(MatchEscapedName(i), Token::ESCAPED_NAME);
update(MatchUnescapedName(i), Token::UNESCAPED_NAME);
update(MatchWhitespaceAndComments(i), Token::SPACE);
if (token == Token::UNMATCHED) throw LexingException("Invalid query.");
tokens.emplace_back(token, original_.substr(i, len));
i += len;
// If we notice execute, we possibly create a trigger which has defined statements.
// The statements will be parsed separately later on so we skip it for now.
if (utils::IEquals(tokens.back().second, "execute")) {
// check if it's CREATE TRIGGER query
std::span token_span{tokens};
// query could start with spaces and/or comments
if (token_span.front().first == Token::SPACE) {
token_span = token_span.subspan(1);
}
// we need to check that first and third elements are correct keywords
// CREATE<SPACE>TRIGGER<SPACE>trigger-name...EXECUTE
// trigger-name (5th element) can also be "execute" so we verify that the size is larger than 5
if (token_span.size() > 5 && utils::IEquals(token_span[0].second, "create") &&
utils::IEquals(token_span[2].second, "trigger")) {
unstripped_chunk = original_.substr(i);
break;
}
}
}
std::vector<std::string> token_strings;
// A helper function that stores literal and its token position in a
// literals_. In stripped query text literal is replaced with a new_value.
// new_value can be any value that is lexed as a literal.
auto replace_stripped = [this, &token_strings](int position, const auto &value, const std::string &new_value) {
literals_.Add(position, storage::v3::PropertyValue(value));
token_strings.push_back(new_value);
};
// Copy original tokens because we need to use original case in named
// expressions and keywords in tokens will be lowercased in the next loop.
auto original_tokens = tokens;
// For every token in original query remember token index in stripped query.
std::vector<int> position_mapping(tokens.size(), -1);
// Convert tokens to strings, perform filtering, store literals and nonaliased
// named expressions in return.
for (int i = 0; i < static_cast<int>(tokens.size()); ++i) {
auto &token = tokens[i];
// We need to shift token index for every parameter since antlr's parser
// thinks of parameter as two tokens.
int token_index = token_strings.size() + parameters_.size();
switch (token.first) {
case Token::UNMATCHED:
LOG_FATAL("Shouldn't happen");
case Token::KEYWORD: {
// We don't strip NULL, since it can appear in special expressions
// like IS NULL and IS NOT NULL, but we strip true and false keywords.
if (utils::IEquals(token.second, "true")) {
replace_stripped(token_index, true, kStrippedBooleanToken);
} else if (utils::IEquals(token.second, "false")) {
replace_stripped(token_index, false, kStrippedBooleanToken);
} else {
token_strings.push_back(token.second);
}
} break;
case Token::SPACE:
break;
case Token::STRING:
replace_stripped(token_index, expr::ParseStringLiteral(token.second), kStrippedStringToken);
break;
case Token::INT:
replace_stripped(token_index, expr::ParseIntegerLiteral(token.second), kStrippedIntToken);
break;
case Token::REAL:
replace_stripped(token_index, expr::ParseDoubleLiteral(token.second), kStrippedDoubleToken);
break;
case Token::SPECIAL:
case Token::ESCAPED_NAME:
case Token::UNESCAPED_NAME:
token_strings.push_back(token.second);
break;
case Token::PARAMETER:
parameters_[token_index] = expr::ParseParameter(token.second);
token_strings.push_back(token.second);
break;
}
if (token.first != Token::SPACE) {
position_mapping[i] = token_index;
}
}
if (!unstripped_chunk.empty()) {
token_strings.push_back(std::move(unstripped_chunk));
}
query_ = utils::Join(token_strings, " ");
hash_ = utils::Fnv(query_);
auto it = tokens.begin();
while (it != tokens.end()) {
// Store nonaliased named expressions in returns in named_exprs_.
it = std::find_if(it, tokens.end(),
[](const std::pair<Token, std::string> &a) { return utils::IEquals(a.second, "return"); });
// There is no RETURN so there is nothing to do here.
if (it == tokens.end()) return;
// Skip RETURN;
++it;
// Now we need to parse cypherReturn production from opencypher grammar.
// Skip leading whitespaces and DISTINCT statemant if there is one.
while (it != tokens.end() && it->first == Token::SPACE) {
++it;
}
if (it != tokens.end() && utils::IEquals(it->second, "distinct")) {
++it;
}
// If the query is invalid, either antlr parser or cypher_main_visitor will
// report an error.
// TODO: we shouldn't rely on the fact that those checks will be done
// after this step. We should do them here.
while (it < tokens.end()) {
// Disregard leading whitespace
while (it != tokens.end() && it->first == Token::SPACE) {
++it;
}
// There is only whitespace, nothing to do...
if (it == tokens.end()) break;
bool has_as = false;
auto last_non_space = it;
auto jt = it;
// We should track number of opened braces and parantheses so that we can
// recognize if comma is a named expression separator or part of the
// list literal / function call.
int num_open_braces = 0;
int num_open_parantheses = 0;
int num_open_brackets = 0;
for (;
jt != tokens.end() && (jt->second != "," || num_open_braces || num_open_parantheses || num_open_brackets) &&
!utils::IEquals(jt->second, "order") && !utils::IEquals(jt->second, "skip") &&
!utils::IEquals(jt->second, "limit") && !utils::IEquals(jt->second, "union") &&
!utils::IEquals(jt->second, "query") && jt->second != ";";
++jt) {
if (jt->second == "(") {
++num_open_parantheses;
} else if (jt->second == ")") {
--num_open_parantheses;
} else if (jt->second == "[") {
++num_open_braces;
} else if (jt->second == "]") {
--num_open_braces;
} else if (jt->second == "{") {
++num_open_brackets;
} else if (jt->second == "}") {
--num_open_brackets;
}
has_as |= utils::IEquals(jt->second, "as");
if (jt->first != Token::SPACE) {
last_non_space = jt;
}
}
if (!has_as) {
// Named expression is not aliased. Save string disregarding leading and
// trailing whitespaces.
std::string s;
auto begin_token = it - tokens.begin() + original_tokens.begin();
auto end_token = last_non_space - tokens.begin() + original_tokens.begin() + 1;
for (auto kt = begin_token; kt != end_token; ++kt) {
s += kt->second;
}
named_exprs_[position_mapping[it - tokens.begin()]] = s;
}
if (jt != tokens.end() && jt->second == ",") {
// There are more named expressions.
it = jt + 1;
} else {
// We're done with this return statement
break;
}
}
}
}
std::string GetFirstUtf8Symbol(const char *_s) {
// According to
// https://stackoverflow.com/questions/16260033/reinterpret-cast-between-char-and-stduint8-t-safe
// this checks if casting from const char * to uint8_t is undefined behaviour.
static_assert(std::is_same<std::uint8_t, unsigned char>::value,
"This library requires std::uint8_t to be implemented as "
"unsigned char.");
const uint8_t *s = reinterpret_cast<const uint8_t *>(_s);
if ((*s >> 7) == 0x00) return std::string(_s, _s + 1);
if ((*s >> 5) == 0x06) {
auto *s1 = s + 1;
if ((*s1 >> 6) != 0x02) throw LexingException("Invalid character.");
return std::string(_s, _s + 2);
}
if ((*s >> 4) == 0x0e) {
auto *s1 = s + 1;
if ((*s1 >> 6) != 0x02) throw LexingException("Invalid character.");
auto *s2 = s + 2;
if ((*s2 >> 6) != 0x02) throw LexingException("Invalid character.");
return std::string(_s, _s + 3);
}
if ((*s >> 3) == 0x1e) {
auto *s1 = s + 1;
if ((*s1 >> 6) != 0x02) throw LexingException("Invalid character.");
auto *s2 = s + 2;
if ((*s2 >> 6) != 0x02) throw LexingException("Invalid character.");
auto *s3 = s + 3;
if ((*s3 >> 6) != 0x02) throw LexingException("Invalid character.");
return std::string(_s, _s + 4);
}
throw LexingException("Invalid character.");
}
// Return codepoint of first utf8 symbol and its encoded length.
std::pair<int, int> GetFirstUtf8SymbolCodepoint(const char *_s) {
static_assert(std::is_same<std::uint8_t, unsigned char>::value,
"This library requires std::uint8_t to be implemented as "
"unsigned char.");
const uint8_t *s = reinterpret_cast<const uint8_t *>(_s);
if ((*s >> 7) == 0x00) return {*s & 0x7f, 1};
if ((*s >> 5) == 0x06) {
auto *s1 = s + 1;
if ((*s1 >> 6) != 0x02) throw LexingException("Invalid character.");
return {((*s & 0x1f) << 6) | (*s1 & 0x3f), 2};
}
if ((*s >> 4) == 0x0e) {
auto *s1 = s + 1;
if ((*s1 >> 6) != 0x02) throw LexingException("Invalid character.");
auto *s2 = s + 2;
if ((*s2 >> 6) != 0x02) throw LexingException("Invalid character.");
return {((*s & 0x0f) << 12) | ((*s1 & 0x3f) << 6) | (*s2 & 0x3f), 3};
}
if ((*s >> 3) == 0x1e) {
auto *s1 = s + 1;
if ((*s1 >> 6) != 0x02) throw LexingException("Invalid character.");
auto *s2 = s + 2;
if ((*s2 >> 6) != 0x02) throw LexingException("Invalid character.");
auto *s3 = s + 3;
if ((*s3 >> 6) != 0x02) throw LexingException("Invalid character.");
return {((*s & 0x07) << 18) | ((*s1 & 0x3f) << 12) | ((*s2 & 0x3f) << 6) | (*s3 & 0x3f), 4};
}
throw LexingException("Invalid character.");
}
// From here until end of file there are functions that calculate matches for
// every possible token. Functions are more or less compatible with Cypher.g4
// grammar. Unfortunately, they contain a lof of special cases and shouldn't
// be changed without good reasons.
//
// Here be dragons, do not touch!
// ____ __
// { --.\ | .)%%%)%%
// '-._\\ | (\___ %)%%(%%(%%%
// `\\|{/ ^ _)-%(%%%%)%%;%%%
// .'^^^^^^^ /` %%)%%%%)%%%'
// //\ ) , / '%%%%(%%'
// , _.'/ `\<-- \<
// `^^^` ^^ ^^
int StrippedQuery::MatchKeyword(int start) const { return kKeywords.Match<tolower>(original_.c_str() + start); }
int StrippedQuery::MatchSpecial(int start) const { return kSpecialTokens.Match(original_.c_str() + start); }
int StrippedQuery::MatchString(int start) const {
if (original_[start] != '"' && original_[start] != '\'') return 0;
char start_char = original_[start];
for (auto *p = original_.data() + start + 1; *p; ++p) {
if (*p == start_char) return p - (original_.data() + start) + 1;
if (*p == '\\') {
++p;
if (*p == '\\' || *p == '\'' || *p == '"' || *p == 'B' || *p == 'b' || *p == 'F' || *p == 'f' || *p == 'N' ||
*p == 'n' || *p == 'R' || *p == 'r' || *p == 'T' || *p == 't') {
// Allowed escaped characters.
continue;
} else if (*p == 'U' || *p == 'u') {
int cnt = 0;
auto *r = p + 1;
while (isxdigit(*r) && cnt < 8) {
++cnt;
++r;
}
if (!*r) return 0;
if (cnt < 4) return 0;
if (cnt >= 4 && cnt < 8) {
p += 4;
}
if (cnt >= 8) {
p += 8;
}
} else {
return 0;
}
}
}
return 0;
}
int StrippedQuery::MatchDecimalInt(int start) const {
if (original_[start] == '0') return 1;
int i = start;
while (i < static_cast<int>(original_.size()) && isdigit(original_[i])) {
++i;
}
return i - start;
}
int StrippedQuery::MatchOctalInt(int start) const {
if (original_[start] != '0') return 0;
int i = start + 1;
while (i < static_cast<int>(original_.size()) && '0' <= original_[i] && original_[i] <= '7') {
++i;
}
if (i == start + 1) return 0;
return i - start;
}
int StrippedQuery::MatchHexadecimalInt(int start) const {
if (original_[start] != '0') return 0;
if (start + 1 >= static_cast<int>(original_.size())) return 0;
if (original_[start + 1] != 'x') return 0;
int i = start + 2;
while (i < static_cast<int>(original_.size()) && isxdigit(original_[i])) {
++i;
}
if (i == start + 2) return 0;
return i - start;
}
int StrippedQuery::MatchReal(int start) const {
enum class State { START, BEFORE_DOT, DOT, AFTER_DOT, E, E_MINUS, AFTER_E };
State state = State::START;
auto i = start;
while (i < static_cast<int>(original_.size())) {
if (original_[i] == '.') {
if (state != State::BEFORE_DOT && state != State::START) break;
state = State::DOT;
} else if ('0' <= original_[i] && original_[i] <= '9') {
if (state == State::START) {
state = State::BEFORE_DOT;
} else if (state == State::DOT) {
state = State::AFTER_DOT;
} else if (state == State::E || state == State::E_MINUS) {
state = State::AFTER_E;
}
} else if (original_[i] == 'e' || original_[i] == 'E') {
if (state != State::BEFORE_DOT && state != State::AFTER_DOT) break;
state = State::E;
} else if (original_[i] == '-') {
if (state != State::E) break;
state = State::E_MINUS;
} else {
break;
}
++i;
}
if (state == State::DOT) --i;
if (state == State::E) --i;
if (state == State::E_MINUS) i -= 2;
return i - start;
}
int StrippedQuery::MatchParameter(int start) const {
int len = original_.size();
if (start + 1 == len) return 0;
if (original_[start] != '$') return 0;
int max_len = 0;
max_len = std::max(max_len, MatchUnescapedName(start + 1));
max_len = std::max(max_len, MatchEscapedName(start + 1));
max_len = std::max(max_len, MatchKeyword(start + 1));
max_len = std::max(max_len, MatchDecimalInt(start + 1));
if (max_len == 0) return 0;
return 1 + max_len;
}
int StrippedQuery::MatchEscapedName(int start) const {
int len = original_.size();
int i = start;
while (i < len) {
if (original_[i] != '`') break;
int j = i + 1;
while (j < len && original_[j] != '`') {
++j;
}
if (j == len) break;
i = j + 1;
}
return i - start;
}
int StrippedQuery::MatchUnescapedName(int start) const {
auto i = start;
auto got = GetFirstUtf8SymbolCodepoint(original_.data() + i);
if (got.first >= parser::lexer_constants::kBitsetSize || !kUnescapedNameAllowedStarts[got.first]) {
return 0;
}
i += got.second;
while (i < static_cast<int>(original_.size())) {
got = GetFirstUtf8SymbolCodepoint(original_.data() + i);
if (got.first >= parser::lexer_constants::kBitsetSize || !kUnescapedNameAllowedParts[got.first]) {
break;
}
i += got.second;
}
return i - start;
}
int StrippedQuery::MatchWhitespaceAndComments(int start) const {
enum class State { OUT, IN_LINE_COMMENT, IN_BLOCK_COMMENT };
State state = State::OUT;
int i = start;
int len = original_.size();
// We need to remember at which position comment started because if we fail
// to match comment finish we have a match until comment start position.
int comment_position = -1;
while (i < len) {
if (state == State::OUT) {
auto got = GetFirstUtf8SymbolCodepoint(original_.data() + i);
if (got.first < parser::lexer_constants::kBitsetSize && kSpaceParts[got.first]) {
i += got.second;
} else if (i + 1 < len && original_[i] == '/' && original_[i + 1] == '*') {
comment_position = i;
state = State::IN_BLOCK_COMMENT;
i += 2;
} else if (i + 1 < len && original_[i] == '/' && original_[i + 1] == '/') {
comment_position = i;
if (i + 2 < len) {
// Special case for an empty line comment starting right at the end of
// the query.
state = State::IN_LINE_COMMENT;
}
i += 2;
} else {
break;
}
} else if (state == State::IN_LINE_COMMENT) {
if (original_[i] == '\n') {
state = State::OUT;
++i;
} else if (i + 1 < len && original_[i] == '\r' && original_[i + 1] == '\n') {
state = State::OUT;
i += 2;
} else if (original_[i] == '\r') {
break;
} else if (i + 1 == len) {
state = State::OUT;
++i;
} else {
++i;
}
} else if (state == State::IN_BLOCK_COMMENT) {
if (i + 1 < len && original_[i] == '*' && original_[i + 1] == '/') {
i += 2;
state = State::OUT;
} else {
++i;
}
}
}
if (state != State::OUT) return comment_position - start;
return i - start;
}
} // namespace memgraph::query::v2::frontend

View File

@@ -0,0 +1,103 @@
// Copyright 2022 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.
#pragma once
#include <string>
#include <unordered_map>
#include "query/v2/parameters.hpp"
#include "utils/fnv.hpp"
namespace memgraph::query::v2::frontend {
// Strings used to replace original tokens. Different types are replaced with
// different token.
const std::string kStrippedIntToken = "0";
const std::string kStrippedDoubleToken = "0.0";
const std::string kStrippedStringToken = "\"a\"";
const std::string kStrippedBooleanToken = "true";
/**
* StrippedQuery contains:
* * stripped query
* * literals stripped from query
* * hash of stripped query
*/
class StrippedQuery {
public:
/**
* Strips the input query and stores stripped query, stripped arguments and
* stripped query hash.
*
* @param query Input query.
*/
explicit StrippedQuery(const std::string &query);
/**
* Copy constructor is deleted because we don't want to make unnecessary
* copies of this object (copying of string and vector could be expensive)
*/
StrippedQuery(const StrippedQuery &other) = delete;
StrippedQuery &operator=(const StrippedQuery &other) = delete;
/**
* Move is allowed operation because it is not expensive and we can
* move the object after it was created.
*/
StrippedQuery(StrippedQuery &&other) = default;
StrippedQuery &operator=(StrippedQuery &&other) = default;
const std::string &query() const { return query_; }
const auto &original_query() const { return original_; }
const auto &literals() const { return literals_; }
const auto &named_expressions() const { return named_exprs_; }
const auto &parameters() const { return parameters_; }
uint64_t hash() const { return hash_; }
private:
// Return len of matched keyword if something is matched, otherwise 0.
int MatchKeyword(int start) const;
int MatchString(int start) const;
int MatchSpecial(int start) const;
int MatchDecimalInt(int start) const;
int MatchOctalInt(int start) const;
int MatchHexadecimalInt(int start) const;
int MatchReal(int start) const;
int MatchParameter(int start) const;
int MatchEscapedName(int start) const;
int MatchUnescapedName(int start) const;
int MatchWhitespaceAndComments(int start) const;
// Original query.
std::string original_;
// Stripped query.
std::string query_;
// Token positions of stripped out literals mapped to their values.
// TODO: Parameters class really doesn't provide anything interesting. This
// could be changed to std::unordered_map, but first we need to rewrite (or
// get rid of) hardcoded queries which expect Parameters.
Parameters literals_;
// Token positions of query parameters mapped to their names.
std::unordered_map<int, std::string> parameters_;
// Token positions of nonaliased named expressions in return statement mapped
// to their original (unstripped) string.
std::unordered_map<int, std::string> named_exprs_;
// Hash based on the stripped query.
uint64_t hash_;
};
} // namespace memgraph::query::v2::frontend

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,50 @@
// Copyright 2022 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.
#pragma once
#include <functional>
#include <string>
#include <unordered_map>
#include "query/v2/bindings/typed_value.hpp"
#include "storage/v3/view.hpp"
#include "utils/memory.hpp"
namespace memgraph::query::v2 {
class DbAccessor;
namespace {
const char kStartsWith[] = "STARTSWITH";
const char kEndsWith[] = "ENDSWITH";
const char kContains[] = "CONTAINS";
const char kId[] = "ID";
} // namespace
struct FunctionContext {
DbAccessor *db_accessor;
utils::MemoryResource *memory;
int64_t timestamp;
std::unordered_map<std::string, int64_t> *counters;
storage::v3::View view;
};
/// Return the function implementation with the given name.
///
/// Note, returned function signature uses C-style access to an array to allow
/// having an array stored anywhere the caller likes, as long as it is
/// contiguous in memory. Since most functions don't take many arguments, it's
/// convenient to have them stored in the calling stack frame.
std::function<TypedValue(const TypedValue *arguments, int64_t num_arguments, const FunctionContext &context)>
NameToFunction(const std::string &function_name);
} // namespace memgraph::query::v2

2540
src/query/v2/interpreter.cpp Normal file

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,445 @@
// Copyright 2022 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.
#pragma once
#include <gflags/gflags.h>
#include "query/v2/auth_checker.hpp"
#include "query/v2/bindings/cypher_main_visitor.hpp"
#include "query/v2/bindings/typed_value.hpp"
#include "query/v2/config.hpp"
#include "query/v2/context.hpp"
#include "query/v2/cypher_query_interpreter.hpp"
#include "query/v2/db_accessor.hpp"
#include "query/v2/exceptions.hpp"
#include "query/v2/frontend/ast/ast.hpp"
#include "query/v2/frontend/stripped.hpp"
#include "query/v2/metadata.hpp"
#include "query/v2/plan/operator.hpp"
#include "query/v2/plan/read_write_type_checker.hpp"
#include "query/v2/stream.hpp"
#include "query/v2/stream/streams.hpp"
#include "query/v2/trigger.hpp"
#include "storage/v3/isolation_level.hpp"
#include "storage/v3/name_id_mapper.hpp"
#include "utils/event_counter.hpp"
#include "utils/logging.hpp"
#include "utils/memory.hpp"
#include "utils/settings.hpp"
#include "utils/skip_list.hpp"
#include "utils/spin_lock.hpp"
#include "utils/thread_pool.hpp"
#include "utils/timer.hpp"
#include "utils/tsc.hpp"
namespace EventCounter {
extern const Event FailedQuery;
} // namespace EventCounter
namespace memgraph::query::v2 {
inline constexpr size_t kExecutionMemoryBlockSize = 1UL * 1024UL * 1024UL;
class AuthQueryHandler {
public:
AuthQueryHandler() = default;
virtual ~AuthQueryHandler() = default;
AuthQueryHandler(const AuthQueryHandler &) = delete;
AuthQueryHandler(AuthQueryHandler &&) = delete;
AuthQueryHandler &operator=(const AuthQueryHandler &) = delete;
AuthQueryHandler &operator=(AuthQueryHandler &&) = delete;
/// Return false if the user already exists.
/// @throw QueryRuntimeException if an error ocurred.
virtual bool CreateUser(const std::string &username, const std::optional<std::string> &password) = 0;
/// Return false if the user does not exist.
/// @throw QueryRuntimeException if an error ocurred.
virtual bool DropUser(const std::string &username) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual void SetPassword(const std::string &username, const std::optional<std::string> &password) = 0;
/// Return false if the role already exists.
/// @throw QueryRuntimeException if an error ocurred.
virtual bool CreateRole(const std::string &rolename) = 0;
/// Return false if the role does not exist.
/// @throw QueryRuntimeException if an error ocurred.
virtual bool DropRole(const std::string &rolename) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual std::vector<TypedValue> GetUsernames() = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual std::vector<TypedValue> GetRolenames() = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual std::optional<std::string> GetRolenameForUser(const std::string &username) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual std::vector<TypedValue> GetUsernamesForRole(const std::string &rolename) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual void SetRole(const std::string &username, const std::string &rolename) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual void ClearRole(const std::string &username) = 0;
virtual std::vector<std::vector<TypedValue>> GetPrivileges(const std::string &user_or_role) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual void GrantPrivilege(const std::string &user_or_role, const std::vector<AuthQuery::Privilege> &privileges) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual void DenyPrivilege(const std::string &user_or_role, const std::vector<AuthQuery::Privilege> &privileges) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual void RevokePrivilege(const std::string &user_or_role,
const std::vector<AuthQuery::Privilege> &privileges) = 0;
};
enum class QueryHandlerResult { COMMIT, ABORT, NOTHING };
class ReplicationQueryHandler {
public:
ReplicationQueryHandler() = default;
virtual ~ReplicationQueryHandler() = default;
ReplicationQueryHandler(const ReplicationQueryHandler &) = default;
ReplicationQueryHandler &operator=(const ReplicationQueryHandler &) = default;
ReplicationQueryHandler(ReplicationQueryHandler &&) = default;
ReplicationQueryHandler &operator=(ReplicationQueryHandler &&) = default;
struct Replica {
std::string name;
std::string socket_address;
ReplicationQuery::SyncMode sync_mode;
std::optional<double> timeout;
ReplicationQuery::ReplicaState state;
};
/// @throw QueryRuntimeException if an error ocurred.
virtual void SetReplicationRole(ReplicationQuery::ReplicationRole replication_role, std::optional<int64_t> port) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual ReplicationQuery::ReplicationRole ShowReplicationRole() const = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual void RegisterReplica(const std::string &name, const std::string &socket_address,
const ReplicationQuery::SyncMode sync_mode, const std::optional<double> timeout,
const std::chrono::seconds replica_check_frequency) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual void DropReplica(const std::string &replica_name) = 0;
/// @throw QueryRuntimeException if an error ocurred.
virtual std::vector<Replica> ShowReplicas() const = 0;
};
/**
* A container for data related to the preparation of a query.
*/
struct PreparedQuery {
std::vector<std::string> header;
std::vector<AuthQuery::Privilege> privileges;
std::function<std::optional<QueryHandlerResult>(AnyStream *stream, std::optional<int> n)> query_handler;
plan::ReadWriteTypeChecker::RWType rw_type;
};
/**
* Holds data shared between multiple `Interpreter` instances (which might be
* running concurrently).
*
* Users should initialize the context but should not modify it after it has
* been passed to an `Interpreter` instance.
*/
struct InterpreterContext {
explicit InterpreterContext(storage::v3::Shard *db, InterpreterConfig config,
const std::filesystem::path &data_directory);
storage::v3::Shard *db;
std::optional<double> tsc_frequency{utils::GetTSCFrequency()};
std::atomic<bool> is_shutting_down{false};
AuthQueryHandler *auth{nullptr};
AuthChecker *auth_checker{nullptr};
utils::SkipList<QueryCacheEntry> ast_cache;
utils::SkipList<PlanCacheEntry> plan_cache;
TriggerStore trigger_store;
utils::ThreadPool after_commit_trigger_pool{1};
const InterpreterConfig config;
query::v2::stream::Streams streams;
storage::v3::LabelId NameToLabelId(std::string_view label_name) {
return storage::v3::LabelId::FromUint(query_id_mapper.NameToId(label_name));
}
storage::v3::PropertyId NameToPropertyId(std::string_view property_name) {
return storage::v3::PropertyId::FromUint(query_id_mapper.NameToId(property_name));
}
storage::v3::EdgeTypeId NameToEdgeTypeId(std::string_view edge_type_name) {
return storage::v3::EdgeTypeId::FromUint(query_id_mapper.NameToId(edge_type_name));
}
private:
// TODO Replace with local map of labels, properties and edge type ids
storage::v3::NameIdMapper query_id_mapper;
};
/// Function that is used to tell all active interpreters that they should stop
/// their ongoing execution.
inline void Shutdown(InterpreterContext *context) { context->is_shutting_down.store(true, std::memory_order_release); }
class Interpreter final {
public:
explicit Interpreter(InterpreterContext *interpreter_context);
Interpreter(const Interpreter &) = delete;
Interpreter &operator=(const Interpreter &) = delete;
Interpreter(Interpreter &&) = delete;
Interpreter &operator=(Interpreter &&) = delete;
~Interpreter() { Abort(); }
struct PrepareResult {
std::vector<std::string> headers;
std::vector<query::v2::AuthQuery::Privilege> privileges;
std::optional<int> qid;
};
/**
* Prepare a query for execution.
*
* Preparing a query means to preprocess the query and save it for
* future calls of `Pull`.
*
* @throw query::v2::QueryException
*/
PrepareResult Prepare(const std::string &query, const std::map<std::string, storage::v3::PropertyValue> &params,
const std::string *username);
/**
* Execute the last prepared query and stream *all* of the results into the
* given stream.
*
* It is not possible to prepare a query once and execute it multiple times,
* i.e. `Prepare` has to be called before *every* call to `PullAll`.
*
* TStream should be a type implementing the `Stream` concept, i.e. it should
* contain the member function `void Result(const std::vector<TypedValue> &)`.
* The provided vector argument is valid only for the duration of the call to
* `Result`. The stream should make an explicit copy if it wants to use it
* further.
*
* @throw utils::BasicException
* @throw query::v2::QueryException
*/
template <typename TStream>
std::map<std::string, TypedValue> PullAll(TStream *result_stream) {
return Pull(result_stream);
}
/**
* Execute a prepared query and stream result into the given stream.
*
* TStream should be a type implementing the `Stream` concept, i.e. it should
* contain the member function `void Result(const std::vector<TypedValue> &)`.
* The provided vector argument is valid only for the duration of the call to
* `Result`. The stream should make an explicit copy if it wants to use it
* further.
*
* @param n If set, amount of rows to be pulled from result,
* otherwise all the rows are pulled.
* @param qid If set, id of the query from which the result should be pulled,
* otherwise the last query should be used.
*
* @throw utils::BasicException
* @throw query::v2::QueryException
*/
template <typename TStream>
std::map<std::string, TypedValue> Pull(TStream *result_stream, std::optional<int> n = {},
std::optional<int> qid = {});
void BeginTransaction();
void CommitTransaction();
void RollbackTransaction();
void SetNextTransactionIsolationLevel(storage::v3::IsolationLevel isolation_level);
void SetSessionIsolationLevel(storage::v3::IsolationLevel isolation_level);
/**
* Abort the current multicommand transaction.
*/
void Abort();
private:
struct QueryExecution {
std::optional<PreparedQuery> prepared_query;
utils::MonotonicBufferResource execution_memory{kExecutionMemoryBlockSize};
utils::ResourceWithOutOfMemoryException execution_memory_with_exception{&execution_memory};
std::map<std::string, TypedValue> summary;
std::vector<Notification> notifications;
explicit QueryExecution() = default;
QueryExecution(const QueryExecution &) = delete;
QueryExecution(QueryExecution &&) = default;
QueryExecution &operator=(const QueryExecution &) = delete;
QueryExecution &operator=(QueryExecution &&) = default;
~QueryExecution() {
// We should always release the execution memory AFTER we
// destroy the prepared query which is using that instance
// of execution memory.
prepared_query.reset();
execution_memory.Release();
}
};
// Interpreter supports multiple prepared queries at the same time.
// The client can reference a specific query for pull using an arbitrary qid
// which is in our case the index of the query in the vector.
// To simplify the handling of the qid we avoid modifying the vector if it
// affects the position of the currently running queries in any way.
// For example, we cannot delete the prepared query from the vector because
// every prepared query after the deleted one will be moved by one place
// making their qid not equal to the their index inside the vector.
// To avoid this, we use unique_ptr with which we manualy control construction
// and deletion of a single query execution, i.e. when a query finishes,
// we reset the corresponding unique_ptr.
std::vector<std::unique_ptr<QueryExecution>> query_executions_;
InterpreterContext *interpreter_context_;
// This cannot be std::optional because we need to move this accessor later on into a lambda capture
// which is assigned to std::function. std::function requires every object to be copyable, so we
// move this unique_ptr into a shrared_ptr.
std::unique_ptr<storage::v3::Shard::Accessor> db_accessor_;
std::optional<DbAccessor> execution_db_accessor_;
std::optional<TriggerContextCollector> trigger_context_collector_;
bool in_explicit_transaction_{false};
bool expect_rollback_{false};
std::optional<storage::v3::IsolationLevel> interpreter_isolation_level;
std::optional<storage::v3::IsolationLevel> next_transaction_isolation_level;
PreparedQuery PrepareTransactionQuery(std::string_view query_upper);
void Commit();
void AdvanceCommand();
void AbortCommand(std::unique_ptr<QueryExecution> *query_execution);
std::optional<storage::v3::IsolationLevel> GetIsolationLevelOverride();
size_t ActiveQueryExecutions() {
return std::count_if(query_executions_.begin(), query_executions_.end(),
[](const auto &execution) { return execution && execution->prepared_query; });
}
};
template <typename TStream>
std::map<std::string, TypedValue> Interpreter::Pull(TStream *result_stream, std::optional<int> n,
std::optional<int> qid) {
MG_ASSERT(in_explicit_transaction_ || !qid, "qid can be only used in explicit transaction!");
const int qid_value = qid ? *qid : static_cast<int>(query_executions_.size() - 1);
if (qid_value < 0 || qid_value >= query_executions_.size()) {
throw InvalidArgumentsException("qid", "Query with specified ID does not exist!");
}
if (n && n < 0) {
throw InvalidArgumentsException("n", "Cannot fetch negative number of results!");
}
auto &query_execution = query_executions_[qid_value];
MG_ASSERT(query_execution && query_execution->prepared_query, "Query already finished executing!");
// Each prepared query has its own summary so we need to somehow preserve
// it after it finishes executing because it gets destroyed alongside
// the prepared query and its execution memory.
std::optional<std::map<std::string, TypedValue>> maybe_summary;
try {
// Wrap the (statically polymorphic) stream type into a common type which
// the handler knows.
AnyStream stream{result_stream, &query_execution->execution_memory};
const auto maybe_res = query_execution->prepared_query->query_handler(&stream, n);
// Stream is using execution memory of the query_execution which
// can be deleted after its execution so the stream should be cleared
// first.
stream.~AnyStream();
// If the query finished executing, we have received a value which tells
// us what to do after.
if (maybe_res) {
// Save its summary
maybe_summary.emplace(std::move(query_execution->summary));
if (!query_execution->notifications.empty()) {
std::vector<TypedValue> notifications;
notifications.reserve(query_execution->notifications.size());
for (const auto &notification : query_execution->notifications) {
notifications.emplace_back(notification.ConvertToMap());
}
maybe_summary->insert_or_assign("notifications", std::move(notifications));
}
if (!in_explicit_transaction_) {
switch (*maybe_res) {
case QueryHandlerResult::COMMIT:
Commit();
break;
case QueryHandlerResult::ABORT:
Abort();
break;
case QueryHandlerResult::NOTHING:
// The only cases in which we have nothing to do are those where
// we're either in an explicit transaction or the query is such that
// a transaction wasn't started on a call to `Prepare()`.
MG_ASSERT(in_explicit_transaction_ || !db_accessor_);
break;
}
// As the transaction is done we can clear all the executions
// NOTE: we cannot clear query_execution inside the Abort and Commit
// methods as we will delete summary contained in them which we need
// after our query finished executing.
query_executions_.clear();
} else {
// We can only clear this execution as some of the queries
// in the transaction can be in unfinished state
query_execution.reset(nullptr);
}
}
} catch (const ExplicitTransactionUsageException &) {
query_execution.reset(nullptr);
throw;
} catch (const utils::BasicException &) {
EventCounter::IncrementCounter(EventCounter::FailedQuery);
AbortCommand(&query_execution);
throw;
}
if (maybe_summary) {
// return the execution summary
maybe_summary->insert_or_assign("has_more", false);
return std::move(*maybe_summary);
}
// don't return the execution summary as it's not finished
return {{"has_more", TypedValue(true)}};
}
} // namespace memgraph::query::v2

121
src/query/v2/metadata.cpp Normal file
View File

@@ -0,0 +1,121 @@
// Copyright 2022 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 "query/v2/metadata.hpp"
#include <algorithm>
#include <compare>
#include <string>
#include <string_view>
namespace memgraph::query::v2 {
namespace {
using namespace std::literals;
constexpr std::string_view GetSeverityLevelString(const SeverityLevel level) {
switch (level) {
case SeverityLevel::INFO:
return "INFO"sv;
case SeverityLevel::WARNING:
return "WARNING"sv;
}
}
constexpr std::string_view GetCodeString(const NotificationCode code) {
switch (code) {
case NotificationCode::CREATE_CONSTRAINT:
return "CreateConstraint"sv;
case NotificationCode::CREATE_INDEX:
return "CreateIndex"sv;
case NotificationCode::CREATE_STREAM:
return "CreateStream"sv;
case NotificationCode::CREATE_SCHEMA:
return "CreateSchema"sv;
case NotificationCode::CHECK_STREAM:
return "CheckStream"sv;
case NotificationCode::CREATE_TRIGGER:
return "CreateTrigger"sv;
case NotificationCode::DROP_CONSTRAINT:
return "DropConstraint"sv;
case NotificationCode::DROP_REPLICA:
return "DropReplica"sv;
case NotificationCode::DROP_INDEX:
return "DropIndex"sv;
case NotificationCode::DROP_SCHEMA:
return "DropSchema"sv;
case NotificationCode::DROP_STREAM:
return "DropStream"sv;
case NotificationCode::DROP_TRIGGER:
return "DropTrigger"sv;
case NotificationCode::EXISTANT_CONSTRAINT:
return "ConstraintAlreadyExists"sv;
case NotificationCode::EXISTANT_INDEX:
return "IndexAlreadyExists"sv;
case NotificationCode::LOAD_CSV_TIP:
return "LoadCSVTip"sv;
case NotificationCode::NONEXISTANT_INDEX:
return "IndexDoesNotExist"sv;
case NotificationCode::NONEXISTANT_CONSTRAINT:
return "ConstraintDoesNotExist"sv;
case NotificationCode::REGISTER_REPLICA:
return "RegisterReplica"sv;
case NotificationCode::REPLICA_PORT_WARNING:
return "ReplicaPortWarning"sv;
case NotificationCode::SET_REPLICA:
return "SetReplica"sv;
case NotificationCode::START_STREAM:
return "StartStream"sv;
case NotificationCode::START_ALL_STREAMS:
return "StartAllStreams"sv;
case NotificationCode::STOP_STREAM:
return "StopStream"sv;
case NotificationCode::STOP_ALL_STREAMS:
return "StopAllStreams"sv;
}
}
} // namespace
Notification::Notification(SeverityLevel level) : level{level} {};
Notification::Notification(SeverityLevel level, NotificationCode code, std::string title, std::string description)
: level{level}, code{code}, title(std::move(title)), description(std::move(description)){};
Notification::Notification(SeverityLevel level, NotificationCode code, std::string title)
: level{level}, code{code}, title(std::move(title)){};
std::map<std::string, TypedValue> Notification::ConvertToMap() const {
return std::map<std::string, TypedValue>{{"severity", TypedValue(GetSeverityLevelString(level))},
{"code", TypedValue(GetCodeString(code))},
{"title", TypedValue(title)},
{"description", TypedValue(description)}};
}
std::string ExecutionStatsKeyToString(const ExecutionStats::Key key) {
switch (key) {
case ExecutionStats::Key::CREATED_NODES:
return std::string("nodes-created");
case ExecutionStats::Key::DELETED_NODES:
return std::string("nodes-deleted");
case ExecutionStats::Key::CREATED_EDGES:
return std::string("relationships-created");
case ExecutionStats::Key::DELETED_EDGES:
return std::string("relationships-deleted");
case ExecutionStats::Key::CREATED_LABELS:
return std::string("labels-added");
case ExecutionStats::Key::DELETED_LABELS:
return std::string("labels-removed");
case ExecutionStats::Key::UPDATED_PROPERTIES:
return std::string("properties-set");
}
}
} // namespace memgraph::query::v2

92
src/query/v2/metadata.hpp Normal file
View File

@@ -0,0 +1,92 @@
// Copyright 2022 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.
#pragma once
#include <cstdint>
#include <map>
#include <string>
#include <string_view>
#include <type_traits>
#include "query/v2/bindings/typed_value.hpp"
namespace memgraph::query::v2 {
enum class SeverityLevel : uint8_t { INFO, WARNING };
enum class NotificationCode : uint8_t {
CREATE_CONSTRAINT,
CREATE_INDEX,
CREATE_SCHEMA,
CHECK_STREAM,
CREATE_STREAM,
CREATE_TRIGGER,
DROP_CONSTRAINT,
DROP_INDEX,
DROP_REPLICA,
DROP_SCHEMA,
DROP_STREAM,
DROP_TRIGGER,
EXISTANT_INDEX,
EXISTANT_CONSTRAINT,
LOAD_CSV_TIP,
NONEXISTANT_INDEX,
NONEXISTANT_CONSTRAINT,
REPLICA_PORT_WARNING,
REGISTER_REPLICA,
SET_REPLICA,
START_STREAM,
START_ALL_STREAMS,
STOP_STREAM,
STOP_ALL_STREAMS,
};
struct Notification {
SeverityLevel level;
NotificationCode code;
std::string title;
std::string description;
explicit Notification(SeverityLevel level);
Notification(SeverityLevel level, NotificationCode code, std::string title, std::string description);
Notification(SeverityLevel level, NotificationCode code, std::string title);
std::map<std::string, TypedValue> ConvertToMap() const;
};
struct ExecutionStats {
public:
// All the stats have specific key to be compatible with neo4j
enum class Key : uint8_t {
CREATED_NODES,
DELETED_NODES,
CREATED_EDGES,
DELETED_EDGES,
CREATED_LABELS,
DELETED_LABELS,
UPDATED_PROPERTIES,
};
int64_t &operator[](Key key) { return counters[static_cast<size_t>(key)]; }
private:
static constexpr auto kExecutionStatsCountersSize = std::underlying_type_t<Key>(Key::UPDATED_PROPERTIES) + 1;
public:
std::array<int64_t, kExecutionStatsCountersSize> counters{0};
};
std::string ExecutionStatsKeyToString(ExecutionStats::Key key);
} // namespace memgraph::query::v2

View File

@@ -0,0 +1,71 @@
// Copyright 2022 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.
#pragma once
#include <algorithm>
#include <utility>
#include <vector>
#include "storage/v3/property_value.hpp"
#include "utils/logging.hpp"
/**
* Encapsulates user provided parameters (and stripped literals)
* and provides ways of obtaining them by position.
*/
namespace memgraph::query::v2 {
struct Parameters {
public:
/**
* Adds a value to the stripped arguments under a token position.
*
* @param position Token position in query of value.
* @param value
*/
void Add(int position, const storage::v3::PropertyValue &value) { storage_.emplace_back(position, value); }
/**
* Returns the value found for the given token position.
*
* @param position Token position in query of value.
* @return Value for the given token position.
*/
const storage::v3::PropertyValue &AtTokenPosition(int position) const {
auto found = std::find_if(storage_.begin(), storage_.end(), [&](const auto &a) { return a.first == position; });
MG_ASSERT(found != storage_.end(), "Token position must be present in container");
return found->second;
}
/**
* Returns the position-th stripped value. Asserts that this
* container has at least (position + 1) elements.
*
* @param position Which stripped param is sought.
* @return Token position and value for sought param.
*/
const std::pair<int, storage::v3::PropertyValue> &At(int position) const {
MG_ASSERT(position < static_cast<int>(storage_.size()), "Invalid position");
return storage_[position];
}
/** Returns the number of arguments in this container */
auto size() const { return storage_.size(); }
auto begin() const { return storage_.begin(); }
auto end() const { return storage_.end(); }
private:
std::vector<std::pair<int, storage::v3::PropertyValue>> storage_;
};
} // namespace memgraph::query::v2

Some files were not shown because too many files have changed in this diff Show More